Exported Projects:

Front Matter

1. Introduction

2. Cellular Foundations of Neuropharmacology

3. Molecular Foundations of Neuropharmacology

4. Receptors

5. Modulation of Synaptic Transmission

6. Amino Acid Transmitters

7. Acetylcholine

8. Norepinephrine and Epinephrine

9. Dopamine

10. Serotonin (5-Hydroxytryptamine), Histamine, and Adenosine

11. Neuroactive Peptides

12. Cellular Mechanisms in Learning and Memory

13. Treating Neurological and Psychiatric Diseases

THE BIOCHEMICAL BASIS OF
NEUROPHARMACOLOGY - 8th Ed.
(2003)
Front Matter
NOTICE
The Biochemical Basis of Neuropharmacology

TITLE PAGE
THE BIOCHEMICAL BASIS OF NEUROPHARMACOLOGY

EIGHTH EDITION

JACK R. COOPER, Ph.D.

Emeritus Professor of Pharmacology
Yale University School of Medicine

FLOYD E. BLOOM, M.D.

Chairman, Department of Neuropharmacology
The Scripps Research Institute

ROBERT H. ROTH, Ph.D.

Professor of Pharmacology and Psychiatry
Yale University School of Medicine

2003

COPYRIGHT PAGE
This book is dedicated to the memory of Nicholas J. Giarman,
colleague and dear friend

Oxford New York

Auckland Bangkok Buenos Aires Cape Town Chennai
Dar es Salaam Delhi Hong Kong Istanbul Karachi Kolkata
Kuala Lumpur Madrid Melbourne Mexico City Mumbai
Nairobi Sao Paulo Shanghai Taipei Tokyo Toronto

Copyright 1970, 1974, 1978, 1982, 1986, 1991, 1996,

2003 by Oxford University Press, Inc.

Published by Oxford University Press, Inc.

198 Madison Avenue, New York, New York, 10016
http://www.oup-usa.org

Oxford is a registered trademark of Oxford University Press.

All rights reserved. No part of this publication may be

reproduced,
stored in a retrieval system, or transmitted, in any form or by
any means,
electronic, mechanical, photocopying, recording, or otherwise,
without the prior permission of Oxford University Press.

Library of Congress Cataloging-in-Publication Data

Cooper, Jack R., 1924-
The biochemical basis of neuropharmacology /
Jack R. Cooper, Floyd E. Bloom, Robert H. Roth.-8th ed.
p.; cm. Includes bibliographical references and index.
ISBN 0-19-514007-9 (cloth)-ISBN 0-19-514008-7 (pbk.)
1. Neurochemistry. 2. Neuropharmacology.
I. Bloom, Floyd E. II. Roth, Robert H., 1939-III. Title.
[DNLm. 1. Neuropharmacology. 2. Nerve Tissue-chemistry.
3. Neurotransmitters-physiology.
QV 76.5 C777b 2003] QP356.3. C66 2003 615 . 78-dc21
2002025192

987654321
Printed in the United States of America
on acid-free paper

PREFACE
Preface to the Eighth Edition

We must confess it came as somewhat of a shock when it was

pointed out that since the first edition of this book was
published in 1970, we have attempted to educate a whole
generation of neuroscience-oriented individuals. While we
modestly lower our eyes at this accomplishment, we feel a bit
frustrated these days because neuroscience has exploded so
dramatically in the past 30 years that it is becoming
increasingly difficult to know what to cover and in what detail
without swamping the reader with too much information. We
have tried to compensate for the brevity of the discussions by
providing appropriate references at the end of each chapter.

Aside from the usual updating of material, the major change

in this edition is an extensive rewriting of the chapter on
memory and learning, to emphasize that genes involved in
behavior are not immutable but their expression can be
modified by transcription factors. Thus, with respect to
learning, that old question about which is more important,
nature or nurture, genetics or environment, should be
answered with the question, which leg is more important for
walking, the left or the right?

J. R. C.
F. E. B.
R. H. R.

CONTENTS
1. Introduction, 1
2. Cellular Foundations of Neuropharmacology, 7
3. Molecular Foundations of Neuropharmacology, 39
4. Receptors, 65
5. Modulation of Synaptic Transmission, 85
6. Amino Acid Transmitters, 105
7. Acetylcholine, 151
8. Norepinephrine and Epinephrine, 181
9. Dopamine, 255
10. Serotonin (5-Hydroxytryptamine), Histamine, and
Adenosine, 271
11. Neuroactive Peptides, 321
12. Cellular Mechanisms in Learning and Memory, 357
13. Treating Neurological and Psychiatric Diseases, 373

(+/-) Show / Hide Bibliography

Send Feedback Teton Server (5.0.0) - ©2007 Teton Data Title Updates
Customer Service Systems User
800.901.5494 Send Us Your Comments Responsibilities
Training Center
What's New
1. Introduction
Neuropharmacology can be defined simply as the study of
drugs that affect nervous tissue. This, however, is not a
practical definition since a great many drugs whose
therapeutic value is extraneural can affect the nervous
system. For example, the cardiotonic drug digitalis will not
uncommonly produce central nervous system (CNS) effects
ranging from blurred vision to disorientation. For our
purposes, we must accordingly limit the scope of
neuropharmacology to those drugs specifically employed to
affect the nervous system. The domain of neuropharmacology
would thus include psychotropic drugs that affect mood and
behavior, anesthetics, sedatives, hypnotics, narcotics,
anticonvulsants, analgesics, and a variety of drugs that affect
the autonomic nervous system.

Since, with few exceptions, the precise molecular mechanism

of action of these drugs is unknown and recitations of their
absorption, metabolism, therapeutic indications, and toxic
liability can be found in most textbooks of pharmacology, we
have chosen to take a different approach to the subject. We
will concentrate on the biochemistry and physiology of
nervous tissue, emphasizing neurotransmitters, and will
introduce the neuropharmacological agents where their action
is related to the subject under discussion. Thus, a discussion
of lysergic acid diethylamide (LSD) is included in the chapter
on serotonin, and a suggested mechanism of action of the
antipsychotic drugs is found in Chapters 9 and 13.

It is not difficult to justify this focus on either real or proposed

neurotransmitters since the drugs act at junctions rather than
on the events that occur with axonal conduction or within the
cell body. Except for local anesthetics, which interact with
axonal membranes, all neuropharmacological agents whose
mechanisms of action are to some extent documented seem
to be involved primarily with synaptic events. This finding
appears to be quite logical in view of the regulatory
mechanisms in the transmission of nerve impulses. The
extent to which a neuron is depolarized or hyperpolarized will
depend largely on its excitatory and inhibitory synaptic inputs,
and these inputs must obviously involve neurotransmitters,
neuromodulators, or neurohormones. What is enormously
difficult to comprehend is the contrast between the action of a
drug on a simple neuron, which causes it either to fire or not
to fire, and the wide diversity of CNS effects, including subtle
changes in mood and behavior, which that same drug will
induce. As will become clearer in subsequent chapters, at the
molecular level, an explanation of the action of a drug is often
possible; at the cellular level, an explanation is sometimes
possible; but at the behavioral level, our ignorance is
abysmal. There is no reason to assume, for example, that a
drug that inhibits the firing of a particular neuron will
therefore produce a depressive state in an animal: there may
be dozens of unknown intermediary reactions involving
transmitters and modulators between the demonstration of
the action of a drug on a neuronal system and the ultimate
effect on behavior.

However, the fact that there are compounds with a specific

chemical structure to control a given pathological condition is
an exciting experimental finding since it suggests an approach
that the neuropharmacologist can use to clarify normal as
well as abnormal brain chemistry and physiology. For
instance, the use of drugs that affect the adrenergic nervous
system has uncovered basic and hitherto unknown neural
properties, such as the uptake, storage, and release of the
biogenic amines. Recognition of the analogy between curare
poisoning in animals and myasthenia gravis in humans led to
the understanding of the cholinergic neuromuscular
transmission problem in myasthenia gravis and to subsequent
treatment with anticholinesterases.
We have already referred to neuroactive agents involved in
synaptic transmission as neurotransmitters, neuromodulators,
and neurohormones, so definitions are now in order. Although
we can define these terms in a strict, rigid fashion, it will be
apparent, as noted later, that it is an exercise in futility to
apply these definitions to a neuroactive agent as a
classification unless one understands its activity and specifies
its locus of action. Briefly, the traditional definition of a
neurotransmitter states that the compound must be
synthesized and released presynaptically; it must mimic the
action of the endogenous compound that is released on nerve
stimulation; and, where possible, a pharmacological identity is
required where drugs that either potentiate or block
postsynaptic responses to the endogenously released agent
also act identically to the suspected neurotransmitter that is
administered. Conventionally, based on studies of
acetylcholine at the neuromuscular junction, transmitter
action was thought to be a brief and highly restricted point-
to-point process. If one takes the word modulation literally,
then a neuromodulator has no intrinsic activity but is active
only in the face of ongoing synaptic activity, where it can
modulate transmission either pre- or postsynaptically. In
many instances, however, a modulating agent does produce
changes in conductance or membrane potential. Typically,
modulatory effects involve a second-messenger system. A
neurohormone has intrinsic activity; can be released from
both neuronal and nonneuronal cells; and, most important to
the definition, travels in the circulation to act at a site distant
from its release site. Just how far a neurohormone has to
travel before it loses its neurotransmitter status and becomes
a neurohormone has never been determined.

We stated earlier that, while we could define these terms, it

would be of little use to pigeonhole known neuroactive
compounds until the site of action and the activity of the
agent were specified. For example, dopamine is a certified
neurotransmitter in the striatum, yet it is released from the
hypothalamus and travels through the hypophyseal circulation
to the pituitary, where it inhibits the release of prolactin.
Here, it obviously fits the definition of a neurohormone.
Similarly, serotonin is a neurotransmitter in the raphe nuclei,
yet at the facial motor nucleus it acts primarily as a
neuromodulator and secondarily as a transmitter. Most
peptides, with their multiple activities in the brain and gut,
are generally considered to be neuromodulators, yet
substance P fulfills the criteria of a transmitter at sensory
afferents to the dorsal horn of the spinal cord. In sum, the
plethora of exceptions to the aforementioned definitions of
transmitter, modulator, and hormone has generated
confusion in the literature. Better to describe the activity of a
neuroactive agent at a specified site rather than attempt to
give a profitless definition.

The multidisciplinary aspects of pharmacology in general are

particularly relevant in the field of neuropharmacology, where
a pure neurophysiologist or neurochemist would be severely
handicapped in elucidating drug action at the molecular level.
Since neuropharmacology is not a specific discipline with its
own technology, the neuropharmacologist should be aware of
the methodologies that are available for the total dissection of
a biological problem at all levels of resolution from the
molecular to the behavioral.

Until relatively recently, medicinal chemists felt fortunate if

they could produce a few dozen potential therapeutic agents a
year. However, in the past several years, technologies utilizing
combinatorial chemistry and high-throughput screening have
produced libraries that can yield over 5000 potential drugs a
week. It is therefore not a problem these days to develop
neuropharmacological agents. A major problem is finding
ways to get these agents, small organic molecules as well as
peptides and proteins, into the brain and into the correct
location. Further, in the case of gene replacement, it is
necessary to ensure that the gene is not only expressed but
also programmed to make just the right amount of protein. To
date, a number of novel strategies have been employed to
circumvent the blood-brain barrier (see Chapter 2) and to
prevent enzymatic destruction of the agent before it gets to
the brain. An example of one technique is the administration
of levodopa (L-DOPA) plus a peripherally acting DOPA
decarboxylase inhibitor. Unlike dopamine, DOPA via the
neutral amino acid transport system can penetrate the brain,
where it will be decarboxylated to dopamine, the neuroactive
agent that is depleted in parkinsonism (see Chapter 9). In
addition to this prodrug approach, other techniques involve
encapsulating the drug with lipids or biodegradable polymers,
coupling the drug to a molecule that possesses a specific
transport mechanism, utilizing retrograde transport to deliver
trophic factors to specific sites in the brain, or transiently
opening the blood-brain barrier (e.g., injecting a
hyperosmotic solution or a bradykinin analog). Gene transfer
is a considerably more difficult technology, which is still
developing. Virus-based vectors, such as adenovirus, herpes
simplex virus, or Salmonella, that have been rendered
nonpathogenic along with a cloned gene can be directly
injected into the target cell. This is the in vivo approach. In
the ex vivo approach to gene therapy, cells are removed from
the patient, cultured, modified with a vector and the cloned
gene, and then injected back into the patient. In both
instances, delivery of the gene to the appropriate site may
still be problematic.

In science, one measures something. One must know what to

measure, where to measure it, and how to measure it. This
sounds rather obvious, but the student should be aware that,
particularly in the neural sciences, these seemingly simple
tasks can be enormously difficult. For example, suppose one
were interested in elucidating the presumed biochemical
aberration in schizophrenia. What would one measure?
Adenosine triphosphate? Glucose? Ascorbic acid?
Unfortunately, this problem had been zealously investigated
early on by people who measured everything they could think
of, generally in the blood, in their search for differences
between normal individuals and schizophrenics. As could be
predicted, the problem was not solved. (It may be assumed,
however, that these studies produced a large population of
anemic schizophrenics with all this bloodletting.) In recent
times, it has been demonstrated that antipsychotic drugs
block a dopamine receptor (see Chapter 9). Although this
biochemical reaction takes place immediately in test tubes
containing brain tissue, patients who are given antipsychotic
medication do not show beneficial effects for about 2 weeks.
The inference, therefore, is that the drug itself and its
biochemical reaction do not produce the ameliorative effect;
rather, it is the adaptation of the brain to the presence of the
drug that is beneficial. The question then is what is this
adaptation; the answer is that we still do not know what to
measure (but see Chapter 13).

Deciding where to measure something in neuroscience is

complicated by the heterogeneity of nervous tissue: in
general, unless one has a particular axon to grind, it is
preferable to use peripheral nerves rather than the CNS.
Suburban neurochemists (they work on the peripheral
nervous system) have an easier time than their CNS
counterparts since it is a question not only of which region of
the brain to use for the test preparation but also of which of
the multitude of cell types within each area to choose. If a
project involved a study of amino acid transport in nervous
tissue, for example, would one use isolated nerve-ending
particles (synaptosomes), glial cells, neuronal cell bodies in
culture, a myelinated axon, or a ganglion cell? Up to the
present time, most investigators have used cortical brain
slices, but the obvious disadvantage of this preparation is that
one has no idea which cellular organelle takes up the amino
acid.

How to measure something is a surprisingly easy question to

answer, at least if one is dealing with simple molecules. With
the recent advances in microseparation techniques and in
fluorometric, radiometric, and immunological assays, there is
virtually nothing that cannot be measured with a high degree
of specificity and sensitivity. In this regard, one should be
careful not to overlook the classic bioassay, which tends to be
scorned by young investigators but is in fact largely
responsible for striking progress in our knowledge of both the
prostaglandins and the opiate receptors with their peptide
agonists. The major problem is with macromolecules. How
can neuronal membranes be quantified, for example, if
extraneuronal constituents are an invariable contaminant and
if markers to identify unequivocally a cellular constituent are
often lacking? The quantitative and spatial measurement of
receptors utilizing autoradiography is also a key problem (see
Chapter 4).

This discussion of measurement is meant to point out that

what appears on the surface to be the simplest part of
research can in fact be very difficult. It is vital that students
learn not to accept data without a critical appraisal of the
procedures that were employed to obtain the results. Current
trends in the neural sciences that are related to
neuropharmacology include identifying subclasses of ion
channels, utilizing molecular genetics to uncover genes whose
expression is activated or suppressed by exposure to drugs or
peptides, neural cartography (the mapping of transmitters
and neuroactive peptides in the CNS), searching for toxins
with specific effects on conduction or transmission, cloning
and characterizing receptors and ion channels, and identifying
trophic factors involved in synaptogenesis and neuronal
regulation. It can also easily be predicted that within the next
few years an intensive search will be undertaken to explain
the function of the thousands of receptor subtypes. Clearly,
neuropharmacological agents will be invaluable probes in this
search.
In this introductory chapter, we have flitted over a number of
topics relevant to neuropharmacology. Now it is time to get
down to serious business.

SELECTED REFERENCES

Nestler, E. J., S. E. Hyman, and R. C. Malenka (2001).

Molecular Neuropharmacology. McGraw-Hill, New York.

Siegel, G. J., B. W. Agranoff, R. W. Albers, S. K. Fisher, and M.

D. Uhler (1999). Basic Neurochemistry, 6th ed. Lippincott-
Raven, Philadelphia.

(+/-) Show / Hide Bibliography

Send Feedback Teton Server (5.0.0) - ©2007 Teton Data Title Updates
Customer Service Systems User
800.901.5494 Send Us Your Comments Responsibilities
Training Center
What's New
2. Cellular Foundations of
Neuropharmacology
INTRODUCTION
As we begin to consider the particular problems that underlie
the analysis of drug actions in the central nervous system, it
may be asked "Just what is so special about nervous tissue?"
Nerve cells have two special properties that distinguish them
from all other cells in the body. First, they can conduct
bioelectric signals for long distances without any loss of signal
strength. Second, they possess specific intercellular
connections with other nerve cells and with innervated tissues
such as muscles and glands. These connections determine the
types of information a neuron can receive and the range of
responses it can yield in return.

CYTOLOGY OF THE NERVE CELL

Introduction
We do not need the high resolution of the electron microscope
to identify the characteristic structural features of the nerve
cell. The classic studies of the Spanish Nobel Prize-winning
cytologist Santiago Ramon y Cajal demonstrated the
heterogeneous size and shape of neurons as individual cells.
An inescapable rule of neurocytology and neuroanatomy is
that structures have several synonymous names. So, for
example, we find that the body of the nerve cell is also called
the soma and the perikaryon literally, "the part that
surrounds the nucleus." A fundamental scheme classifies
nerve cells by the number of cytoplasmic processes they
possess. In the simplest case, the perikaryon has one
process, called an axon. The best example of this cell type is
the sensory neuron, whose perikarya occur in groups in the
sensory or dorsal root ganglia. In this case, the axon conducts
the signal, which was generated by the sensory receptor in
the skin or other organs, centrally through the dorsal root into
the spinal cord or cranial nerve nuclei. At the next step of
complexity, we find neurons possessing two processes: the
bipolar nerve cells. The sensory receptor nerve cells of the
retina, the olfactory mucosa, and the auditory nerve are of
this form, as is a class of small nerve cells of the brain known
as granule cells.

All other nerve cells tend to fall into the class known as
multipolar nerve cells. These cells possess only one axon or
efferent-conducting process, which may be short or long, be
branched or straight, and possess a recurrent or collateral
branch that feeds back onto the same type of nerve cell from
which the axon arises. Their main differences relate to the
extent and size of the receptive field of the neuron, termed
the dendrites or dendritic tree. In silver-stained preparations
for the light microscope, the branches of dendrites look like
trees in wintertime, although they may be long and smooth,
be short and complex, or bear short spines like a cactus. It is
on these dendritic branches as well as on the cell body where
the termination of axons from other neurons makes the
specialized interneuronal communication point known as the
synapse.

Regardless of their shape and size, neurons have very

characteristic cytoplasmic organelles. Neuronal cytoplasm in
the perikaryon and dendrites, but not in the axons, is rich in
rough and smooth endoplasmic reticula, connoting an
emphasis on their secretory activity. In all structural
compartments, neurons are rich in mitochondria, connoting
their dependence on high rates of adenosine triphosphate
(ATP) generation. In addition, neuronal processes, the axons
and dendrites, exhibit prominent microtubules to maintain
their polarized shapes.
The Synapse
The characteristic specialized contact zone that has been
presumptively identified as the site of functional interneuronal
communication is the synapse. It contains special organelles.
As the axon approaches the site of its termination, it exhibits
structural features not found more proximally. Most striking is
the occurrence of dilated regions of the axon (varicosities),
within which are clustered large numbers of microvesicles
(synaptic vesicles). Synaptic vesicles tend to be spherical in
shape, with diameters varying between 400 and 1200 A.
Depending on the type of fixation used, the shape and
staining properties of the vesicles can be related to their
neurotransmitter content. The nerve endings also exhibit
mitochondria but not microtubules unless the varicosity is a
"preterminal" region of an axon as it extends toward its
terminal target. One or more of these varicosities may form a
specialized contact with one or more dendritic branches
before the ultimate termination. Such endings are known as
en passant terminals. In this sense, the term nerve terminal,
or nerve ending, connotes a functional transmitting site rather
than the end of the axon.

Electron micrographs of synaptic regions in the central

nervous system reveal a specialized contact zone between the
axonal nerve ending and the postsynaptic structure (Fig. 2-1).
Cell types arising from the embryonic ectoderm often have
such specialized intercellular contact zones, which are
generally presumed to maintain the structural integrity of the
cells within a layer. In the nervous system, the specialized
contact zone at synapses has been viewed as the site of
active chemical transmission and response. This conjecture
posed substantial controversy in the era before any of the
molecules associated with presynaptic transmitter release or
postsynaptic transmitter response had been characterized.
However, the controversy now seems to have ended with the
identification of many of the specific proteins that have
precise functional properties in and on the surface of the
synaptic vesicles and at sites along the inside of the
presynaptic membrane associated with vesicle docking and
release. Similarly, the direct chemical identification at discrete
sites within the postsynaptic surface of the specialized
contacts of receptor proteins and other proteins capable of
modifying the response to neurotransmitters has given near
consensus to the functional inference of a synaptic active
zone. In some neurons, especially the single-process and
small-granule cell types, the dendrite may also be structurally
specialized to store and release transmitter.

Figure 2-1. High-power view of two nerve

terminals contacting a small dendritic spine
containing one round mitochondrion (M). At this
magnification, the synaptic vesicles can be seen
clearly, as can the zones of specialized contact
(Sy). Astrocyte processes containing glycogen (A)
can be seen. Note that the larger nerve terminal
makes a specialized contact on the small terminal
(axoaxonic) as well as on the dendrite
(axodendritic). x12,000.

Glia
A second element in the maintenance of the neuron's integrity
depends on a type of cell known as neuroglia. There are two
main types of neuroglia, together termed macroglia.

The first is called the fibrous astrocyte, a descriptive term

based on its starlike shape when viewed in the light
microscope and on the fibrous nature of its cytoplasmic
organelles, which can be seen with both light and electron
microscopy. Astrocytes are found mainly in regions of axons
and dendrites; they tend to surround and closely contact the
adventitial surface of blood vessels. Functions such as
insulation (between conducting surfaces) and organization (to
surround and separate functional units of nerve endings and
dendrites) have been empirically attributed to the astrocyte,
mainly on the basis of its structural characteristics. However,
studies of carbohydrate-metabolizing properties of astrocytes
in culture have demonstrated their capacity to accumulate
glucose, synthesize glycogen, and provide two-carbon energy
substrates to neurons. Furthermore, the link between glucose
uptake into astrocytes is also activity-dependent and
regulated by extracellular cations, as well as by at least some
neurotransmitters. These findings have led to the suggestion
that the regional localization of glucose uptake, as
demonstrated in the human brain by positron emission
tomography and in the brain of other species by 2-
deoxyglucose autoradiography, reflects primarily the uptake
of glucose by astrocytes and not by neurons.

The second major type of neuroglia is known as the

oligodendrocyte. It is called the satellite cell when it occurs
close to nerve cell bodies and the Schwann cell when it occurs
in the peripheral nervous system. The cytoplasm of the
oligodendrocyte is characterized by rough endoplasmic
reticulum, but its most prominent characteristic is the
enclosure of concentric layers of its own surface membrane
around the axon. These concentric layers come together so
closely that the oligodendrocyte cytoplasm is completely
squeezed out and the original internal surfaces of the
membrane become fused, presenting the ringlike appearance
of the myelin sheath in cross section. Along the course of an
axon, which may be many centimeters, many
oligodendrocytes are required to constitute its myelin sheath.
At the boundary between adjacent portions of the axon
covered by separate oligodendrocytes, there is an uncovered
axonal portion known as the node of Ranvier.

Many central axons and certain elements of the peripheral

autonomic nervous system do not possess myelin sheaths.
Even these axons, however, are not bare or exposed directly
to the extracellular fluid; rather, they are enclosed within
single invaginations of the astrocyte surface membrane.
Because of this close relationship between the conducting
portions of the nerve cell, its axon, and the astrocyte, it is
easy to see the origin of the proposition that the astrocyte
may contribute to the nurture of the nerve cell.

There is yet a third nonneuronal cell class in the brain, termed

microglia, that is gaining increasing importance as a
pharmacological target. Microglia are of mesodermal origin
and related to macrophages and monocytes. Some microglia
reside within the brain, often around blood vessels. During
events that cause tissue necrosis, such as stroke, trauma, or
infection, however, macrophages enter the brain and secrete
chemical signals (cytokines, see Chapter 11) to recruit
lymphocytes and leukocytes to seal off and repair the tissue
damage.

Brain Permeability Barriers

While the unique cytological characteristics of neurons and
glia are sufficient to establish the complex intercellular
relationships of the brain, there is yet another
histophysiological concept to consider. Numerous chemical
substances pass from the bloodstream into the brain at rates
that are far slower than for entry into all other organs in the
body. There are similar slow rates of transport between the
cerebrospinal fluid and the brain, although there is no good
standard in other organs against which to compare this latter
movement.

These permeability barriers appear to be the end result of

numerous contributing factors that present diffusional
obstacles to chemicals on the basis of molecular size, charge,
solubility, and specific carrier systems. The difficulty has not
been in establishing the existence of these barriers but in
determining their mechanisms. When the relatively small
protein horseradish peroxidase (molecular weight = 43,000
daltons), is injected intravenously into mice, its eventual
location within the tissue can be demonstrated
histochemically with the electron microscope. As opposed to
the easy transvascular movement of this substance across
muscle capillaries, in the brain the peroxidase molecule hardly
penetrates the continuous layer of vascular endothelial cells at
all. The endothelial cells of brain capillaries differ from those
of other tissues in that the intercellular zones of membrane
apposition are much more highly developed in the brain and
virtually continuous along all surfaces of these cells.
Furthermore, cerebral vascular endothelial cells lack
pinocytotic vesicles, considered to be the transvascular carrier
systems of both large and small molecules in other tissues.
Recent studies suggest that a very reduced transcytosis may
occur in brain capillary endothelial cells.

Since the enzyme marker can barely go through or between

the endothelial cells, an operationally defined barrier exists.
Whether the same barrier is also applicable to highly charged
lipophobic small molecules cannot be determined from these
observations. As neuropharmacologists, what concerns us
more are the factors that retard the entrance of these smaller
molecules, such as norepinephrine and serotonin, their amino
acid precursors, or drugs that affect the metabolism of these
and other neurotransmitters. Charged molecules can,
however, diffuse widely through the extracellular spaces of
the brain when permitted entry via the cerebrospinal fluid.

Astrocytes are currently thought to elicit expression of the

proteins that constitute the blood-brain barrier in cerebral
capillary endothelial cells. Recent research suggests that a
model blood-brain barrier system may be attainable under
culture conditions. So long as central astrocytes are viable,
brain fragments experimentally transplanted to vascular beds,
such as the anterior chamber of the eye, that would normally
lack a blood-brain barrier, nevertheless retain a functional
barrier when revascularized. This suggests that the properties
of the barrier reside not in the endothelial cells themselves
but in some functional response to the adjacent central
astrocytes.

Substances that have difficulty entering the brain, in general,

also have difficulty leaving it. Thus, when monoamines are
increased in concentration by blocking their catabolism (see
Chapter 9), high levels of amine persist until the inhibiting
agents are metabolized or excreted. One such excretory route
is the acid-transport system, by which the choroid plexus
and/or brain parenchymal cells actively secrete acid
catabolites, as well as drugs such as penicillin or zidovudine,
which one might like to keep in. This step can be blocked by
the drug probenecid, resulting in increased brain and
cerebrospinal fluid amine catabolite and drug levels.

The choroid plexus may also be an interface between the

peripheral vascular system and the immune response system,
where antigens for which immunosurveillance is required can
be recognized and immune responses mounted. Lymphocytes
may normally "wander" through the brain in search of
immune targets, although how they can slip through the
endothelial cell barriers without reducing their normal barrier
functions remains unclear.

Since the precise nature of these barriers still cannot be

formulated, students would be wise to avoid the "great wall of
China" concept and lean toward the possibility of a series of
variously placed, progressively selective filtration sites that
discriminate substances on the basis of several molecular
characteristics. With lipid-soluble, weak electrolytes a
characteristic of most centrally acting drugs transport occurs
by a process of passive diffusion. Thus, a drug will penetrate
the endothelial cell only in the undissociated form and at a
rate consonant with its lipid solubility and its pKa (negative
logarithm of acid ionization constant).

Specialized sets of neurons known as the circumventricular

organs exist within discrete sites along the linings of the
cerebroventricles but are functionally on the blood side of the
blood-brain barrier. Considered to be "windows" through
which the normally excluded central milieu can monitor the
components of the bloodstream, these neurons can
communicate directly with neurons well within the enclosure
of the blood-brain barrier.

BIOELECTRIC PROPERTIES OF THE NERVE

CELL

Introduction
Given these structural details, we can now turn to the second
striking feature of nerve cells, namely, their bioelectric
property. However, even for this introductory presentation, we
must understand certain basic concepts of the physical
phenomena of electricity in order to have a working
knowledge of the bioelectric characteristics of living cells.

The initial concept is that of a difference in potential existing

within a charged field, as occurs when charged particles are
separated and prevented from randomly redistributing
themselves. When a potential difference exists, the amount of
charge per unit of time that will flow between the two sites
(i.e., current flow) depends on the resistance separating
them. If the resistance tends to 0, no net current will flow
since no potential difference can exist in the absence of a
measurable resistance. If the resistance is extremely high,
only a minimal current will flow; and that will be proportional
to the electromotive force or potential difference between the
two sites. The relationship between voltage, current, and
resistance is Ohm's law: V = I R.
When we come to measuring the electrical properties of living
cells, these basic physical laws apply but with one exception.
The pioneer electrobiologists, who did their work before the
discovery and definition of the electron, developed a
convention for the flow of charges based not on the electrons
but on the flow of positive charges. Therefore, since in
biological systems the flow of charges is not carried by
electrons but by ions, the direction of flow is expressed in
terms of the movement of positive charges. To analyze the
electrical potentials of a living system, we use small
electrodes (a microprobe for detecting current flow or
potential), electronic amplifiers for increasing the size of the
current or potential, and oscilloscopes or polygraphs for
displaying the potentials observed against a time base.

Membrane Potentials
If we take two electrodes and place them on the outside of a
living cell or tissue, we will find little, if any, difference in
potential. However, if we injure a cell so as to break its
membrane or insert one ultrafine electrode across the
otherwise intact membrane, we will find a potential difference
such that the inside of the cell is 50 mV or more negative with
respect to the extracellular electrode (Fig. 2-2). This
transmembrane potential difference has been found in almost
all types of living cell in which it has been sought; such a
membrane is said to be electrically polarized. By passing
negative ions into the cell through the microelectrode (or
extracting cations), the inside can be made more negative
(hyperpolarized). If positive current is applied to the inside of
the cell, the transmembrane potential difference is decreased
and the potential is said to be depolarized. The potential
difference across the membrane of most living cells can be
accounted for by the relative distribution of the intracellular
and extracellular ions.
The extracellular fluid is particularly rich in sodium (Na) and
relatively low in potassium (K). Inside the cell, the cytoplasm
is relatively high in K content and very low in Na. While the
membrane of the cell permits K ions (K+) to flow back and
forth with relative freedom, it resists the movement of Na
ions (Na+) from the extracellular fluid to the inside of the cell.
Since K+ can cross the membrane, they tend to flow along
the concentration gradient, which is highest inside the cell.
K+ diffusion out of the cell leaves a relative negative charge
behind, owing to the negative charges of the macromolecular
proteins. As the negative charge inside the cell begins to
build, the further diffusion of K+ from inside to outside is
retarded. Eventually, an equilibrium point will be reached that
is proportional to certain physical constants and to the
relative concentrations of intracellular and extracellular K+
and chloride (Cl-) ions. These concepts of ionic diffusion
potentials across semipermeable membranes apply generally
not only to nerve and muscle but also to blood, glandular, and
other cells large enough to have their transmembrane
potential measured.

Figure 2-2. At the top is shown a hypothetical

neuron (N1) receiving a single excitatory pathway
(E) and a single inhibitory pathway (I). A
stimulating electrode (S) has been placed on the
nerve cell's axon; microelectrode 1 is extracellular
to nerve cell 1, while microelectrode 2 is in the cell
body, and microelectrode 3 is in its nerve
terminal. Microelectrode 4 records from within
postsynaptic cell 2. The potentials and current,
recorded by each of these electrodes, are
compared through a "black box" of electronics
with a distant extracellular grounded electrode and
displayed on an oscilloscope screen. When the cell

is resting and the electrode is on the outside of

is resting and the electrode is on the outside of
the cell, no potential difference is observed (1). In
the resting state, electrode 2 records a steady
potential difference between inside and outside of
approximately 250 mV (2). While recording from
electrode 2 and stimulating the inhibitory pathway,
the membrane potential is hyperpolarized during
the inhibitory postsynaptic potential (2 + I). When
recording from electrode 2 and stimulating the
excitatory pathway, a subthreshold stimulus (ST)
produces an excitatory postsynaptic potential
indicated by a brief depolarization of the resting
membrane potential (2 + E). When the excitatory
effects are sufficient to reach threshold (T), an
action potential is generated which reverses the
inside negativity to inside positivity (2 + E). On
the lower scale, potentials recorded by electrodes
3 (blue line) and 4 (black line) are compared on
the same time base following axonal stimulation of
nerve cell 1, which is assumed to be excitatory.
The point of stimulus is seen as an electrical
artifact at point S. The action potential generated
at the nerve terminal occurs after a finite lag
period due to the conduction time (C) of the axon
between the stimulating electrode and the nerve
terminal. The action potential in the nerve ending
does not directly influence postsynaptic cell 2 until
after the transmitter has been liberated and can
react with nerve cell 2's membrane, causing the
excitatory postsynaptic potential. The time
between the beginning of the action potential
recorded by microelectrode 3 and the excitatory
postsynaptic potential recorded by electrode 4 is
the time required for excitation secretion coupling
in the nerve terminal and the liberation of
sufficient transmitter to produce effects on nerve
cell 2. Electrode 5 is a patch clamp electrode
attached to neuron 2; it indicates the effects of
 attached to neuron 2; it indicates the effects of
transmitters acting through receptors located
elsewhere on the neuron's surface, mediating
channel opening (O) and closing (C) events
through intracellular second messengers.

Membrane Ion Pumps

When the nerve cell or muscle fiber can be impaled by
electrodes to record transmembrane potential, the relation
between the membrane potential and external K+
concentration can be directly tested by exchanging the
extracellular fluid for artificial solutions of varying K
concentration. When this experiment is performed on muscle
cells, we find that the membrane potential bears a linear
relationship to the external K concentration at normal to high
K concentrations but that it deviates from this linear
relationship when the external K concentration is less than
normal. To account for this discrepancy, we must reexamine
an earlier statement. While the plasma membranes of nerve
and muscle cells and other types of polarized cell are
relatively impermeable to the flow of Na ions along the high
concentration gradient from extracellular to intracellular, they
are not completely impermeable.

Radioisotope experiments can establish that a certain amount

of Na "leaks" into the resting cell from outside. The amount of
measurable Na entering the cell is sufficient to double the
intracellular Na concentration in approximately 1 hour if there
is not some opposing process to maintain the relatively low
intracellular Na concentration. The process that continuously
maintains the low intracellular Na concentration is known as
active Na transport or, colloquially, as the sodium pump. This
pump mechanism ejects Na from the inside of the cell against
the high concentration and electrical gradients forcing it in.
However, the pump does not handle Na exclusively but
requires the presence of extracellular K+. Thus, when a Na
ion is ejected from the cell, a K ion is incorporated into the
cell, giving the mechanism yet another apt name, the Na-K
ATPase exchanger.

When the external K+ concentration is near normal, the

transmembrane potential, which is based mainly on K
concentration differences, behaves as if there were actually
more extracellular K than really exists. This is because the
Na-K exchange mechanism elevates the amount of K coming
into the cell. Remember that K permeability is relatively high
and that K+ tends to diffuse out of the cell because of its
concentration gradient but to diffuse into the cell because of
charge attraction. Therefore, two factors operate to drive K+
into the cell in the presence of relatively low external K+
concentration: (1) the electrical gradient across the
membrane and (2) the Na-K pump mechanism. The latter
system could be considered electrogenic since at low external
K+ concentrations it modifies the electrical status of the
muscle membrane. Other metabolic pumps simply exchange
cationic species across the membrane and are
nonelectrogenic. The relative electrogenicity of a pump may
depend on the ratio of the exchange cations (i.e., 1:1, 2:2, or
3:2). The pump is immediately dependent on metabolic
energy and can be blocked by several metabolic poisons, such
as dinitrophenol and the rapid-acting cardiac glycoside
ouabain. Astrocytes have such pumps too.

The Uniqueness of Nerves

All that we have said regarding the transmembrane ionic
distributions applies equally to the red blood cell or glia as to
the neuron. Thus, possession of a transmembrane potential
difference is not sufficient to account for the neuron's
bioelectric properties. However, applying depolarizing currents
across the membrane can bring out the essential difference
between the red blood cell and the nerve cell. When the red
blood cell membrane is depolarized, the difference in potential
across the cell passively follows the imposed polarization.
However, when a nerve cell membrane, such as the giant
axon of an invertebrate, is depolarized from a resting value of
approximately 270 mV to approximately 210 to 215 mV, an
explosive self-limiting process occurs, namely, the action
potential. In the action potential, the transmembrane
potential is reduced not merely to 0 but beyond 0 so that the
inside of the membrane now becomes positive with respect to
the outside. This overshoot may extend for 10 to 30 mV in
the positive direction. Because of this explosive response to
an electrical depolarization, the nerve membrane is said to be
"electrically excitable," and the resultant voltage polarity shift
is the action potential. While astrocytes can also show
variable membrane potentials, they do not exhibit this
excitability.

Analysis of Action Potentials

In an elegant series of pioneering experiments that are now
classic, Hodgkin, Huxley, and Katz analyzed the various ionic
steps responsible for the action potential. When the cell
begins to depolarize in response to stimulation, the current
flow across the membrane is carried by K. As the membrane
becomes more depolarized, the resistance to Na decreases
(i.e., Na conductance increases) and more Na enters the cell
along its electrical and concentration gradients. As Na enters,
the membrane becomes more and more depolarized, which
further increases the conductance to Na and thus depolarizes
the membrane more and faster. Such conductance changes
are "voltage-dependent." This self-perpetuating process
continues, driven by the flow of Na ions moving toward their
equilibrium distribution, which should be proportional to the
original extracellular and intracellular concentrations of Na.
However, the peak of the action potential does not attain the
equilibrium potential predicted on the basis of transmembrane
Na concentrations because of a second phase of events. The
voltage-dependent increase in Na+ conductance and the
consequent depolarization also activate a voltage-dependent
K+ conductance, and K flow then increases along its
concentration gradient from inside to outside the cell. This
process restricts the height of the reversal potential since it
tends to maintain the inside negativity of the cell and begins
to reduce the membrane conductance to Na, thus making the
action potential a self-limiting phenomenon. In most nerve
axons, the action potential lasts for approximately 0.2 to 0.5
milliseconds, depending on the type of fiber and the
temperature at which it is measured.

Once the axon has been sufficiently depolarized to reach

threshold for an action potential, the wave of activity travels
at a rate proportional to the diameter of the axons (through
which the bioelectric currents will flow). In large axons, the
rate is further accelerated by the insulation provided by
myelin sheaths, restricting the flow of transmembrane
currents to the opening at the nodes of Ranvier. Therefore,
instead of the action potential propagating from minutely
contiguous sites of the membrane, the action potential in the
myelinated axon leaps from node to node. This saltatory
conduction is consequently much more rapid.

The threshold level for an all-or-none action potential is also

inversely proportional to the diameter of the axon: large
myelinated axons respond to low values of imposed
stimulating current, whereas fine and unmyelinated axons
require much greater depolarizing currents. Local anesthetics
block activation of Na conductance, preventing depolarization.

Once the threshold has been reached, a complete action

potential will not develop if it occurs too quickly after the
preceding action potential, an interval termed the refractory
phase. This phase varies for different types of excitable nerve
and muscle cell and appears to be related to the activation
process that increases Na conductance, a phenomenon that
has a finite cycling period. That is, the membrane cannot be
reactivated before a finite interval of time has occurred. K
conductance increases with the action potential and lasts
slightly longer than the activation of Na conductance. This
results in a prolonged phase of after-hyperpolarization due to
the continued redistribution of K from inside to outside the
membrane. If the axonal membrane is artificially maintained
at a transmembrane potential equal to the K+ equilibrium
potential, no after-hyperpolarization can be seen.

Ion Channels
The experiments of Hodgkin and Huxley defined the kinetics
of cation movement during nerve membrane excitation
without constraint on the mechanisms accounting for the
movement of ions through the membrane. The discovery that
drugs can selectively block cation movement and that Na+
permeability (blocked by tetrodotoxin) can be separated from
K+ permeability (blocked by tetraethylammonium [TEA])
made more detailed analysis of ion movement mechanisms
feasible. Membrane physiologists now agree that there are
several ion-specific pathways that form separate and
independent "channels" for passive movement of Na+, K+,
Ca2+, and Cl-. Thanks to several spectacular advances in the
molecular biology of ion channels, it is becoming clear that
these conductance mechanisms are actually quite complex
arrays of interacting proteins. In some cases, the channel
proteins open and close in an all-or-nothing fashion on time
scales of 0.1 to 10 milliseconds to provide aqueous channels
through the plasma membrane that ions can traverse. More
precisely, channel macromolecules can exist in several
interconvertible conformations, only one of which permits ion
movement. The conformational shifts from one form to
another are sensitive to the bioelectric fields operating on the
membrane; by facilitating or retarding the conformational
shifts, the ion channels are "gated." In this concept, Ca2+
acts at the membrane surface to alter permeability only by
virtue of the effect its charge has on the electric fields of
otherwise fixed (mainly negative) organic charges. The
altered fields in turn can gate the channels because a part of
the channel protein is able to sense the field and, thus,
modulate the conformational shifts that open or close the
gate. When ions flow across the membrane, the ionic current
changes the membrane potential and other membrane
properties.

From a variety of experimental methods, including those that

can sample single ion channels in cultured neurons and other
excitable cells, a large number of specific ion channels have
been described. The current terminology recognizes four
types of ion channel: (1) nongated, passive ion channels,
previously referred to as "leakage channels," which are
continuously open; (2) voltage-gated (i.e., voltage-sensitive)
channels, in which channel opening and closing is affected by
the membrane potential inside of the cell; (3) chemically
gated channels, the opening and closing of which is affected
by receptors on the external plasma membrane, such as
those affected by drugs and other transmitters; and (4) ion-
gated channels, whose opening and closing is affected by
shifts in intracellular ion concentrations. Ion-gated channels
are often also sensitive to membrane potential and to
external regulatory receptors, and chemically gated channels
are often also voltage-sensitive. These various modes of
interaction provide an extremely rich spectrum of responses,
thus greatly complicating what were once simple rules of
excitability and ion conductance regulation. Additional forms
of more complex ion flow regulation provide the means by
which neurons communicate to their target cells through
junctional transmission. As will be seen below, advances in
the molecular biology of the ion channels have revealed
important principles of their structure and function.

Junctional Transmission
While these ionic mechanisms appear to account adequately
for the phenomena occurring in the propagation of an action
potential down an axon, they do not per se explain what
happens when the action potential reaches the nerve ending.
At the nerve ending, the membrane of the axon is separated
from the membrane of the postjunctional nerve cell, muscle,
or gland by an intercellular space of 50 to 200 A (Fig. 2-1).
When an electrode can be placed in both the terminal axon
and the postsynaptic cell, depolarization of the nerve terminal
does not result in a direct instantaneous shift in the
transmembrane potential of the postsynaptic element, except
in cases where the connected cells are electronically coupled.
With this exception, the junctional site seldom exhibits direct
electrical excitability like the axon.

Postsynaptic Potentials
With the advent of microelectrode techniques for recording
the transmembrane potential of nerve cells in vivo, it became
possible to determine the effects of stimulation of nerve
pathways that had previously been shown to cause either
excitation or inhibition of synaptic transmission. From such
studies, Eccles (1964) observed that subthreshold excitatory
stimuli would produce postsynaptic potentials with time
durations of 2 to 20 milliseconds. The excitatory postsynaptic
potentials algebraically accumulate with other contemporary
excitatory and inhibitory postsynaptic potentials. Most
importantly, the duration of these postsynaptic potentials is
longer than can be accounted for on the basis of electrical
activity in the preterminal axon or on the electronic
conductive properties of the postsynaptic membrane (Fig. 2-
2). This latter observation combined with the fact that
synaptic sites are not directly electrically excitable provides
conclusive evidence that central synaptic transmission must
be chemical: the prolonged time course is compatible with a
rapidly released chemical transmitter whose time course of
action is terminated by local enzymes, diffusion, and reuptake
by the nerve ending.

By such experiments, it was possible to work out the basic

ionic mechanisms for inhibitory and excitatory postsynaptic
potentials. When an ideal excitatory pathway is stimulated,
the presynaptic element liberates an excitatory transmitter,
which activates ionic conductance of the postsynaptic
membrane. This response leads to an increase in one or more
transmembrane ionic conductances, depolarizing the
membrane toward the Na equilibrium potential. In the resting
state, as has already been discussed, the membrane resides
near the K equilibrium potential. If the depolarization reaches
the threshold for activating adjacent voltage-dependent
conductances, an all-or-none action potential (spike) will be
triggered. For many neurons, the axon hillock has the lowest
spike threshold. If the resultant depolarization is insufficient
to reach threshold, the cell can still discharge if additional
excitatory postsynaptic potentials summate to threshold.

The postsynaptic potential resulting from the stimulation of an

ideal inhibitory pathway to the postsynaptic cell has been
explained in terms of the fact that an inhibitory transmitter
selectively activates channels for Cl- or K+, resulting in
diffusion of ions and hyperpolarization of the membrane. This
counterbalances the excitatory postsynaptic potentials.

Because the sites of synaptic or junctional transmission are

electrically inexcitable, the postsynaptic membrane potential
can be maintained at various levels by applying current
through intracellular electrodes and changing the intracellular
concentrations of various ions. By such maneuvers, it is
possible to poise the membrane at or near the so-called
equilibrium potentials for each of the ionic species and to
determine the ionic species whose equilibrium potential
corresponds to the conductance change caused by the
synaptic transmitter. This is the most molecular test for the
identification of synaptic transmitter substance actions.
(However, certain objections can be raised to this test in
terms of those nerve endings making junctional contacts on
distal portions of the dendritic tree. Here, the postsynaptic
potentials may be incompletely transmitted to the cell body,
where the recording electrode is placed.)

On re-reading the above section, note the use of the term

ideal. It is generally considered that classically acting
neurotransmitters produce their effects on receptor-coupled
ion conductances that are voltage-independent; that is, the
receptor will alter the coupled ion channel regardless of the
membrane potential at the moment. Nevertheless, many
nonclassic transmitters seem to operate on receptors coupled
to voltage-sensitive mechanisms. Transmitters whose
receptors are associated with intracellular second-messenger
systems (e.g., activation of cyclic nucleotide synthesis)
frequently produce these more complex forms of interaction
(see Chapter 5). Similarly, many neuropeptides appear to
affect certain of their target cells by modifying responses to
other transmitters, while not showing any direct shifts in
membrane potential or conductance when tested for actions
on their own. For example, the -adrenergic actions of locus
ceruleus neurons on their central targets produce excitability
changes that depend on which other afferent systems are
activated synchronously (see Conditional Actions of
Transmitters, below).

Two other aspects of ionic mechanisms bear some mention.

First, despite our preoccupation with action potentials and
their modification, Bullock has pointed out that the most
numerous central nervous system neurons, the small single-
process type of granule-like cell, may conduct its neuronal
business within its restricted small spatial domain with no
need ever to fire a spike. Second, those neurons that do fire
spikes may sometimes do so unconventionally, using an influx
of Ca2+ ions (voltage-sensitive Ca conductance) rather than
Na+. This Ca spike may represent a mechanism to transmit
activity from the cell body out to the dendritic system and
may play a functional role in those neurons whose dendrites
can also release transmitter, such as the catecholamine cell
body nuclei. Thus, the simplified ideal version of ionic
mechanisms may be only one of many regulatory
mechanisms between connected cells. In addition, through
the use of ion-sensitive fluorescent dyes, such as fura-2, it
has been possible to demonstrate that astrocytes in long-term
cell cultures not only can exhibit dynamic ionic responses very
much like those of neurons but also can do so in waves of
coordinated activity. The student is advised to maintain an
appreciative awareness of these potentially complex
interaction systems. In the following section, we examine
some of the less classic synaptic events and their
advantageous properties.

During the 1980s, it became fashionable to accept transmitter

response data only when they were recorded from stable
neurons in vitro. In addition to slices of mammalian brains,
other preferred systems included "model" neurons (i.e.,
neuronally derived cell lines, neuron-glia hybrid cells, and
endocrine cells of normal or tumor origin previously exploited
in lieu of real neurons). As a result, many experimental
findings previously obtained with great difficulty from intact
living brains were neglected in deference to the new, simpler
preparations. Still beyond the pale, given the frenzy to work
in vitro, is the reexamination of these transmitter-related
effects in the intact nervous system, where their role in
specific circuitry operations and interactions could actually be
evaluated under living conditions.

The transductive mechanisms now considered to be part of

the normal repertoire of regulatory processes include a
variety of transmitter-regulated (transmitter-dependent) and
voltage-sensitive ionic conductances that were formerly
regarded as unusual for the mammalian central nervous
system despite their nearly simultaneous demonstration in
invertebrate ganglia, mammalian autonomic ganglia, and
selected central mammalian neurons. Most of these unusual
transmitter-regulated conductance mechanisms relate to an
unexpectedly large number of distinctive ionic conductances
for Ca2+ and K+, with more modest expansions in the
channels for Cl- or Na+ currents.

Given the onslaught of molecular biological characterizations

for ion channels (see Chapter 3), it is important for the
student to recognize that within a specific functional category
of ion conductance (i.e., Na+, K+, Ca+), there are subtypes
of functional responses that are ligand-specific and that may
be carried out by more than one ion channel protein
(ionophore) complex. These precisely defined channel
proteins can now be examined in intimate molecular detail to
dissect how drugs, toxins, and imposed voltages can alter the
excitability of a neuron.

Calcium Channels Among the multiple voltage-sensitive

Ca2+ conductances described in neurons, three are most
consistent. The first is a transient, low-threshold Ca2+
conductance (T). This Ca2+ conductance is inactive at resting
membrane potentials but is "deinactivated" by modest
hyperpolarizations, providing a feature of oscillatory behavior.
It is most frequently inhibited by Cd2+ or Co2+ and, in some
cases, by Ni, Mg, or Mn as well; it can be activated by Ba2+.
The second Ca2+ conductance channel is a slowly
inactivating, high-threshold Ca2+ conductance (L) seen
mainly in nerve terminals. The third is a transient, high-
threshold Ca conductance (N) observed in the soma and
dendrites of large neurons in the neocortex, olfactory cortex,
and hippocampal formation. The latter are blocked by Mn2+,
Co2+, and Cd2+ and activated by Ba2+ and TEA; these
responses may be inhibited functionally by endogenous
purinergic receptors. The N-type Ca2+ channels have also
been well studied in sympathetic neurons, where they are
regulated through three separate transductive pathways, each
of which may be engaged by different neurotransmitters and
their specific intracellular mechanisms (see Chapter 5).

Potassium Channels At least three types of K channel

have been described in central neurons: (1) the "A" or "A-
like" fast, transient K conductances inhibited by 4-
aminopyridine, Ba2+, or Co2+; (2) the so-called anomalous
rectifying K conductances (see below), of which the M current
(closed by cholinergic muscarinic receptors) is one example;
(3) the Ca-activated K conductances, blocked by Co, Mn, Cd,
and some neurotransmitters (see Siggins and Gruol, 1986,
and Hille, 1992). Although most data on these K+ channel
effects are pharmacological, the properties can clearly
regulate cell firing and response patterns in distinct manners.
By closing the M current, muscarinic receptors transduce
cholinergic signals into more effective depolarization, once
partial depolarization brings this channel into play.
Somatostatin can oppose this effect, forcing the M channel to
open. The latter effect may be mediated intracellularly by
second messengers derived from arachidonic acid metabolism
(see Chapter 5). By blocking the Ca-activated K channel of
central neurons, the transmitter receptors for -adrenergic
agonists, 5-hydroxytryptamine, histamine, and corticotropin-
releasing hormone enhance the ability of responsive neurons
to follow long depolarizing pulses, thereby generating longer
trains of spikes per afferent impulse.

Depending on the specific cells in which they were recognized

(even Ca-activated K channels exist in many glands and
completely nonneural cell types) and the conditions and
possible inhibitors that may have been evaluated, as many as
12 different K conductances have been proposed. For
example, many K channels are linked to second-messenger
mediation (e.g., the channels activated by GABAB receptors
and by D2 dopamine receptors on rat substantia nigra
neurons). With patch-clamp analyses of single K channels in
locus ceruleus neurons acutely isolated from 1- to 7-day-old
rats, opioids (at receptors), somatostatin, and 2-
adrenergic agonists seem to open a K channel that is not
voltage-dependent and is directly regulated by receptor
occupancy through a G protein (see Chapter 4) but with no
known intervening second messenger.

Other Ion-Specific Channels A "persistent" Na+

conductance was first observed in cerebellar Purkinje neurons
and later in hippocampal pyramidal neurons, as well as
neurons throughout the neuraxis. This conductance provides
long-lasting but low-amplitude depolarization, which does not
lead directly to neuronal firing but, rather, provides a bias
from which the conventional fast Na channels can produce full
spike initiation. Persistent Na channels are typically blocked
by tetrodotoxin and activated by TEA.

Slow Postsynaptic Potentials

Most of the postsynaptic potentials described by Eccles (1964)
were relatively short, usually 20 milliseconds or less, and
appeared to result from passive changes in ionic conductance.
Postsynaptic potentials of slow onset and several seconds'
duration have been described (Fig. 2-2), both of a
hyperpolarizing nature and of a depolarizing nature. While
such prolonged postsynaptic potentials could be the result of
either prolonged release of transmitter or persistence of the
transmitter at postsynaptic receptor sites, there is substantial
support for the possibility that slow postsynaptic potentials
could also be caused by other forms of synaptic
communication. Many of these slow synaptic potentials are
not accompanied by the expected increase in transmembrane
ionic conductances but instead by increased transmembrane
ionic impedance. Although multiple hypothetical explanations
have been offered for such responses, the actual molecular
mechanisms remain obscure. Among the more promising
leads, transmitters such as the catecholamines can activate
the synthesis of cyclic nucleotides, which in turn can activate
intraneuronal protein kinases, which can phosphorylate
specific membrane proteins (see Chapter 5). The
phosphorylation of a membrane-mounted ion channel protein
would be expected to alter its ionic permeability, and perhaps
such changes lie at the root of the membrane effects of
several types of neurotransmitter (also see Chapter 12).

Conditional Actions of Transmitters

Frequently, transmitters produce novel actions unlike those of
classically conceived transmitters. These unconventional
actions suggest that broader definitions are useful for
conceptualizing the range of regulatory signals involved in
interneuronal communication and for examining transmitter
actions. For example, when the -adrenergic effects of locus
ceruleus stimulation are examined, target cell responses no
longer adhere to standard concepts of inhibition. Rather, they
appear to fit better the designation of "biasing" or "enabling."
The latter indicates that the enabling transmitter (in this
example, norepinephrine) can enhance or amplify the
effectiveness of other transmitter actions converging on the
common target neurons during the time period of the
enabling circuit's activity (see Chapter 8). These -adrenergic
actions can enhance either excitatory or inhibitory afferents, a
general effect referred to as enabling or, more ambiguously,
modulatory. Some pharmacological actions of neuropeptides
have been described as having the opposite effect, or
disenabling (e.g., the effects of opioid peptides on the
excitatory actions of sensory transmitters within the spinal
cord) (see Chapter 11). This story is more complex
(surprised?) because neuropeptides coexist with amino acid
and amine transmitters.

To reexplore the issue of time course on the more complex

interactions, it may be useful to speak of conditional and
unconditional actions. Unconditional actions are those that a
given transmitter evokes by itself (i.e., in the absence of
other transmitters acting on the common target cell).
Conditional actions, occurring either pre- or postsynaptically,
include, but would not be limited to, the type of enhancement
that is subsumed by enabling. In such a conditional
interaction, each transmitter would act at its own pre- or
postsynaptic transmitter receptor and interact on that target
cell when both transmitters occupy their receptors
simultaneously.

Thus, there are abundant circuits, abundant transmitters,

and, for each of these, many classes of chemically coupled
systems that can transduce the effects of active transmitter
receptors. These receptors can operate either actively or
passively, conditionally or unconditionally, over a wide range
of time through nonspecific, dependent, or independent
metabolic events. Clearly, neurons have a broad but finite and
as yet incompletely characterized repertoire of molecular
responses that messenger molecules (transmitters,
hormones, and drugs) can elicit. The power of the chemical
vocabulary of such components is their combinatorial capacity
to act conditionally and coordinately and to integrate the
temporal and spatial domains within the nervous system.
Transmitter Secretion
We have already seen that the cellular machinery of the
neuron suggests that it functions as a secretory cell.
Secretion of synaptic transmitters is the activity-locked
expression of neuronal activity induced by depolarization of
the nerve terminal. Recently, it has been possible to separate
the excitation-secretion coupling process of the presynaptic
terminal into at least two distinct phases. This has been made
possible through analysis of the action of the puffer fish
poison tetrodotoxin, which blocks the electrical excitation of
the axon but does not block the release of transmitter
substance from the depolarized nerve terminal. The best of
these experiments have been performed in the giant synaptic
junctions of the squid stellate ganglion, in which the nerve
terminals are large enough to be impaled by recording and
stimulating microelectrodes and with recording from the
postsynaptic and presynaptic neurons. In this case, when
tetrodotoxin blocks conduction of action potentials down the
axon, electrical depolarization of the presynaptic terminal still
results in the appearance of an excitatory postsynaptic
potential in the ganglion neuron. Since tetrodotoxin
selectively blocks voltage-dependent Na+ conductance, the
excitation secretion must be coupled more closely to other
ions. Present evidence strongly favors the view that voltage-
sensitive Ca2+ conductance is required for transmitter
secretion. Thus, the spike-generating and conducting events
rest on voltage-dependent ion conductance changes, while
synaptic events rest on voltage-independent or voltage-
sensitive conductance.

Biochemical, ultrastructural, and physiological experiments

have led to the concept that transmitter molecules are stored
within vesicles in the nerve terminal and that the Ca-
dependent excitation-secretion coupling within the
depolarized nerve terminal requires the transient exchange of
vesicular contents into the synaptic cleft. It is unclear whether
the vesicle simply undergoes rapid fusion with the presynaptic
specialized membrane to allow the transmitter stored in the
vesicle to diffuse out or whether the process of exocytotic
release simultaneously requires insertion of the vesicle
membrane into the synaptic plasma membrane, reappearing
later by the reverse process, namely, endocytosis.
Information on the lipid and protein components of the two
types of membrane once suggested that long-term fusion-
endocytosis cycles were unlikely, but more recent data are
compatible with either fusion release or contact release. In
noradrenergic vesicles, for example, the transmitter is stored
in very high concentrations in ternary complexes involving
ATP, Ca, and possibly additional lipids or lipoproteins.
Unfortunately, neurochemically homogeneous vesicles from
central synapses have never been completely purified, and
therefore all such analyses remain somewhat open to
interpretation. For other molecules under active consideration
as neurotransmitters, storage within brain synaptic vesicles
has been extremely difficult to document chemically. The
difficulties arise from the fact that homogenization of the
brain to prepare synaptosomes disrupts both structural and
functional integrity, and under these conditions the failure to
demonstrate that amino-acid transmitters are stored in
vesicles is rationalized as uncontrollable leakage. With
electron microscopy and autoradiography, however, sites
accumulating transmitters for which there is a high-affinity,
energy-dependent uptake process can be demonstrated.
Under these conditions, authentic "synaptic terminals" are
identified, but glial processes are also labeled. The vesicle
story is further discussed in Chapter 8.

In some cases, release of the transmitter can be modulated

"presynaptically" by the neuron's own transmitter
(autoreceptors). Autoreceptors are conceived to be receptors
that are generally distributed over the surface of a neuron
and are sensitive to the transmitter secreted by that neuron.
In the case of the central dopamine-secreting neurons, such
receptors have been related to the release of the transmitter
and to its synthesis. Such effects seem to be achieved
through receptor mechanisms different from those by which
the same transmitter molecule acts postsynaptically.
Presynaptic release may also be modified (by receptors other
than auto-receptors), by coreleased neuropeptides, or by the
effects of transmitters released by other neurons in the
vicinity of the terminal or the cell body.

ANALYSIS OF MEMBRANE ACTIONS OF

DRUGS AND TRANSMITTERS IN VITRO
The development of methods for the nearly complete
functional maintenance of central neurons in vitro for several
hours, such as tissue slice preparations, and for several days
to weeks in single-cell or explant culture systems has led to a
proliferation of additional electrophysiological methods to
examine transmitter and drug action. In slice preparations,
neuronal targets can be readily localized by inspection and
intracellular electrodes can be inserted into suitably large
neurons under visual control, while additional stimulating
electrodes may activate sources of afferent circuitry within
the slice. Transmitters and drugs can then be applied to the
whole slice by superfusion within oxygenated buffered salt
solutions or more locally by micropressure pulse or
iontophoretic application methods. In many cases, excellent
intracellular recordings can be obtained for long periods of
time because there are no annoying respiratory, cardiac, or
other movements to dislodge the electrode.

When long-term intracellular recordings can be obtained, two

additional sources of information on transmitter actions can
be analyzed. In voltage-clamp analysis, the experimenter
inserts one or two electrodes into the cell and by injecting
current holds the membrane potential of the neuron at a
constant value. The cell is usually poised at a membrane
potential more negative than resting in order to prevent
spontaneous spikes. Transmitter action is monitored by the
amount of current required to keep the membrane potential
constant and thus measures transmembrane current flow
directly. However, since the membrane potential stays
constant, any of the nonlinear properties of that neuron's
response that could occur when sufficient depolarization has
occurred will be prevented.

The clamp can also be quickly changed to a new level of

membrane potential, and the neuron's responses to this shift
provide the basis for a pharmacological dissection (e.g., with
ion channel blockers or ion substitutions) of the degree to
which the effects of a transmitter can be explained as ion
dependent or voltage dependent. Many of the actions ascribed
to neuropeptides and some of those ascribed to monoamines
fit the concept of voltage dependent, since they are modest
effects at best at resting membrane potential levels, but they
emerge as more substantive effects when the responding
neuron is depolarized or hyperpolarized by other convergent
transmitters.

Another method, termed noise analysis, also examines ion

channel activity more directly than the standard in vivo
methods. This method assumes that ion channels are either
open or closed and that they switch instantaneously, and do
so independently, between the two conditions. Using
intracellular electrodes, the fluctuations in membrane
potential (or, if voltage clamping is used, the fluctuations in
membrane current flow) are analyzed statistically to infer the
conductance of individual types of channel and the mean time
they are open in the absence or presence of the transmitter
to be analyzed.

When the neurons considered to be the appropriate targets of

a specific transmitter can be maintained in long-term tissue
culture, patch-clamp analysis, the current superstar of
membrane action analysis methods, can be _applied. This
method offers the ability to study the behavior of single-ion
channels under conditions of almost unbelievable precision.
Special "fire-polished" microelectrodes are placed on the
neuron's surface, and a slight vacuum is applied to the pipette
to attain a very tight junction with the exposed surface of the
neuronal membrane, thus requiring near-nude neurons for
best application. The resulting cell-electrode junction will have
such a high electrical resistance (gigohms) that the patch of
enclosed membrane within the microelectrode's tip will be
essentially isolated from the rest of the cell. Current flowing
within that patch can then be analyzed independently of the
responses of the rest of the neuron. With state-of-the-art,
low-noise amplifiers, current flow through individual channels
can be monitored and transmitter actions evaluated in terms
of open time, amplitudes (number of channels opened), and
closing times.

In addition, with clever micromanipulations, patch clamps can

be done in three configurations. In the cell-attached mode,
the pipette is sealed to the intact cell and measurements are
made with no further physical disruption. However, further
application of slight vacuum allows the patch of enclosed
membrane to be removed from the cell but with enclosed ion
channels still viable and responsive. In the inside-out patch,
the previously intracellular surface will be on the exterior of
the sealed membrane patch and simulations of changes in
intracellular ions or, for example, catalysts of protein
phosphorylations can examine the ion channel for regulation.
It is also possible to demonstrate that, with clever handling
and further negative pressure before pulling the membrane
patch off the cell, an outside-out patch can be obtained. Here,
the original patch is ruptured, the perimeter remains
attached, and then the surrounding external membrane
segments reseal once they are excised from the cell surface.
By placing the outside surface into solutions with differing
doses of transmitter, drug, or ion channel toxin, it is possible
to analyze very discrete, single-channel pharmacology.

AN APPROACH TO
NEUROPHARMACOLOGICAL ANALYSIS
The business of analyzing bioelectrical potentials can be very
complicated, even when restricted to changes in single
neurons or to small portions of contiguous neurons; but if we
restrict our examination of centrally active drugs to the
effects on single cells, we can ask rather precise questions.
For example, does drug X act on resting membrane potential
or resistance, on an electrogenic pump, or on the Na- or K-
activation phase of the action potential? Or does it block or
modulate the effects of junctional transmission between two
specific groups of cells?

To employ the modern powerful and precise

electrophysiological tools, we must determine first the most
likely target neuron to study. Earlier neuropharmacologists
were forced to rely on much coarser tools, such as the
electroencephalogram or sensory evoked potentials measured
by electrodes placed on the scalp. Those methods, at their
best, could measure the population response of a group of
neurons to a drug, something that single-unit analysis can do
only after many single recordings are collated. Macroelectrode
methods are receiving increased attention again, since, as
noninvasive methods, they can be used to examine drug
actions clinically.

APPROACHES
If, as modern-day neuropharmacologists, we are chiefly
concerned with uncovering the mechanisms of action of drugs
in the brain, there are several avenues along which we can
organize our attack. We could choose to examine the way in
which drugs influence the perception of sensory signals by
higher integrative centers of the brain. This is compatible with
a single-neuron and ionic conductance types of analysis,
directed, say, at how drugs affect inhibitory postsynaptic
potentials. Drugs that cause convulsions, such as strychnine,
have been analyzed in this respect; but all types of inhibitory
postsynaptic potential are not affected by strychnine.

A second basic approach would be to use both

macroelectrodes and microelectrodes to compare the drug
responses of single units and populations of units in the same
brain region. This approach is clearly limited, however, unless
we understand the intimate functional relations between the
multiple types of cell found even within one region of the
brain.

A third approach is also possible. We could choose to separate

the effects of drugs between those affecting the generation of
action potential and its propagation and those acting on
junctional transmission. For this type of analysis, we must
identify the chemical synaptic transmitter for the junctions to
be studied. Many of the interpretative problems already
alluded to can be attacked through this approach. Thus, as
might be expected, there is likely to be more than one type of
excitatory and inhibitory transmitter substance, and a
convulsant drug may affect the response to one type of
inhibitory transmitter without affecting another. Moreover, a
drug may have specific regional effects in the brain if it affects
a unique synaptic transmitter there. In fact, using this
approach, it may be possible to find drug effects not directly
reflected in electrical activity at all but related more to the
catabolic or anabolic systems maintaining the required
functional levels of transmitter. We conclude this chapter by
considering the techniques for identifying the synaptic
transmitter for particular synaptic connections. The chapters
that follow are organized to present in detail our current
understanding of putative central neurotransmitter
substances.

IDENTIFICATION OF SYNAPTIC
TRANSMITTERS

Introduction
How then do we identify the substance released by nerve
endings? The entire concept of chemical junctional
transmission arose from the classic experiments of Otto
Loewi, who demonstrated chemical transmission by
transferring the ventricular fluid of a stimulated frog heart
onto a nonstimulated frog heart, thereby showing that the
effects of the nerve stimulus on the first heart were
reproduced by the chemical activity of the solution flowing
onto the second heart. Since the phenomenon of chemical
transmission originated from studies of peripheral autonomic
organs, these peripheral junctions have become convenient
model systems for central neuropharmacological analysis.

Certain interdependent criteria have been developed to

identify junctional transmitters. By common-sense analysis,
one would suspect that the most important criterion would be
that a substance suspected of being a junctional transmitter
must be demonstrated to be released from the prejunctional
nerve endings when the nerve fibers are selectively
stimulated. This criterion was relatively easily satisfied for
isolated autonomic organs in which only one or, at most, two
nerve trunks enter the tissue and the whole system can be
isolated in an organ bath. In the central nervous system,
however, satisfaction of this criterion presumes (1) that the
proper nerve trunk or set of nerve axons can be selectively
stimulated and (2) that release of the transmitter can be
detected in the amounts released by single nerve endings
after one action potential. This last subcriterion is necessary
since we wish to restrict our analysis to the first set of
activated nerve endings and not to examine the substances
released by the secondary and tertiary interneurons in the
chain, some of which might reside quite close to the primary
endings. For many years, satisfaction of the second criterion
was almost impossible because collection devices were so
large as to injure the brain and detection methods were
insufficiently sensitive. However, newer technologies, such as
in vivo microdialysis, tissue voltammetry, and antibody-coated
carbon filaments, have been applied to detect release
effectively.

Localization
Because it is difficult, if not impossible, to identify the
substance released from single nerve endings by selective
stimulation, the next best evidence would be to prove that a
suspected synaptic transmitter resides in the presynaptic
terminal of our selected nerve pathway. Normally, we would
expect the enzymes for synthesizing and catabolizing this
substance also to be in the vicinity of this nerve ending, if not
actually part of the nerve-ending cellular machinery. In the
case of neurons secreting peptides or simple amino acid
transmitters, however, these metabolic requirements may
need further consideration. To document the presence of
neurotransmitter, several types of specific cytochemical
method for both light microscopy and electron microscopy
have been developed. More commonly employed is the
biochemical population approach, which analyzes the regional
concentrations of suspected synaptic transmitter substances.
However, presence per se indicates neither releasability nor
neuroeffectiveness (e.g., acetylcholine in the nerve-free
placenta or serotonin in the enterochromaffin cell). Although
it has generally been considered that a neuron makes only
one transmitter and secretes that same substance
everywhere that synaptic release occurs, neuropeptide
exceptions to this rule have become common.
Synaptic Mimicry: Drug Injections
A third criterion arising from peripheral autonomic nervous
system analysis is that the suspected exogenous substance
mimics the action of the transmitter released by nerve
stimulation. In most pharmacological studies of the nervous
system, drugs are administered intravascularly or onto one of
the external or internal surfaces of the brain. The substances
could also be directly injected into a given region of the brain,
although the resultant structural damage would have to be
controlled and the target verified histologically. Analysis of the
effects of drugs given by each of these various gross routes of
administration is quite complex.

We know that diffusional barriers selectively retard entry from

the bloodstream of many types of molecule into the brain.
These barriers have been demonstrated for most of the
suspected central synaptic agents. In addition, we suspect
that extracellular catabolic enzymes could destroy the
transmitter as it diffuses to the postulated site of action. A
further complicating aspect of these gross methods of
administration is that the interval of time from the
administration of the agent to the recording of the response is
usually quite long (several seconds to several minutes) in
comparison with the intervals required for junctional
transmission (milliseconds). The delay in response further
reduces the likelihood of detecting the primary site of action
on one of a chain of neurons.

Microelectrophoresis
The student will now realize how important it is to have
methods of drug administration equal in sophistication to
those with which the electrical phenomena are detected. The
most practical micromethod of drug administration yet
devised is based on the principle of electrophoresis.
Micropipettes are constructed in which one or several barrels
contain an ionized solution of the chemical substance under
investigation. The substance is applied by appropriately
directing the current flow. The microelectrophoretic technique,
when applied with controls to rule out the effects of pH,
electrical current, and diffusion of the drug to neighboring
neurons, has overcome the major limitations of classic
neuropharmacological techniques. An alternative especially
effective for testing poorly ionized molecules, such as
neuropeptides, is to fill the delivery capillary with lower
concentrations and then to "puff" very small volumes onto the
test neuron by air pressure. Frequently, a multiple-barreled
electrode is constructed from which one records the
spontaneous extracellular discharges of single neurons while
other attached pipettes are utilized to apply drugs. One can
also construct an intracellular microelectrode glued to an
extracellular drug-containing multielectrode so that the
transmembrane effects of these suspected transmitter agents
can be compared with the effects of nerve pathway
stimulation. The intracellular electrode can also be used to
poise the relative polarization of the membrane and to allow
us to detect whether the applied suspected transmitter and
that released by nerve stimulation cause the membrane to
approach identical ionic equilibrium potentials.

Considerable experimentation with this technique has made

possible certain generalizations regarding the actions of each
putative neurotransmitter that has been studied. These
substances are reviewed in detail in each of the chapters that
follows. However, it should be borne in mind that certain
substances have more or less invariable actions; for example,
-aminobutyrate and glycine always inhibit, while glutamate
and aspartate always excite. Insofar as we know, these
actions arise from increased membrane conductances to Na,
K, Ca, or Cl in every case. Other substances have many kinds
of effect, depending on the nature of the cell whose receptors
are being tested. Thus, acetylcholine frequently excites but
can also inhibit, and the receptors for either response can be
nicotinic or muscarinic. Similarly, dopamine, norepinephrine,
and serotonin almost always inhibit; but they have a few
excitatory actions that are probably not completely
artifactual.

Pharmacology of Synaptic Effects

The fourth criterion for identifying a synaptic transmitter
requires identical pharmacological effects of drugs
potentiating or blocking postsynaptic responses to both the
neurally released and the administered samples. Because the
pharmacological effects are often not identical (most "classic"
blocking agents are extrapolated to the brain from effects on
peripheral autonomic organs), this fourth criterion is often
satisfied indirectly with a series of circumstantial pieces of
data. Recently, with the advent of drugs that block the
synthesis of specific transmitter agents, the pharmacology for
certain families of transmitters has been improved.

Electrophysiological analysis of drug and transmitter actions

in the central nervous system was traditionally accomplished
in terms of single-cell activity in vivo. Four types of
physiological response served as the major indices to
compare exogenously applied transmitter candidates and
drugs with the effects of endogenous transmitters: (1)
spontaneous activity, (2) orthodromic synaptically evoked
activity, (3) antidromic activity, and (4) relative responses to
independently acting excitatory or inhibitory transmitter
released from another experimental source, such as another
barrel of a multibarrel pipette. These techniques have been
most successful when applied to large neurons, whose
selected afferent pathways can be stimulated and the
transmembrane effects specifically analyzed. However, when
the unit recording techniques are applied to the intact
mammalian brain, visualization and selection of the neuron
under investigation are almost impossible. This is only one of
the reasons why ex vivo brain slice methods and short-term
tissue culture preparations, analyzed with voltage clamping
and patch clamping, have gained in popularity.

THE STEPS OF SYNAPTIC TRANSMISSION

Let us now conclude this chapter by briefly examining the
mechanisms of presumed synaptic transmission for the
mammalian central nervous system. Each step in such
transmission constitutes one of the potential sites of central
drug action (Fig. 2-3). A stimulus activates an all-or-none
action potential in a spiking axon by depolarizing its
transmembrane potential above the threshold level. The
action potential propagates unattenuated to the nerve
terminal, where ion fluxes activate a mobilization process
leading to transmitter secretion and transmission to the
postsynaptic cell. From companion biochemical experiments
(to be described in the following chapters), the transmitter
substance is believed to be stored within the microvesicles or
synaptic vesicles seen in nerve endings by electron
microscopy. In certain types of nerve junction, miniature
postsynaptic potentials can be seen in the absence of
conducted presynaptic action potentials. These miniature
potentials have a quantal effect on the postsynaptic
membrane in that occasional potentials are statistical
multiples of the smallest measurable potentials. The
biophysical quanta have been related to the synaptic vesicles,
although the proof for this relationship remains
circumstantial.

When the transmitter is released from its storage site by the

presynaptic action potential, the effects on the postsynaptic
cells cause either excitatory or inhibitory postsynaptic
potentials, depending on the nature of the postsynaptic cell
receptor for the particular transmitter agent. If sufficient
excitatory postsynaptic potentials summate temporally from
various inputs onto the cell, the postsynaptic cell will
integrate these potentials and give off its own all-or-nothing
action potential, which is then transmitted to each of its own
axon terminals, and the process continues.

In trying to solve the problem of interneuronal chemical

communication, it may be useful, nevertheless, to maintain
an open mind with regard to three dimensions by which
neuronal circuits can be characterized: (1) the spatial domain
(those areas of the brain or peripheral receptive fields that
feed onto a given cell and those areas into which that cell
sends its efferent signals); (2) the temporal domain (the time
spans over which the spatial signals are active); and (3) the
functional domain (the mechanism by which the secreted
transmitter substance operates on the receptive cell). When
the receptive cell is closely coupled in time, space, and
function to the secreting cell, almost everyone would agree
that "real" synaptic actions have occurred. When the effects
are long-lasting and widely distributed, however, many would
prefer to call this action something else, even though the
molecular agonist is stored in and released presynaptically
from neurons onto the nerve cells they contact.

Figure 2-3. Twelve steps in the synaptic

transmission process are indicated in this idealized
synaptic connection. Step 1 is transport down the
axon. Step 2 is the electrically excitable
membrane of the axon. Step 3 involves the
organelles and enzymes present in the nerve
terminal for synthesizing, storing, and releasing
the transmitter, as well as for the process of active
reuptake. Step 4 includes the enzymes present in
the extracellular space and within the glia for
catabolizing excess transmitter released from
nerve terminals. Step 5 is the postsynaptic
receptor that triggers the response of the
postsynaptic cell to the transmitter. Step 6 shows
the organelles within the postsynaptic cells which
 the organelles within the postsynaptic cells which
respond to the receptor trigger. Step 7 is the
interaction between genetic expression of the
postsynaptic nerve cell and its influences on the
cytoplasmic organelles that respond to transmitter
action. Step 8 includes the possible "plastic" steps
modifiable by events at the specialized synaptic
contact zone. Step 9 includes the electrical portion
of the nerve cell membrane that, in response to
the various transmitters, is able to integrate the
postsynaptic potentials and produces an action
potential. Step 10 is the continuation of the
information transmission by which the
postsynaptic cell sends an action potential down
its axon. Step 11, release of transmitter, is
subjected to modification by a presynaptic
(axoaxonic) synapse; in some cases, an analogous
control can be achieved between dendritic
elements. Step 12, release of the transmitter from
a nerve terminal or secreting dendritic site, may
be further subjected to modulation through
autoreceptors that respond to the transmitter
which the same secreting structure has released.
Glia (G) can accumulate (4) released transmitters.

SELECTED REFERENCES
Alexander, S. P. H and J. A. Peters (2000). Receptor and ion
channel supplement. Trends Pharmacol. Sci. 21 (Suppl.), 1-
120.

Armstrong, C. M. and B. Hille (1998). Voltage-gated ion

channels and electrical excitability. Neuron 20, 371-380.

Aston-Jones, G. and G. R. Siggins (1994). Electrophysiology.

In Psychopharmacology: The Fourth Generation of Progress
(F. E. Bloom and D. J. Kupfer, eds.), Raven Press, New York,
pp. 41-64.

Bloom, F. E. (2001). Neurohumoral transmission and the

central nervous system. In The Pharmacological Basis of
Therapeutics, 10th ed. (Hardman, J. G., Limbird, L. E. and
Gilman, A. G., Eds.). McGraw-Hill, New York, pp. 293-320.

Bullock, T. H. (1980). Spikeless neurones: where do we go

from here? Soc. Exp. Biol. Semin. Ser. 6, 269-284.

Catterall, W. A. (1993). Structure and function of voltage-

gated ion channels. Trends Neurosci. 16, 500-506.

Eccles, J. C. (1964). The Physiology of Synapses. Academic

Press, New York.

Hille, B. (1992). Ionic Channels of Excitable Membranes, 2nd

ed. Sinauer Associates, Sunderland, MA.

Hodgkin, A. L. and A. F. Huxley (1952). Currents carried by

sodium and potassium ions through the membrane of the
giant axon of Loligo. J. Physiol. 116, 449.

Katz, B. (1966). Nerve, Muscle and Synapse. McGraw-Hill,

New York.

Llinas, R. R. (1988). The intrinsic electrophysiological

properties of mammalian neurons: insights into central
nervous system function. Science 242, 1654-1660.

Loewi, O. (1921). Uber humorale Ubertragbarkeit der

Herznervenwirkung. Pflugers Arch. 189, 239.

Magistretti, P. J., L. Pellerin, and J.-L. Martin (1994). Brain

energy metabolism. an integrated cellular perspective. In
Psychopharmacology: The Fourth Generation of Progress (F.
E. Bloom and D. J. Kupfer, eds.). Raven Press, New York, pp.
657-670.

Malenka, R. C. and R. A. Nicoll (1999). Long-term

potentiation a decade of progress? Science 251, 1870-1874.

Nicoll, R. A. (1988). The coupling of neurotransmitters to ion

channels in the brain. Science 241, 545-553.

Siggins, G. R., and D. L. Gruol (1986). Synaptic mechanisms

in the vertebrate central nervous system. In Handbook of
Physiology. Intrinsic Regulatory Systems of the Brain, vol. IV
(F. E. Bloom, ed.). American Physiological Society, Bethesda,
MD, pp. 1-114.

Zigmond, M. J., F. E. Bloom, S. C. Landis, J. L. Roberts, and L.

R. Squire, eds. (1999). Fundamental Neuroscience, Academic
Press, San Diego.

(+/-) Show / Hide Bibliography

Send Feedback Teton Server (5.0.0) - ©2007 Teton Data Title Updates
Customer Service Systems User
800.901.5494 Send Us Your Comments Responsibilities
Training Center
What's New
3. Molecular Foundations of
Neuropharmacology
INTRODUCTION
Complete understanding of the basis for a drug's actions on
the brain requires knowledge of all the molecules involved.
However, until the last decade or so, most of the molecules
involved in drug actions on the nervous system were
recognized by their actions rather than their precise molecular
structures or cellular compartmentalization. Thanks to
advances in molecular biology, a growing number of these
critical molecules can be specified in highly accurate terms to
the level of their atomic structure. Nevertheless, as
neuropharmacologists, the terms of reference remain very
much the same. A drug is said to act selectively when it elicits
responses from discrete populations of cells that possess
"drug-recognizing" macromolecules, or receptors. Most drug
receptors involve sites where neurotransmitters act. Some
resemble, at the molecular level, specific molecular features
of a neurotransmitter. However, drugs may also act by
regulating intracellular enzymes critical for normal transmitter
synthesis or breakdown or removal from the extracellular
spaces of the brain. Receptors recognize drugs for a variety of
reasons, which will be explored in subsequent chapters. Once
having made that recognition, the activated receptor usually
interacts with other molecules to alter membrane properties
or intracellular metabolism. These cellular changes in turn
regulate the interactions between cells in circuits. These
circuit changes regulate the performance of functional
systems (like the sensory, motor, or vegetative control
systems) and eventually the behavior of the whole organism.

Thus, understanding the actions of drugs on the function of

the brain, whether it be in terms of single cells or behavior, is
a multilevel, multifaceted process that begins with and builds
upon the concept of molecular interactions. Even beginning
students of drug action on the nervous system will probably
accept this statement as a reasonable hypothetical principle.
In practice, however, this principle is severely compromised
because most of the molecules in a very complex organ like
the brain remain unknown.

When Watson and Crick deduced the three-dimensional,

double-helical structure of DNA in 1953, the implications for
the coding and replication of genetic information were
recognized but could not be experimentally tested. Almost 25
years of effort were required before the new biological
technology was launched. During this interval, it became clear
how to combine genes and gene fragments from multicellular
organisms with those of viruses, fungi, and bacteria to
produce new genetic instructions and novel gene products. At
last, the concept of a gene and its cellular product attained
concrete form. Almost immediately, neuroscientists, who are
always ready to exploit new technologies, began to apply
these methods to the brain.

The power of molecular biological methods is realized from

several related but independent developments: (1) the ability
to clone genetic information (i.e., to isolate a selected
segment and accurately reproduce it in large amounts), (2)
the ability to determine the nucleic acid sequence of the
selected gene segment (i.e., to read the complete molecular
structure of a gene), and (3) the ability to practice genetic
engineering (i.e., to perturb and control gene expression and
to alter the structure of gene products by chemically
modifying precise sites in the molecular structure of the
genes). Within a decade the possibilities for applying this
basic triad of powerful tools were dramatically revealed by
two additional innovative technologies: the polymerase chain
reaction, or PCR (by which large amounts of specific nucleic
acid sequences can be produced without prior purification,
cloning, or even a complete knowledge of their sequences),
and the ability to create transgenic animals (i.e., to transfer
synthetic genes into embryonic cells to make new mice, pigs,
and cows to the experimentalist's specifications). Second-
generation refinements of the latter two technologies have
allowed PCR to serve as a means to find novel gene products
within a family of genes in which some segments have been
conserved (e.g., the transmembrane domains of certain
receptors and transporters). The transgenic strategy can now
reproducibly provide mice that are good disease models,
either by creating mice lacking a specific gene (knockout
mutations) presumed to be essential for one or another
transductive pathway or by extending the original application
of transgenic mice to overexpress selected genes, such as the
mutated forms identified in human monogenic diseases like
Huntington's disease (see Chapter 13).

All of these developments have contributed to a very rapid

advance toward a truly molecular basis for the understanding
of the nervous system and the way it can be altered by drug
actions. In this chapter, we explore these molecular
foundations.

CELLULAR VARIATION
Before we can deal effectively with the critical details of
molecules that regulate the function of the nervous system
and mediate the responses to drugs that act there, we must
briefly consider how the cells of the brain differ from the other
cells of the body and what it is that allows for differences
between types of cell.

Except for erythrocytes, all mammalian cells have a nucleus

that separates the basic units of genetic information from the
cytoplasm. The cytoplasm and its organelles allow the cell to
generate energy, which the cell uses to synthesize the
structural and enzymatic molecules that give it and its
enveloping plasma membrane the functional properties by
which it contributes to the overall operation of the organism.
Liver cells, kidney cells, skin cells, white blood cells, and cells
of the nervous system all possess these basic similarities and
individual properties. In an individual organism, all of these
somatic cells (lymphocytes being the only exception),
including the cells of the central and peripheral nervous
systems, possess entirely the same basic set of genetic
information. Lymphocytes can rearrange their genes
extensively to refine the antibodies they must produce to
meet all challenges.

The total set of potential genetic instructions of an individual,

its genome, is composed of basic instructional units-the
genes-each having a specific location on a specific
chromosome. When there are multiple versions (or alleles) of
a given gene, those expressed in a given individual are
considered the genotype of that individual. Each type of
specialized cell in an individual's organs expresses a subset of
genes that encode the special structural, enzymatic, and
other functional proteins that endow the cell with its size,
shape, location, and other physiological characteristics. This
set of characteristic features expressed by a cell is termed its
phenotype.

Classically, the phenotype of the cell classes in the brain (i.e.,

neurons or glia) has been described according to its physical
features (e.g., location, size, shape, connections) and
transmitter-specific neurochemistry. Understanding the
genetic basis for cell typing affords the opportunity to explain
how neurons of the same shape and size can have different
connections or transmitters, what this diversity and
redundancy achieves, and why some neurons and neuronal
systems may be more vulnerable to destructive insults than
others. For now, we can assume that these cellular
archetypes and the almost innumerable variants within each
of the two large classes (neurons and glia) are the outward
reflection of the corresponding specific expression of subsets
of their genes. After a brief overview, additional complexities
will be described. Neuropharmacology students recently
exposed to molecular biology or biochemistry courses may
find the next few sections "old hat." However, we include this
material so that everyone reaching the end of this chapter will
have the fundamentals necessary for a full appreciation of the
starting material (see Fig. 3-1).

1. Genetic information is stored in the form of long strands of

deoxyribonucleic acid (DNA) chains and selectively expressed
in specific cells. Thus, each specialized cell attains its
specialized functional and structural status by expressing a
subset of all of its genetic instructions. To express selective
segments of the genome, the DNA-encoded information is
converted, or transcribed, into a second similar molecular
form as strands of ribonucleic acid (RNA) under the control of
special proteins and RNA-synthesizing enzymes
(polymerases) that perform the transcription steps. We will
deal in slightly more detail with the actual molecular
properties of DNA and RNA below, but the immense amount
of detail simplified here is beyond our scope.

2. The primary transcript form (the heterogeneous nuclear

RNA or hnRNA) is then edited by several rapid steps and
exported from the nucleus (through nuclear pores) to the
cytoplasm. The edited RNA transcript, or messenger RNA
(mRNA), is then translated by special cytoplasmic organelles,
also made of RNA and proteins, called ribosomes. The
translation is a chemical language shift from the nucleic acid
code of the RNA into the amino acid sequence of the protein
that is to be expressed.

3. In some brain cells, such as the glia, which make their

proteins largely for use within the cell, the ribosomes occur
freely within the cytoplasm. In neurons and other dedicated
secretory cells, the translated protein undergoes
posttranslational processing, in which the protein's structure
may be modified to attain the folded, globular, or linear
structural properties that allow it to become associated with
the proper intracellular compartments (e.g., within the plasma
membrane or within the cytoplasm) where it is intended to
function. As more and more proteins have been revealed in
molecular structure and cellular location, at least three main
types are now specified. There are two types of integral
membrane protein, with the N terminus either extracellular
(type I) or intracytoplasmic (type II), as well as peripheral
membrane proteins close to the cytoplasmic surface of the
membrane but not penetrating the lipid bilayers. There are
also wholly cytoplasmic proteins.

4. In cells, like neurons, that transport large amounts of

protein products within the cell's interior for purposes of
secretion, more extensive posttranslational processing occurs.
In such cells, the ribosomes are physically associated with a
special set of endoplasmic reticular membranes, giving the
membranes a "rough" appearance, for which they are known
as the "rough endoplasmic reticulum." Within the channels of
this inside-the-cell network, the newly synthesized proteins
are led to a set of smooth endoplasmic reticular membranes,
the Golgi apparatus, where they are packaged into secretory
organelles for transport to the secretory, or releasing,
segments of the cell. Specific domains of the mRNAs, and
hence the translation products of these secretory proteins,
allow the internal apparatus of the cell to guide the to-be-
secreted products from one compartment to another.

All of these organelle systems of the cell are essential for the
selective transcription, translation, and packaging or
compartmentalization of the specific proteins by which a given
class of cells attains its specific phenotype. These structural
elements of cell biology were well known to classic cytologists
long before the molecular mechanisms underlying these
events were understandable. Interested readers will now
begin to comprehend the special analytic advantages that
arise from the methods of molecular biology.

Figure 3-1. A schematic overview of the basic

steps and cellular compartments involved in
determining the specific phenotype of a neuron
(above) and in cloning the mRNAs that allow the
neuron to translate its genetic information into
specific proteins (below). An mRNA is converted to
a single-stranded DNA, which is further converted
into a double-stranded segment that is then
inserted into a plasmid. Bacteria are infected with
individual plasmids, and individual plasmid-
infected bacteria are grown into colonies.
Replicates of the culture plate are screened by
nucleic acid or antibody probes to select clones of
interest.

FUNDAMENTAL MOLECULAR INTERACTIONS

Introduction
The cornerstone discovery of molecular biology was the
formulation by Watson and Crick in 1953 of the double-
stranded helix model of DNA structure. The insightful model
they developed provided a coherent integration of the regular
X-ray crystallographic structure of partially purified DNA with
the previously known quantitative chemical data on the
relative frequency within DNA of its four nucleotide bases,
thus explaining why the purine-pyrimidine pairs adenine (A)-
thymidine (T) and guanine (G)-cytosine (C) occur in precisely
equal frequency. The Watson-Crick molecular model for DNA
also accurately predicted the basic mechanism of DNA
replication and accurate repair.
Nucleic Acid Base Pairing Complementarity
In the Watson-Crick double-helix, two right-handed helical
polynucleotide chains coil around the same central axis,
making a complete helical turn every 10 nucleotides (Fig. 3-
2). In the interior of the helix, the purine and pyrimidine
bases (A with T and G with C) are paired through hydrogen
bonding of their complementary structures, placing the
phosphate groups around the outside of the helix. The
structural focal point for gene expression is the precise
molecular complementarity between the primary sequences of
nucleotide bases in one strand of the DNA helix with the
antiparallel sequence of the second strand. The strand that
encodes the genetic information is termed the sense strand.
Wherever a particular base occurs in the sense strand, there
will be a complementary base, and only that base in the
antisense strand, such that A always pairs with T, and vice
versa, and G always pairs with C, and vice versa.

The base pair complementarity allows for duplication of the

genetic information in dividing cells. This is accomplished by
enzymes known as DNA polymerases, which replicate each
single strand back into double strands according to the single
strand's template. The double-stranded complementarity also
provides a means to repair the DNA should it be damaged
since whichever single strand survives the damage can act as
a template for the repair.

In a similar manner, the information-bearing, or sense, strand

of the helical DNA chain is copied into a complementary
single-stranded RNA during the process of transcription. RNA
is thus a single-stranded complementary copy of the DNA
antisense strand (so that its sequence resembles closely that
of the DNA sense strand; see Fig. 3-2). RNA differs chemically
with the substitutions of uridine for thymidine and ribose
phosphates for deoxyribose phosphates. Transcription is
accomplished by enzymes known as RNA polymerases. The
affinity of the base pairs along sequences of a single DNA
strand for their complementary base pair sequences in DNA or
RNA are so precise that small segments can be used as
probes for the detection of homologous sequences between
large domains of DNA and RNA because the molecular
complementarity will allow the probe to bind to its
complementary structure only when there is a long sequence
of consistent match. The ability of a single-stranded nucleic
acid to bind, or hybridize, to its complementary sequence is
an essential component of many molecular biological
techniques.

Figure 3-2. The arrangement of bases within the

DNA double helix and the relationships between
nucleic acid base pairs, RNA, and amino acids
during the process of transcription and translation.

The Genetic Code

To translate genetic information from sequences of RNA into
linear sequences of amino acids in proteins requires strict
coding so that the 20 or so amino acids commonly found in
proteins can be specified by various combinations of the four
nucleotides. Through an ingenious series of important
experiments, which space does not permit us to describe
here, it was demonstrated that sets of three RNA bases
(triplets) provide the code words that specify which amino
acid will be incorporated into protein according to the order of
the triplet codons in the gene and mRNA (Fig. 3-2). Other
triplet sequences mark the point at which synthesis would
begin, end, or be modified (by removing domains and linking
the remainders together, called splicing).

Often, mRNAs encode far more protein sequence than is

represented in the final form of the processed gene product.
One feature that was detected only through recent cellular
biological insight was the signal peptide, a 15- to 30-amino
acid sequence at the N-terminal end of the encoded gene
product in which the amino acids are highly hydrophobic. The
signal peptide is a near-constant feature of proteins intended
for secretion, such as neuropeptides; its function seems to be
to guide the nascent protein chain through the endoplasmic
reticular membrane for subsequent packaging by the Golgi
membrane apparatus. Proteins that will not be secreted lack
signal peptides. Thus, the structure of novel proteins predicts
whether or not they are likely to be secretory products.

DNA Segments for Genes Are Interrupted

In addition to these remarkable explanations of gene
expression and translation, another totally unanticipated
wrinkle of gene regulation soon emerged. In the late 1970s,
with the analysis of the genes encoding immunoglobulins and
hemoglobins, it was recognized that the basic organization of
genes in eukaryotic cells (cells with nuclei, as in all
multicellular organisms) did not follow the principles that had
been uncovered from the study of prokaryotes (cells, like
bacteria, that lack a separable genetic compartment).
Instead, higher organisms (and some viruses) have their
gene segments split up into coding regions from which gene
products are expressed (exons). Intervening regions of DNA
(introns) separate the exons, but the RNA versions of introns
are not found in the mRNA.

This interrupted DNA coding structure leads to two outcomes.

(1) The primary gene transcript, formally termed the hnRNA,
will contain extra RNA sequences, and the introns must be
edited out before the mRNA can successfully direct protein
synthesis by ribosomes (see Fig. 3-3). The editing process
opens up the hnRNA, removes introns, and resplices the cut
ends. (2) In some cells, including neurons, the composition of
the transcribed mRNA can also be edited by splicing out
certain exon segments. The editing-splicing process (or
"alternative splicing") provides a means by which a gene
containing several exons can give rise to several different
gene product proteins with some shared domains and some
unique domains. When gene products share similar nucleotide
and protein sequences, we often speak of them as a
"structural family" (see Chapter 11).

This editing and splicing may seem to be an unnecessarily

complicated route to follow for a process that is intended to
translate important genes with great fidelity into equally
important enzymatic and structural proteins. However, the
reader should be cautioned that such biological complexity
almost always implies important and unanticipated regulatory
control and enrichment. In the case of gene regulation and
expression, the added complexities offer the means by which
new life forms can evolve.

Figure 3-3. The relationship between DNA base

sequences in introns and exons (a-g), the
resulting primary RNA transcript, the subsequently
edited forms of mRNA (two different forms of
hypothetical editing and splicing are depicted),
and the resulting proteins, which can then
undergo posttranslational processing to yield small
peptides or to add carbohydrates or other
chemical modifications to the molecule.

Nucleic Acid Sequence Determinations

The final fundamental procedural development that
accelerated discoveries made with molecular biological
methods was the body of techniques to determine the
sequence of DNA molecules, even if they are several thousand
base pairs in length. The methods for doing this sequencing
are clearly outside the province of this book. Two different
approaches to sequence determination (that of Maxam and
Gilbert and that of Sanger), each based on some very clever
chemistry, had profound enough importance to merit Nobel
recognition. Today, these methods have been automated into
sequencer devices, speeding the process and reducing human
error. From these DNA sequences, it is possible to deduce the
nucleotide sequences of the RNA and thereby the amino acid
sequence of the protein product. The structures of the DNA
and of the gene product can then be analyzed in the
cumulative knowledge of previously determined sequences,
often via computer. In addition to structural comparison of
gene or RNA sequences, important clues to the functional
features of the encoded protein may be inferred from the
domains of hydrophobic, hydrophilic, or other consensus
structural sites for possible posttranslational modification.

Once-Over Quickly Cloning

mRNAs can also be converted into DNA by the class of virus
enzymes known as reverse transcriptases. The copied double-
stranded DNA form, or cDNA, can be incorporated (inserted)
into specific sites within an infectious vector, or plasmid. The
sites for insertion are selected by identifying DNA sequences
that can be "cut" by the actions of restriction endonucleases,
enzymes from purified bacterial sources that cleave DNA
sequences at specific palindromic repeated sequence sites.
Using plasmids of known DNA sequence that can be tailored
to include the proper restriction cleavage sites and allow for
DNA insertion, the same enzyme can then later be used to
cleave out the insert. The restriction sites for insertion are
typically chosen within plasmid genes that code for some
discernible functional property (e.g., antibiotic resistance).
Thus, interruption of coding and expression (when insertion
has been successful) leads to loss of that functional property
and results in a means to identify which plasmids have
successful inserts. Each plasmid can generally incorporate
only one cDNA insert, and, with a great excess of host
bacteria, each insert-bearing plasmid will infect only a single
host (almost always a specific type of Escherichia coli).

By growing these infected bacteria in such a way that each

individual bacterium gives rise to a colony of identical bacteria
bearing identical replicates of the plasmid and insert, the DNA
has been cloned. The cDNA can then be recovered from the
plasmid through another exposure to the restriction enzyme
selected for the original opening of the plasmid insertion site.
Thus, in a relatively few steps, one can start with a mixture of
mRNAs in widely differing proportions from common to very
rare and purify them individually, as well as prepare a
virtually unlimited number of pure samples of the DNA insert.

Polymerase Chain Reaction

In early 1988, the final touches were put onto a startling new
technology for preparing large amounts of specific and rare
nucleic acid sequences through amplification in vitro without
the necessity of first purifying the desired product through
cloning. The changes in molecular biological research wrought
by this new technology were immediate and dramatic, and
the chemistry was relatively quickly rewarded with a Nobel
Prize.

As originally conceived, the goal was to develop a quick

diagnostic tool for the genetic disorder sickle cell
hemoglobinopathy. Using previously selected restriction
enzymes, human DNA was sliced into small sections, one of
which contained the complete hemoglobin gene sequence.
When the DNA was heated to 95 C, the double-stranded DNA
"melted" and the strands separated. Two relatively short DNA
sequences were synthesized (oligonucleotides) to be
complementary to opposite strands of the hemoglobin gene
separated by the part of the gene structure that was known
to be altered in patients with the sickle cell mutation. These
oligonucleotide "primers" hybridized to their complementary
sequences on the single-stranded DNA. In the presence of
large amounts of a purified DNA polymerase and large
amounts of all the deoxynucleotides (A, G, T, and C), the
primers extended to the end of the single strand. In the
original experiments, this span covered only a distance of 20-
30 bases; subsequent modifications now permit amplification
of nucleotide sequences that are several hundred bases long.

By then separating through heat denaturation the dual helical

strands of the newly synthesized material, cooling the
reaction mixture, and adding fresh DNA polymerase and
nucleotide substrates, the cycle of denature, anneal, and
extend reactions, each lasting only a few seconds, can be
rapidly repeated. As long as the polymerase and the
nucleotide substrates remain in excess, the extended
sequences of the first reaction each serve as a template for
opposite strand synthesis in subsequent cycles, thereby
providing a geometric rate of amplification. By identifying a
DNA polymerase from a bacterial strain that survived in the
heat of a Yellowstone geyser (Thermus aquaticus, or Taq for
short), it was possible to develop a procedure that would
survive the heat-denaturation step. Thus, with the Taq
polymerase and large starting amounts of deoxynucleotides, a
series of rapid cycles allows for the exponential in vitro
amplification of the desired gene segment.

If the sequences selected are known to be generally constant

in the genome of the species to be studied, the same primers
can be used to select, amplify, and then evaluate the same
intervening gene segment from many individuals. If the
amplified segment is simply evaluated for its length, it is
possible to determine whether a given individual has had a
major structural gene mutation, such as a deletion or
insertion that will alter the way the restriction enzymes
fragment DNA. However, many other clever applications of
the technology have pushed its advantages even more
broadly than initially imagined. By first using reverse
transcriptase to make double-stranded cDNA, one can also
apply PCR to mRNAs and even make the assessment
quantitative (so-called q-PCR). By modifying the ends of the
probes to be used, it is possible to incorporate special
synthetic sequences that will later make it easier to clone or
sequence the amplified segments or to reinsert modified
versions to assess the functional importance of a specific
sequence domain. Using special synthetic nucleotides that will
form complementary base pairs even if the sequences do not
match precisely, it is possible to amplify homologous
sequences from different but related genes (e.g., as the
transmembrane domains for the seven-transmembrane
domain receptors like the noradrenergic, cholinergic, and
serotonergic receptors).

This highly superficial survey should indicate the facility of

cloning DNA segments (taken directly from genomic digests
or from mRNA copies) and then sequencing the cloned
segments and deducing the structure of their product.
However, we have glossed over the one potentially sour note
in this rhapsody of high-tech molecular music making: how do
you identify which clone carries the insert that encodes the
gene product you want to identify? Suppose you do not even
know what it is you want to look for and clone? As it turns
out, through more cleverness and a few good breaks from
Mother Nature, biologists have found multiple methods to
accomplish these feats of molecular magic, although some of
the screening methods have given new meaning to the phrase
"searching for a needle in a haystack."

MOLECULAR STRATEGIES IN
NEUROPHARMACOLOGY
The immediate applications of molecular biological strategies
within neuropharmacology are shortcuts in molecular isolation
and sequence determination-for instance, to uncover new
peptides or proteins or to provide more complete
understanding of enzymes, receptors, channels, or other
integral proteins of the cell. A rather likely premise holds that
the phenotype of a cell within the nervous system depends on
the structural, metabolic, and regulatory proteins by which it
establishes its recognizable structural and functional
properties. If this is valid, then complex, multifaceted neurons
will probably rely on hundreds, if not thousands, of special-
purpose proteins, many of which may exist in rather limited
amounts. Purifying such rare proteins by the methods that
existed before molecular cloning, especially in the absence of
a functional assay to guide the purification process, is an
overwhelming task, requiring exceptional patience, resources,
and a very large supply of the proper starting tissue material.
For some of the very rare hypothalamic hypophysiotropic
releasing factors (see Chapter 11), hundreds of thousands of
hypothalami were required, as well as the development of
unique purification schemes for each subsequent factor to be
pursued.

Converting the quest for the structure of specific proteins into

a molecular biological quest for the mRNA or gene segment
that encodes this protein greatly facilitates the experimental
analysis, simply because of the powers of cloning,
complementarity, and rapid sequencing. Several methods
have been developed that increase the chances of finding
whatever the researcher is seeking. These methods depend in
part on the nature of the cDNA being pursued and how the
investigator probes either for the insert or for the translation
of the fusion gene product in a cell system capable of
processing the primary translation product into a structural
form that will resemble its natural configuration and
sometimes even its natural function. Given the possibility of
protein-protein interactions, synthetic genes can be translated
into tools to detect unknown proteins with affinity for the
newly discovered gene product, a strategy exploited
shamelessly in the isolation of the vesicle proteins needed for
transmitter secretion.
In addition to their capacity to accelerate the discovery of new
molecules participating in the nervous system's response to
disease or to self-administered drugs, molecular biological
strategies can be used to determine how critical a particular
gene product may be in mediating a cellular event with
behavioral importance-for example, the role of a specific
transmitter's receptor subtype in a cellular event (e.g., long-
term potentiation), which may in turn underlie behavioral
phenomena such as memory formation or recall (see Chapter
12). These functional probes can be achieved through
advanced methods for transgenic animal construction (see
below) in which specific mutations are engineered into any
targeted gene of an experimental animal's genome, leading
eventually to the production of homozygous animals lacking
either that gene completely or the capacity to turn it on and
off as experimentally desired.

Less demanding and less vulnerable to potential confounding

roles of the targeted gene in the developing nervous system
are more short-term manipulations, such as the intracerebral
or intraventricular injection of special nucleotide constructs to
deliver ribozymes (RNA constructions that can target and
degrade specific mRNAs) or antisense oligonucleotides (which
can bind to mRNAs, delay their translation, and enhance their
degradation). A related strategy creates cells or whole
transgenic animals with novel mutations of the proteins in
question that closely resemble but can outcompete with the
protein at its natural binding partner sites (a so-called
dominant negative mutation). In addition to preparing novel
mutant animals by transgenic technologies, it is possible to
insert the special constructs into null cells (cells that normally
do not express the transduction system under investigation)
and, when they express a new receptor in their membrane, to
probe these receptors for more conventional pharmacological
characterization of natural or synthetic agonists and
antagonists. In this way, one might not only develop novel
drugs unique to such receptors but also detect previously
unknown natural ligands for functionally uncharacterized
"orphan" receptors (see Chapter 11).

GENERAL STRATEGIES FOR CLONE

SCREENING AND SELECTION

Introduction
We now briefly explore some of the more important steps of
these exciting discovery protocols.

Enrichment by Tissue Selection and

Preparation
One basic starting point is to select the brain cells or regions
that are presumed to express the molecule to be studied and
then to enrich sources of mRNA to favor the detection of the
one being pursued. Cell lines and even tumors that produce
large amounts of a hormone (e.g., pheochromocytomas or
Vasoactive Intestinal Polypeptide (VIP-omas) [see Chapter
11] or bear large numbers of the desired receptors or
channels (e.g., electroplax or striated muscle) have proved to
be excellent starting materials. Once the cell source is
selected, the desired mRNAs can be further enriched (e.g., by
sucrose gradient centrifugation or electrophoresis), provided
that some characteristics of the mRNA being sought are
known.

Recognizing the Wanted Clone

A general early strategy for detecting desired colonies of
cloned bacteria, known as colony hybridization, illustrates a
general flavor for the operations involved. Bacteria are grown
on special culture plates, from which their colonies can be
copied (transferred as a group by lightly pressing them to
another supporting surface, called "replica plating," thereby
sampling and preserving the spatial identity of all of the
colonies on the plate). The bacteria on the replicate supports
are treated to expose their DNA and screened with radioactive
nucleic acid probes. Colonies that hybridize with the probe
can then be identified by autoradiography. Alternatively, if the
plasmid-carrying inserts were tailored to allow for expression
of the protein encoded by the transferred genetic material, it
might also be possible to identify the desired clones by
immunological reactivity. When a reactive colony has been
identified, its original bacterial colony is recovered from the
original culture plate and the living bacteria are then grown in
large amounts to provide the starting material for DNA
sequence analysis. Modern methods now make these steps
seem primitive.

Building Your Own DNA

If partial protein (or peptide) sequences are known, it is
possible to make predictions of what the mRNA sequence
should be (by back-translating the genetic code for amino
acids) and including enough alternatives to overcome the
ambiguous cases where a specific amino acid may be encoded
by several variant triplets. From this predicted RNA structure,
it is possible to design and synthesize the hypothetical cDNA.
This approach has been used, for example, to create
hypothetical cDNAs for the hypophysiotropic hormones whose
amino acid sequences had been accurately determined "the
old-fashioned way," by earning it one amino acid at a time
from highly purified brain extracts. The main reason to do this
with a biologically active protein or peptide whose structure is
already known would be to get at the complete structure of
its prohormone or to obtain its complete genomic structure
and analyze its regulatory control and expression
mechanisms. As will be seen in Chapter 11, when this is
done, more often than not, the prohormone of the known
peptide is found to encode more than one active product.
However, because of the redundancy of triplet RNA codons for
some amino acids (there are six different codes for leucine
alone), it is generally difficult to acquire a functional full-
length mRNA by predictive synthesis.

An alternative approach is to synthesize a shorter

complementary single-stranded DNA (a so-called
oligodeoxynucleotide probe) and use it as a probe to screen
libraries of clones. Such libraries may be prepared from mRNA
extracts or from special digests of the whole genome. In the
former case, the starting material would be brain, while in the
latter case, in theory, any somatic cells could be used to
prepare the library. With the availability of automated "gene
machines," it is now possible to synthesize a proper probe or
two overnight and use them to screen an awaiting genomic or
cDNA library, thereby determining the complete coding
sequences for a partially purified protein within a few weeks.

On such a screening expedition, likely candidate colonies can

be cross-screened by a second synthetic oligodeoxynucleotide
probe, based on another separate domain of the full protein.
Clones positive for both probes would then have to contain
the gene sequence that encodes the two sequences against
which the probes were made as well as the sequence between
them. This strategy has been used with many neuropeptide
mRNAs (see Chapter 11).

When You Haven't Got a Clue

It is also possible to penetrate the large treasure trove of
cellular proteins that have not yet been identified by
conventional strategies. Given the length of the mammalian
genome and the relatively short list of identified specific
molecules in cells of all classes, we must conclude that there
is an awful lot left to be identified and few clues as to what it
is we do not know. The methods of molecular biology can help
here too.

However, there appears to be a rather select group of ancient

conserved sequences, numbering probably less than 1000,
that nature has found to be important enough to keep
relatively unmodified throughout hundreds of millions of years
of evolution. This core must then be significantly enriched by
the other 99% of the genome. Interestingly for
neuropharmacologists, most neurotransmitter receptor
molecules are not among these ancient conserved sequences,
although at least one 5-hydroxytryptamine (5-HT) receptor
and some of the transporters are. Furthermore, with the
complete sequencing of the genomes of Caenorhabditis
elegans and Drosophila, the paths of evolution, conservation,
and modification of ion channels and receptors are becoming
traceable. The inferences drawn from the first partial
inventories of gene structure and protein motifs perpetuate
the conservation of critical molecular mechanisms of life.

In nonneural tissues, it has been possible to identify the

unique proteins expressed in a male versus a female liver or
those that are unique to thymus-derived versus bone
marrow-derived lymphocytes, by exploiting the fact that a
very high proportion of the proteins expressed in these pairs
of tissues are, for the most part, very similar. Depending on
how the pairs of tissues or cell types are defined and how
similar or dissimilar they actually turn out to be, this method
can be made highly sensitive and can reveal unique
differences in cell-specific gene expression.

The discovery strategy can also be broadened to look for large

sets of tissue-specific genes. For example, Sutcliffe, Milner,
and Bloom employed molecular cloning methods to determine
what proteins are generally made by the brain that are not
found in other major tissues and the degree to which neurons
differ in specific proteins underlying their phenotypes that at
present are defined only empirically. In their original
approach, a cDNA library was prepared from mRNAs extracted
from whole rat brain and the individual cDNAs were
characterized for their ability to hybridize to mRNAs extracted
from brain, liver, or kidney. Those expressed in brain, but not
detectable in extracts of liver or kidney, represented well over
half of the total brain mRNA population (now estimated at
40,000 out of an anticipated total of 80,000-100,000,
although some estimates emerging from the human genome
have been interpreted as being perhaps half this many).
Whatever the final actual inventory of brain-specific genes
turns out to be, perhaps half or more of the genome contains
information pertinent to the generation of neuronal function.
Individual brain-specific clones are then analyzed further by
determining their nucleotide sequence and deducing the
amino acid sequence of its encoded protein. Proteins that are
unique to the database of known sequences can then be
identified further by raising antisera against synthetic
peptides that mimic selected regions of the deduced protein
structure. Subsequent variants of this approach sought genes
expressed in one brain region and not another and even in
one region of the cerebral cortex and not another. One such
hypothalamus-specific mRNA cloned in this manner,
hypocretin, encodes a novel neuropeptide now known to be
involved in the regulation of appetite, blood pressure, and
sleep and to be associated with human narcolepsy.

Yet another strategic approach involves injecting mRNAs from

enriched or prepared sources into frog oocytes or mammalian
cell lines whose genetic composition has been characterized.
By injecting groups of mRNAs and evaluating the oocytes or
cell lines for the response one seeks, it is possible through
trial and error to identify the mRNA for a specific functional
protein, like an ion channel or a cell surface receptor.

An All-Points Search
Would-be molecular neuropharmacologists attending a
seminar involving these approaches may find initial exposure
to the jargon confusing without some orientation.
"Orientation" is an appropriate term, for such presentations
are sprinkled with what may sound like references to compass
points. When DNA is fractionated by restriction enzymes and
the resulting fragments are separated by gel electrophoresis,
it is possible to transfer, or blot, the resulting fragments,
separated mainly on the basis of length, from the acrylamide
gel to a nitrocellulose or nylon support and, there, to analyze
them for the ability to hybridize with cDNA or RNA probes.
This method is referred to as a Southern blot analysis, named
for Dr. E. M. Southern, who started the evolution of this
method. Later, a similar approach was devised in which the
starting material was RNA. Here, the separated RNAs were
blotted for probing with radioactive, single-stranded cDNA
probes. Because the starting material is, speaking in terms of
nucleic acid, the opposite of that in a Southern blot, this RNA
blot is referred to as a Northern blot.

More recently, immunological methods have been used to

probe protein extracts that were separated by acrylamide gel
electrophoresis and then blotted (by an electrical transfer) for
identification by peroxidase or radioisotope-labeled antibodies
to specific protein antigens. The result is termed a Western
blot. If RNA or protein samples are simply dried directly onto
nitrocellulose for probing analysis without first separating
them for size, the resulting blots are termed slot blots or dot
blots. These blots are useful when screening a large number
of clone extracts for inserts or expressed products quickly. As
of this writing, no method has yet been termed an "Eastern
blot."

The resourcefulness with which innovative and automatable

molecular biological methods are being applied to all aspects
of biology grows more and more amazing with each issue of
your favorite journal. For example, the creation of gene chip
arrays allows for the comparison among relatively
homogeneous cell types of the order and magnitude of
changes in gene expression following an experimental
perturbation. Gene probes, applied to essentially microscope-
scale slides, can detect changes in gene expression by array
readers, and the ordering of their expression patterns and the
clusters of genes following similar or opposing patterns can be
extracted from the data by powerful bioinformatic computer
analysis software. As the human and mouse genome
inventories begin to narrow the remaining gaps, it is possible
to imagine a more or less complete list of the genes we have
available. One senses that we are about to experience a
logarithmic increase in the number of specific molecules that
will be fully characterized and from which pharmacological
engineers will shape drug molecules precisely to fit the pocket
of receptors or enzymes for ultimate specificity.

BEYOND THE CLONES

Although we may rightly marvel at the advances that have
been achieved through the use of molecular biological
techniques, all that we have in essence discussed so far in
this chapter is a set of methods that provide novel, powerful,
and accurate ways to identify, isolate, and characterize the
amino acid sequences of a host of intracellular proteins and
their possible subsequent metabolic products and interaction
partners. While this is unquestionably a major advance in the
research armamentarium, it still does not really begin to deal
with a wide range of other important questions that are also
approachable through the molecular tools that recombinant
methodologies provide.

For example, once a cDNA has been proven to represent the

mRNA for a specific molecule, the deduced protein sequence
can be inspected to infer potential functional properties. Thus,
the acetylcholine receptor molecule and the myelin proteolipid
protein exhibit several stretches of 20-24 hydrophobic amino
acids in a row, which is strong presumptive evidence of
membrane-crossing domains and, thus, suggestive of plasma
membrane constitutive proteins. In any case, when the
protein structure can be deduced, the entire molecule or
selected fragments of it can be synthesized and used to raise
antisera. These antisera can then be used to develop
radioimmunoassays for the protein or used in Western blots.
The antisera can also be used for immunocytochemical
analysis of the nervous system, to determine which cell and
which compartments of those cells exhibit the protein that
has been identified. Synthetic fragments can be used to
determine whether the protein domains may be substrates for
posttranslational modification, being processed by further
proteolytic cleavage or by structural modification with
glycosylation, phosphorylation, sulfation, or acylation.
Subcellular fractions and ultrastructural cytochemistry may
suggest organelle specialization or cell surface associations.

The products of cDNA cloning can also be taken back to the

genome, to probe the regions around the location of the
exons to search for the molecular mechanisms that control
transcription. Once the surrounding elements in the genome
have been located, the cDNA probes can be used to
determine the degree to which the mRNA or the underlying
gene exons have been conserved across eukaryotic species
and to determine the position on the chromosomes to which
the gene can be mapped. The chromosomal location of many
proteins has been determined in this or a similar manner.
However, the human genome, and that of most mammals, is
estimated to be on the order of 3 109 bp long. Until the race
to sequence the human genome greatly accelerated the pace
of sequence collection and computer-directed gene
identification, fewer than 20,000 genes had been mapped to
specific chromosomes. In an intermediate step of genome
analysis, markers of unknown genetic functions were mapped
at relatively low density across all human chromosomes. From
those low-resolution maps, the powerful, increasingly
automated genomic analysis routines have already yielded
much more detailed mapping. While a large proportion of the
genes noted so far can be related to previously discovered
molecular motifs, at least 25% of the genes have no such
antecedents. For students of genetic diseases of the nervous
system, the situation is even more complex. Given the
surprising number of mutations associated with known, but
rare, inheritable disorders, from color blindness to deafness to
epilepsy to mental retardation, it seems clear that there are
many ways to impair brain function. Nevertheless, the genetic
roots of complex illnesses such as Alzheimer's disease,
depression, and schizophrenia are beginning to give way to
these powerful methods (see Chapter 13).

The ability to link fragments of DNA with inheritance of

genetic disorders and to specific markers within the digested
fragments helped to establish approximate locations of
genetic mutations, such as the localization of the gene for
Huntington's disease to human chromosome 4. Given the
increased detailed information on length and complexity of
neuronal gene expression, efforts to link specific patterns of
DNA polymorphism (the combinations of unique alleles of
specific genes) may be a critical future development.
Southern blot analysis can track which family members have
inherited specific allele combinations by refinements of
restriction fragment length polymorphism designed to detect
single-nucleotide polymorphisms.

New modifications of these basic strategies are reported

continuously. Although there is a steady growth of molecular
information on the elements of neuronal function, most of the
details that follow in this book deal with those still relatively
few, but prominent, molecules (e.g., the major known
neurotransmitters, their synthetic and catabolic enzymes,
their receptors and response mediators), whose nature is
already partly established. While this list may well be
lengthened ever more rapidly by the shortcuts made possible
by molecular biological methods, students should recognize
that identifying new molecules per se is merely a first step
toward an important pharmacological end but is nowhere near
the true strength of what molecular biology may have to offer
our field.

Even more promising in its implications for

neuropharmacology is the microinjection of a segment of
cloned DNA into the pronucleus of a single-cell zygote (a
fertilized ovum) to create a transgenic mouse. After injecting
a large number of fertilized ova (a success rate of 10% is
considered good), one transfers the eggs to the uterine cavity
of a foster mother that has been pseudobred with a sterile
male. When the pups are born, they are evaluated through
skin fibroblast cultures for incorporation of the injected DNA
and, if possible, for expression of the gene product (e.g.,
overproduction of a circulating hormone). After puberty, their
sperm can also be evaluated for integration of the foreign
DNA sample into germ cells; these are best evaluated by their
ability to transmit the integrated gene to the mouse's own
progeny. If the integration of the foreign DNA is successful
and if the resultant gene structure (at the site of integration)
does not itself produce mutational consequences that may be
troublesome, the founder mice give rise to lines of mice
bearing the transgene. By preparing gene constructs that
induce overproduction of growth hormone, it has been
possible to use the technology to make not only supermice
but superpigs and supergoats as well.

Transgenic mice are now being used for a variety of

experimental purposes. The reverse application of transgenic
technology was actually done first, namely, preparing a
transgene for expression of myelin basic protein and using
that to "cure" the gene mutation of the myelin-deficient
mouse, which is ordinarily unable to make the protein. Other
partially tested strategies include removing a natural gene
(receptor or neurotransmitter transporter) to determine the
effects on cellular or behavioral function-for example,
receptivity to psychostimulants.

Among the more experimental applications of transgenic

technology is the preparation of a gene construct in which the
regulatory domains of a known gene (e.g., neuropeptide) are
coupled with the expression of a novel reporter gene (e.g., an
enzyme that is normally absent in the mouse); under these
conditions, one can evaluate whole-animal treatments (or
whole-cell treatments if cell lines are transfected instead of
embryonic cells) that "turn on the reporter gene," thus
providing inferential evidence of what controls (i.e., elevated
cyclic adenosine monophosphate) the activation of the natural
gene. Finally, applications still somewhat precarious to be
reproducible include the use of a retrovirus to carry a gene
segment to be transfected into cells beyond the single-cell
embryo level; this is especially useful in systems where
partial development can reveal the effects of the added gene.
A third creative but still early application involves transfection
of embryonic stem cells (undifferentiated blast-like cells taken
from the blastocyst level of embryonic development). When
these transfected stem cells are returned to blastocyst
embryos, they spread throughout the developing organism;
and if the researcher is fortunate, the foreign gene will appear
in the germ cell lines as well. Should gene markers ever be
identified that predict the development of single gene-
dependent neuropsychiatric diseases (such as Huntington's,
familial Alzheimer's, or scrapie), it should be possible to
prepare true animal models of the diseases and their
treatment.

MOLECULAR MOTIFS OF TRANSMITTER

RELEASE AND RESPONSE

Introduction
We conclude this foray into molecular neuropharmacology by
quickly considering the consistent structural patterns, or
"molecular motifs," identified in selected categories of
molecules that are essential for synaptic transmission and
thus critical for the content of subsequent chapters. This
survey will also illuminate some of the clever ways that
molecular biology has moved beyond mere discovery to
illumination of the molecular mechanisms of synaptic
transmission.

Sodium Channels

The three basic channels (Na+, K+, and Ca2+) underlying

neuronal excitability were cloned by starting with known short
segments of the proteins obtained from highly purified Na+
channel preparations. Oligonucleotides encoding these
peptide segments and antibodies raised against them were
used to probe cDNA libraries of electric organ mRNA and led
to the sequence determination for what turned out to be the
subunit of the Na channel (there is also a simpler subunit).
Modeled from the distribution of hydrophobicity plots for the
deduced amino acid sequences, the hypothetical -subunit
structure contained four repeats of highly similar six-
transmembrane -helical domains, in which the fifth and sixth
transmembrane domains are separated by extracellular loops,
while the N-terminal, the C-terminal, and the cytoplasmic
loops linking the sixth transmembrane domain to the first
transmembrane domain of the adjacent repeat are
intracellular cytoplasmic loops.

These sequence data were then used to derive the rat brain
Na+ channels, which also turned out to have a similar
structure, at least in the transmembrane domains, though not
the connecting loops. These modeled conformations have
been refined by a variety of functional analyses, including
expression in oocytes to confirm their functional properties
(including sensitivity to channel-blocking toxins) and to
determine pore- and voltage-sensitive domains by mutation,
deletion, and transposition of sequences.

Calcium Channels
Subsequent work, again starting with highly purified protein
sources and cloning from the first few short segments of
amino acid sequences wrung from them, revealed the
molecular characteristics of the Ca channel proteins, which
had previously been categorized on their functional
differences only, as voltage-sensitive Ca2+ channels first
cloned from muscle. As more and more voltage-sensitive
Ca2+ channel subunits and variants were recognized from
muscle and neuronal sources, the complexity of the
heterooligomeric Ca2+ channel proteins has been surprising.
Present data indicate a tetrameric 1 subunit with significant
sequence similarity to the Na channel, with six presumptive
transmembrane domains, a transmembrane 2 subunit
linked by disulfides that is required for complete function, as
is a second adjacent transmembrane subunit and a
peripheral intracytoplasmic subunit. With these basic
subunit characterizations in place, the challenges have been
to define the basis for the functional categorizations of the
previous L, N, and T types of Ca2+ channel recognized
physiologically and their variants. Significant progress in this
challenge has been obtained (see Randall and Benham,
1999).

Potassium Channels

Definition of the structures of K+ channels began with the

cloning of the Shaker gene of Drosophila, which then provided
the structural tools to characterize the multiple forms of
voltage-sensitive and then ligand-gated channel proteins
carrying K+ currents in mammalian neurons. The
crystallization of a bacterial K+ channel has given three-
dimensional reality to the differences in structure and
regulatory function. The basic K+ channel is a presumptive
six-transmembrane structure with a single ionic pore, almost
identical to the conformation of one of the repeat segments in
the Na+ and Ca2+ subunits. In voltage-dependent K+
channels, there is a cytoplasmic peptide between the fifth and
sixth presumptive transmembrane domains, known as "H5"
senses the intracellular polarization. Similarly, the cytoplasmic
loop between transmembrane domains 4 and 5 provides the
means to inactivate the channel from the inside of the cell.
The ligand-gated form of the K+ channel, termed the inward
rectifier, shows an even simpler structure, with two
transmembrane domains and an intervening H5-like voltage-
sensing sequence. Two other variants may account for the
outward rectifier K+ channels with dual ionic pores and either
eight or four transmembrane domains.

Neurotransmitter Receptors
Here, the starting point was protein sequences from highly
purified preparations of electroplax or muscle receptors (for
acetylcholine), brain (for GABA and glycine), or blood cells
(for norepinephrine). Once those structures were in hand,
however, multiple additional members of their receptor
superfamilies were snared by probing cDNA libraries with RNA
probes designed to accept loose (or low-stringency) matches
with similar, but not identical, nucleotide sequences.

As noted in greater detail in Chapter 4, neurotransmitter

receptor cloning studies have so far split the spoils into two
large groups: the ionophore receptors, in which the receptor
is a multimeric ion channel composed of four or five subunits,
each exhibiting a similar presumptive four-transmembrane
domain structure, and the metabotropic, or G protein-
coupled, receptors, each of which is a monomer with seven
conceptual transmembrane domains, although such receptors
may also dimerize. Within the ionophore receptors, the
protein subunits for a given neurotransmitter (acetylcholine
nicotinic receptors, GABA and glycine receptors, most of the
glutamate receptors, and one of the 5-HT receptors), there is
enough similarity to allow for recognition of further receptor
subtyping of different assembled multimers structurally.
Different combinations of GABA subunits expressed in
neurons with a high degree of heterogeneity, for example, can
confer cell type-specific sensitivity to ethanol or
benzodiazepines (see Chapter 7). A surprising feature of one
category of glutamate receptor, the AMPA ( -amino-3-
hydroxy-5-methyl-4-isoxazole propionic acid) receptor (see
Chapter 6), is its dynamic insertion and removal from the
subsynaptic sites under conditions of activation.

The second large category of neurotransmitter receptors

comprises the guanine nucleotide-binding protein-coupled
receptors (for cholinergic muscarinic receptors; all
catecholaminergic, histaminergic, and neuropeptide
receptors; all other 5-HT receptors; and the metabotropic
glutamate receptor). All of these exhibit a presumptive seven-
transmembrane domain configuration, first recognized in
rhodopsin and then found in the first of the cloned adrenergic
receptors. Through expression of mutated mRNA constructs in
null cells, it has been possible to determine quite definitively
which segments of which transmembrane domains are
responsible for ligand recognition and which segments of
which cytoplasmic loops are responsible for interactions with
the G proteins. There is good evidence that at least two G
protein-coupled receptors, the -adrenergic receptor and the
substance P receptor, are internalized after activation by
ligands, temporarily downregulating sensitivity.
Neurotransmitter Transporters
We now turn our attention to a family of molecules that are
essential to the process of transmitter conservation-after-
release. This process, which we used to call simply
"reuptake," has been upgraded with the term transporter,
following cloning. In a relatively brief burst of discoveries in
the early 1990s, the transporters for GABA, glycine, all of the
monoamines, as well as proline and betaine were cloned and
found to express a very similar structural motif consisting of
12 transmembrane domains with substantial conservation
across the transmembrane domain sequences, allowing for
the information from the initial cloning of the GABA
transporter to be extended to the others. In all of these
molecules, concentration of the specific small molecule from
the extracellular space back into the interior of the neuron,
generally the presynaptic terminals, seems to be driven by
cotransport of Na+ and Cl- ions, and the molecules bear a
strong resemblance to the previously known glucose
transporters and to adenylyl cyclase for as yet unclear
reasons.

A second variety of transporter has also been defined that

concentrates cytoplasmic neurotransmitter into the synaptic
vesicles, driven by either a pH or an ionic gradient. While
these transporters also exhibit apparent 12-transmembrane
domain structures, the actual sequences distinguish them
from those that operate in the presynaptic plasma membrane.

Synaptic Vesicle Proteins

In an era of neuropharmacology not too long ago, a minor
parlor game was made of the ability to classify synaptic
vesicles as to their transmitter content on the basis of size,
shape, and cytochemistry. The advent of molecular biology
has radically transformed our views of vesicles from mere
morphological description to arguably the most completely
detailed nervous system organelle in terms of a specifiable
inventory of proteins and interactions during transmitter
secretion. Again, the initial studies started with highly purified
brain synaptic vesicles and a search for the proteins they
contain and, in a marvelous display of Mother Nature's
generosity, managed to reveal the molecular machinery for
secretion across wide realms of the phylogenetic tree. Given
the estimates for the molecular mass of the typical 500 A and
the roughly 600 kDa size of the vesicle transporter, there
could be room for about a dozen more molecules of
approximately 50 kDa. The inventory to date reveals the
following three groups:

1. Vesicle-bound proteins: synapsins (which are substrates for

cAMP-dependent protein kinase and can regulate transmitter
release); synaptotagmin (possibly a Ca-sensor protein) and
synaptobrevin (also known as VAMP for vesicle-associated
membrane protein and a substrate for proteolysis by tetanus
toxin and by some of the botulinum toxins), both of which
seem to be essential for docking on the presynaptic plasma
membrane; and synaptophysin, a vesicle protein of as yet
unknown function (but antibodies to it have served as a useful
means to quantify synapses). Each of these vesicle proteins
has several isoforms such that, it is believed, one of each
class will be present on every synaptic vesicle.

2. Synaptic membrane proteins: the synaptic soup thickens

with two other proteins, SNAP-25, a synapse-associated
protein of 25 kDa, and syntaxin, both of which are substrates
for other botulinum toxin proteases and, thus, also essential
players in the docking and/or fusion steps to release
transmitter.

3. A series of soluble proteins called NSF (an N-

ethylmaleimide-sensitive factor that alters the capacity to
transport vesicles across the cisternae of the Golgi complex in
a clever in vitro assay) and its three soluble NSF-associated
proteins (called -, -, and -SNAPs, of which the form is
quite enriched in brain, while the others are more pronounced
in nonneuronal secretory cells and bear a strong resemblance
to secretory proteins recognized in yeasts for their role in
constitutive secretion).

In the case of the vesicle-related secretion proteins, cloning

provided the detailed sequences and the related isoforms, the
capacity to make small mutations and deletions to define the
docking and interaction domains, and the ability to
reassemble the entire array in null cell assays and to establish
the potential roles for each protein. There are a few more
paralytic bacterial toxins whose substrates in secretion are
still to be accounted for, so we are not likely to be at the end
of this very interesting search. Furthermore, we still do not
know exactly how the vesicles are reformed within the
synaptic terminal following the docking, fusion, and release
steps, presumably involving guanine exchange factor, ADP-
ribosylation factor, and vesicle coat proteins similar to those
budding from the Golgi membranes. A minor mystery still
surrounds the nature of the signal for when it is time to ship
the whole vesicle back up to the perikaryon for an overhaul.

SELECTED REFERENCES
Bargmann, C. A. (1998). Neurobiology of the Caenorhabditis
elegans genome. Science 282, 2028-2033.

Barondes, S. H. (1993). Molecules and Mental Illness. W. H.

Freeman, New York.

Brady, S., D. R. Coleman, and P. Brophy (1998). Subcellular

organization of the nervous system. organelles and their
functions. In Fundamental Neuroscience (M. Zigmond, F. E.
Bloom, S. C. Landis, J. Roberts, and L. Squire, eds.).
Academic Press, New York, pp. 71-106.
Brown, D. A. (2000). The acid test for resting potassium
channels. Curr. Biol. 10, R456-R459.

Crick, F. H. C., L. Barnett, S. Brenner, and R. J. Watts-Tobin

(1961). General nature of the genetic code for proteins.
Nature 192, 1227-1232.

Gilbert, W. (1985). Genes-in-pieces revisited. Science 228,

823.

Gusella, J. F., R. E. Tanzi, M. A. Anderson, W. Hobbs, K.

Gibbons, R. Raschtchian, T. C. Gilliam, M. R. Wallace, N. S.
Wexler, and P. M. Conneally (1984). DNA markers for nervous
system diseases. Science 225, 1320-1325.

Hille, B., C. M. Armstrong, and R. MacKinnon (1999). Ion

channels: from idea to reality. Nat. Med. 5, 1105-1109.

Le Novere, N. and J. P. Changeux (2001). The ligand gated

ion channel database: an example of a sequence database in
neuroscience. Philos. Trans. R. Soc. Lond. B Biol. Sci. 356,
1121-1130.

Lobe, C. G. and A. Nagy (1998). Conditional genome

alteration in mice. Bioessays 20, 200-208.

Randall, A. and C. D. Benham (1999). Recent advances in the

molecular understanding of votage-gated Ca2+ channels. Mol.
Cell. Neurosci. 14, 256-272.

Sanger, F. and A. R. Coulson (1975). A rapid method for

determining sequences in DNA by primed synthesis with DNA
polymerase. J. Mol. Biol. 94, 444-448.

Sutcliffe, J. G. (2001). Open-system approaches to gene

expression in the CNS. J. Neurosci. 21, 8306-8309.
Tjian, R. (1995). Molecular machines that control genes. Sci.
Am. 54-61.

Watson, J. D. and F. H. C. Crick (1953). Molecular structure of

nucleic acids: a structure for deoxyribose nucleic acid. Nature
171, 737-738.

(+/-) Show / Hide Bibliography

Send Feedback Teton Server (5.0.0) - ©2007 Teton Data Title Updates
Customer Service Systems User
800.901.5494 Send Us Your Comments Responsibilities
Training Center
What's New
4. Receptors
INTRODUCTION
The concept that most drugs, hormones, and
neurotransmitters produce their biological effects by
interacting with receptor substances in cells was introduced
by Langley in 1905. It was based on his observations of the
extraordinary potency and specificity with which some drugs
mimicked a biological response (agonists) while others
prevented it (antagonists). Later, Hill, Gaddum, and Clark
independently described the quantitative characteristics of
competitive antagonism between agonists and antagonists in
combining with specific receptors in intact preparations. This
receptor concept has been substantiated in the past several
years by the actual isolation of macromolecular substances
that fit all of the criteria of being receptors. To date, although
not all have been cloned, receptors have been identified for all
of the proven neurotransmitters as well as for histamine,
opioid peptides, neurotensin, vasoactive intestinal peptide,
bradykinin, cholecystokinin, somatostatin, substance P,
insulin, angiotensin II, gonadotropin, glucagon, prolactin, and
thyroid-stimulating hormone. In addition, as noted in Table 4-
1, multiple receptors have been shown to exist for all of the
biogenic amines, acetylcholine (ACh), -aminobutyric acid
(GABA), histamine, opiates, and the amino acid transmitters.
If receptors for all agents (e.g., hormones, trophic factors,
odorants, peptides) in addition to the neurotransmitters were
counted, a total of 1000 would not be surprising. This number
would also include orphan receptors, which are nuclear
receptors that have been cloned because of their similarity to
known receptors but which have no known ligands. Some of
these orphan receptors, when examined, appear to be
transcription factors.

Multiple receptors appear to metastasize at an uncontrollable

rate, but this should be viewed skeptically until a physiological
response to the ligand has been shown or a specific gene has
been cloned and expressed. Some of the receptor subtypes
have been identified only by binding techniques (see below),
which can lead to erroneous conclusions. All of the receptors
for neurotransmitters and peptide hormones that have been
studied, regardless of whether they have been isolated, are
localized on the surface of the cell; only the receptors for
steroid and thyroid hormones are intracellular.

After the discovery that the action of physostigmine (eserine)

resulted from its anticholinesterase activity, it was assumed
that most drugs acted by inhibiting an enzyme. However, it
now appears that with few exceptions (an action on ion
channels or transport proteins) the mechanism of action of
neuroactive drugs usually stems from their effect on specific
receptors. Predictably, the current search for receptors is
among the most intense areas in the neurosciences. This
interest is not purely academic. The recent identification of
adrenergic, dopaminergic, muscarinic, serotonergic, and
histaminergic receptor subtypes has led to the synthesis of
highly selective drugs that are considerably more specific than
their prototypes, which were developed after general
screening for activity. With advances in gene cloning and
expression, more and more receptor subtypes are being
identified, each presumably having its own function. What this
indicates is that future drugs can be designed to fit a single
receptor subtype, thus precluding the side effects of a
nonspecific drug.

DEFINITION
In this rapidly developing field, considerable confusion has
arisen as to what functional characteristics are required of an
isolated, ligand-binding molecule to qualify as a receptor. This
confusion, a semantic problem, developed after the successful
isolation of macromolecules that exhibit selective binding
properties, which made it mandatory to determine whether
this material comprised both the binding element and the
element that initiated a biological response or merely the
former. Some investigators use the term receptor only when
both binding and signal generation occur; they use the term
acceptor if no biological response has been demonstrated.
Others are content to ignore the bifunctional aspect and use
receptor without specifications. In this chapter, we will define
a receptor as the binding or recognition component and refer
to the element involved in the biological response as the
effector, without specifying whether the receptor and effector
reside in the same or separate units. The criteria for receptors
will be dealt with shortly; the biological response that is
generated by the effector obviously has a wider range of
complexity, from a simple one-step coupling to an unknown
number of steps (Fig. 4-1).

Figure 4-1. Schematic model of ligand-receptor

interaction.

ASSAYS
Basically, there are two ways to study the interaction of
neurotransmitters, hormones, or drugs with cells. The first
procedure (and until relatively recently the only one) is to
determine the biological response of an intact isolated organ,
such as the guinea pig ileum, to applied agonists or
antagonists. The disadvantage of this procedure is that one is
obviously enmeshed in a cascade of events beginning with
transport, distribution, and metabolism of the agent before it
even interacts with a receptor, ranging through an unknown
multiplicity of steps before the final biological response of the
tissue is measured. Thus, although studies with agonists may
be interpretable, it is not difficult to envision problems when
antagonists are employed since these compounds may
compete at a level different from receptor binding. Despite
the not unusual problem of a nonlinear relationship between
receptor occupancy and biological response, this approach
has yielded a considerable amount of information. The second
approach to studying receptors is by measuring ligand binding
to a homogenate or slice preparation. This technique became
feasible with the development of ligands of a high specific
radioactivity and a high affinity for the receptor. Here, the
direct method is to incubate labeled agonist or antagonist
with the receptor preparation and then separate the receptor-
ligand complex from free ligand by centrifugation, filtration,
or precipitation. The indirect technique is to use equilibrium
dialysis, where the receptor-ligand complex is determined by
subtracting the ligand concentration in the bath from that in
the dialysis sac.

Recently, a third, electrophysiological approach to identifying

receptors has emerged. Intracellular stimulation and
recording via microelectrodes inserted into a brain slice or
neurons in culture, combined with the application of receptor
agonists and antagonists, can functionally identify receptor
subtypes.

Though not always appreciated, the advantage of using

isolated tissues is that both the efficacy and the functional
activity of an agonist are assessed, in contrast to binding
procedures using broken cell preparations, where only affinity
or a biochemical sequela of binding can be appraised. Ideally,
both techniques should be used, but, alas, this is rarely done.
It should be noted in passing that efficacy and affinity are
independent. To date, it appears that the drugs that exhibit
high affinity but low efficacy have more efficient coupling to
the effector than the reverse situation; therefore, these are
the most potent and selective agents. For a detailed
discussion of the above, the review by Kenakin et al. (1992)
is recommended.
IDENTIFICATION
In the midst of an intensive drive to isolate and characterize
receptors, some zealous investigators have lost sight of the
basic tenets that must be satisfied before it is certain that a
receptor has indeed been isolated. Thus, on occasion,
enzymes, transport proteins, and merely extraneous
lipoproteins or proteolipids that exhibit binding properties
have been mistakenly identified as receptors. Authentic
receptors should have the following properties:

1. Saturability. The great majority of receptors are on the

surface of a cell. Since there are a finite number of receptors
per cell, it follows that a dose-response curve for the binding
of a ligand should reveal saturability. In general, specific
receptor binding is characterized by high affinity and low
capacity, whereas nonspecific binding usually exhibits high
capacity and low-affinity binding that is virtually nonsaturable.

2. Specificity. This is one of the most difficult and important

criteria to fulfill because of the enormous mass of nonspecific
binding sites compared with receptor sites in tissue. For this
reason, in binding assays it is necessary to explore the
displacement of the labeled ligand with a series of agonists
and antagonists that represent both the same and different
chemical structures and pharmacological properties as the
binding ligand. One should also be aware of the avidity with
which inert surfaces bind ligands. For example, substance P
binds tenaciously to glass, and insulin can bind to talcum
powder in the nanomolar range. With agents that exist as
optical isomers, it is of obvious importance to show that the
binding of the ligand is stereospecific. Even here problems
arise. With opiates it is the levorotatory enantiomorph that
exerts the dominant pharmacological effect. Snyder, for
example, has found glass fiber filters that selectively bind the
levorotatory isomer. Specificity obviously means that one
should find receptors only in cells known to respond to the
particular transmitter or hormone under examination.
Furthermore, a correlation should be evident between the
binding affinity of a series of ligands and the biological
response produced by this series. This correlation, the sine
qua non for receptor identification, is unfortunately a criterion
that is not often investigated.

3. Reversibility. Since transmitters, hormones, and most

drugs act in a reversible manner, it follows that their binding
to receptors should be reversible. Also, the ligand of a
reversible receptor should be not only dissociable but
recoverable in its natural (i.e., nonmetabolized) form. This
last dictum distinguishes receptor-agonist interactions but not
receptor-antagonist binding from enzyme-substrate reactions.

4. Restoration of function upon reconstitution. Following the

isolation and identification of the components of the receptor
system, to "put Humpty Dumpty back together again" is the
goal of all receptorologists.

5. Molecular neurobiology. The ultimate identification is to

isolate the gene for a receptor, express it, and demonstrate
the exact similarity of the cloned receptor to the natural one.

It is important to recognize that the quantitative and spatial

measurement of receptors utilizing autoradiography is also a
key problem. Where labeled ligands are employed to map
receptors in brain via light microscopy, a mismatch is often
encountered. Reasons offered for this problem are (1) except
for autoreceptors, neurotransmitters and receptors are
located in different neurons; (2) in addition to the synapse,
receptors and transmitters are found throughout the neuron
and in glial cells; (3) ligands may label only a subunit of a
receptor or only one state of the receptor; and (4)
autoradiography is subject to quenching. With
immunohistochemical peptide mapping, a possible problem is
the recognition by the antibody of a prohormone or,
alternatively, a fragment of a peptide hormone in addition to
the well-recognized problem of cross-reactivity of the
antibody with a physiologically different peptide.

Finally, all drugs do not necessarily act directly on a receptor.

They could bind to a site that is adjacent to a receptor and
thus influence the activity of the receptor.

KINETICS AND THEORIES OF DRUG ACTION

From the law of mass action, the binding of a ligand (L) to its
receptor (R) leads to the following equation:

thus

where k1=association rate constant

k2=dissociation rate constant

[L] = concentration of free ligand

[R] = concentration of free receptor

[LR] = concentration of occupied binding sites

Kd = dissociation equilibrium constant

Since the total number of receptors = [Rt] = [R] + [LR],

If experiments are performed in which the receptor
concentration is kept constant and the ligand concentration is
varied, then a plot of r versus [L] will produce a rectangular
hyperbole, the usual Langmuir adsorption isotherm. Here r
approaches the saturation value of 1. If r is plotted against
log [L], a sigmoid curve will result; log [L] at half-saturation
will give log Kd on the horizontal axis (Fig. 4-2).

This equation can be rearranged as follows:

Now if r/[L] is plotted against r, a straight line will result

(assuming only one set of binding sites) with two intercepts,
the one on the x axis giving the number of binding sites per
molecule and the y intercept yielding I/Kd. This type of plot is
the Scatchard plot (more correctly, Rosenthal plot), widely
used in studying receptor-ligand interactions (Fig. 4-3).
Among the pitfalls encountered in a Scatchard analysis is the
problem that the system is not in true equilibrium.

Another useful representation that can be derived from the

general equation is the Hill plot (Fig. 4-4). If log E/(Emax -E)
is plotted versus log [L] when E is the effect produced and
Emax is the maximum effect, then the slope, indicative of the
nature of the binding, gives a Hill coefficient of unity in cases
where E/Emax is proportional to the fraction of total number
of sites occupied (r). In many situations, the slope turns out
to be a noninteger number different from unity. This finding
indicates that cooperativity may be involved in the binding of
the ligand to the receptor. Cooperativity is the phenomenon
whereby the ligand binding at one site influences, either
positively or negatively, the binding of the ligand at sites on
other subunits of the oligomeric protein. This idea, originally
suggested in 1960 by Monod, Wyman, and Changeux to
explain allosteric enzyme properties, currently offers the most
attractive hypothesis for studying reactions of receptors with
hormones, transmitters, or drugs. We will utilize this
hypothesis later to explain drug action, including the problem
of efficacy (intrinsic activity), spare receptors, and
desensitization.

Another useful analysis when competitive inhibitors of the

receptor binding ligand are studied is the equation derived by
Cheng and Prusoff (1973) in their kinetic analysis of enzyme
inhibitors:

where Ki is the equilibrium dissociation constant of the

competitive inhibitor and I50 is the concentration of the
inhibitor producing 50% inhibition at the concentration of the
labeled ligand that is used in the study.

Clark (1937) produced the first model of drug-receptor action,

known as the "occupation theory," in which the response of a
drug was held to be directly proportional to the fraction of
receptors occupied by the drug. Here, as mentioned earlier,
one should find the usual Langmuir absorption isotherm.
However, instead of the expected rectangular hyperbole when
drug concentration was plotted against the drug bound, in
most cases a sigmoid curve resulted. A second problem with
the occupancy theory is that in many instances only a small
fraction of the total receptors available are occupied, and yet
a maximum response is obtained. These additional receptors,
which may represent as much as 95%-99% of the total, are
referred to as spare receptors. The occupancy theory is
further complicated when the activity of a series of related
agents is explored and the biological response varies from
maximum to 0, even though all of the agents occupy the
receptor. In other words, these agents could be full agonists,
partial agonists that give less than a maximal response, and
antagonists whose occupancy produces no response. This
phenomenon is referred to as efficacy or intrinsic activity of a
drug and is obviously not directly related to the binding
affinity of the drug. A fourth characteristic of drug-receptor
interaction that is sometimes observed is desensitization,
which is the lack or decline of a response to a constant
stimulus.

These problems the sigmoidal dose-response curve, some

anomalous effects with antagonists that spare receptors
might account for, efficacy, and desensitization can be
comfortably fitted into a two-state model of a receptor
analogous to the allosteric model of Monod, Wyman, and
Changeux, proposed independently by Changeux and by
Karlin in 1967.

According to the two-state model, receptors exist in an active

(R) and an inactive (T) state, and each is capable of
combining with the drug (A):
Here, an agonist prefers the R (active configuration of the
receptor), and the efficacy (i.e., intrinsic action) of the drug
will be determined by the ratio of its affinity for the two
states. In contrast, competitive antagonists prefer the T form
of the receptor and will shift the equilibrium to AT. The
sigmoid relationship between the fraction of receptors
activated and the drug concentration (i.e., cooperativity) can
be explained by this model with its equilibrium between R and
T, if one designates T as a subunit of R that binds A and
thereby influences the further binding of A to R. Cooperativity
can also explain anomalous effects of antagonists whenever
the effect of an antagonist persists even in the presence of a
high concentration of agonist. Here, it could be postulated
that by tightly binding to one conformational state of the
receptor, the antagonist inhibits the binding of the agonist.
One might even use this two-state model to account for
desensitization, where T would be a receptor that has been
desensitized perhaps by a local change in the ionic
environment. The ionic environment may be one of the
factors that dictate the two conformational states of a
receptor. Other possibilities include polymerization
(clustering) of the subunits, depolymerization, or
phosphorylation. It should be emphasized that this model is
conjectural, subject to modifications as the need arises.

When one considers the finite number of receptors per cell

and the fact that virtually all receptors are found on the
plasma membrane, it is not surprising that progress in this
area has been slow and laborious. It has been calculated that,
assuming one binding site and an average molecular weight
of 200,000 for a receptor, complete purification of a receptor
protein would require about a 25,000-fold enrichment. The
extraordinary density of ACh receptors in electric tissue has
made this preparation so popular a choice that considerable
information is now available on this cholinergic receptor. Two
snake neurotoxins, Naja siamensis and -bungarotoxin, which
specifically bind nicotinic cholinergic receptors, have been the
key agents that have helped in isolating this receptor.

Also to be considered is the relationship between receptors

and effectors as they interact in the fluid mosaic membrane.
As envisaged by Singer and Nicolson (1972), membranes are
composed of a fluid lipid bilayer that contains globular
protein. Some of these proteins extend through the
membrane, and others are partially embedded in or on the
surface. The hydrophilic ends of the protein protrude from the
membrane, while the hydrophobic ends are localized in the
lipid bilayer. Some of the proteins are immobilized, but
others, floating in an oily sea of lipids, are capable of free
movement. In this concept of membrane fluidity, the receptor
protein would be on the surface of the membrane and the
effector within the membrane. Although the ratio of receptors
to effectors may in some cases be unity (thus explaining
instances in which the occupancy theory is satisfied), it may
also be greater than 1. Consequently, in a situation where
multiple hormones activate a response (e.g., there is only one
form of adenylyl cyclase in a fat cell, but it may be activated
by epinephrine, glucagon, corticotropin, or histamine), it
would be concluded that an excess of receptors over effectors
is present. This model would also explain spare receptors and
is exemplified by the fact that only 3% of insulin receptors
need to be occupied in order to catalyze glucose oxidation in
adipocytes. With the possibility of easy lateral movement of
effector in the membrane, it is also understandable why one
receptor may activate several types of effector. Membrane
fluidity will account for the observation that hormones can
influence the state of aggregation of the receptor, thus giving
rise to either positive or negative cooperativity as determined
in kinetic studies of binding. Although direct evidence of the
interaction and migration of receptors and effectors is difficult
to obtain, current information is easily accommodated by the
fluid mosaic membrane model.

Currently, there are four major groups of receptors known to

be involved in signal transduction (Fig. 4-5), of which the first
two are neurotransmitter-activated. The first group is referred
to as ligand-gated channels or ionotropic channels. It is
composed of multiple subunits with a central pore, which,
when activated, open to permit the passage of Na+, K+,
Ca2+, or Cl-. Thus, depending on which ion is involved, the
membrane potential may be either depolarized or
hyperpolarized. Since no second-messenger biochemical
systems are involved, the effects of neurotransmitters on
these cell surface receptors are very fast, with excitatory and
inhibitory responses occurring in milliseconds. Examples of
these ionotropic receptors are the nicotinic ACh receptor, the
N-methyl-D-aspartate receptor, the GABAA receptor, and the
5-hydroxytryptamine3 (5-HT3) receptor.

The second neurotransmitter-activated receptors that are also

membrane-localized are the G protein-coupled receptors,
referred to as metabotropic receptors. These receptors
mediate slower responses (seconds to minutes) that are
generally modulatory, dampening or enhancing the signal. All
known G protein-coupled receptors contain seven hydrophobic
transmembrane domains that are linked by hydrophilic
groups. Examples of these receptors are muscarinic
cholinergic, adrenergic, dopaminergic, serotonergic,
metabotropic glutaminergic, opiate, peptidergic, and some
purinergic receptors (vida intra). The structure of these
receptors, as exemplified by the 2-adrenergic receptor, is
shown in Figure 4-6.

Metabotropic receptors are a family of guanine nucleotide-

binding proteins with a heterotrimeric structure consisting of
-, -, and -subunits. The G protein signal-transduction cycle
is shown in Figure 4-6. At last count, 22 different G protein
subunits and at least five and 12 subunits had been
identified. These proteins can be roughly classified into four
groups: Gs, Gi, Gq, and G12. Activation of the Gs subunit
family increases adenylyl cyclase activity, opens Ca2+
channels, and inhibits Na+ channels. The Gi subunit family
opens K+ channels, closes Ca2+ channels, inhibits adenylyl
cyclase, and promotes cyclic guanosine monophosphate
phosphodiesterase and probably phospholipase A2. Increased
phospholipase C is the effector for Gq, while with G12, which
activates Rho, a guanosine triphosphate (GTP)-binding
protein, there is yet no other known function (a dysfunctional
family?). Also recently recognized are a family of over two
dozen regulators of G protein signaling (RGS) that promote
desensitization by activating GTPase. Of course, these RGS
proteins, which may have other functions, are in turn
regulated. To add to this bewildering complexity in neuronal
signaling, in addition to the fact that a single receptor can
activate multiple G proteins that may or may not interact,
cloning studies have revealed multiple isoforms of adenylyl
cyclase, phospholipase C, phospholipase A2, and calcium and
potassium channels.

An observation that has helped assign G proteins a role in

signal transduction is that bacterial toxins catalyze the
nicotinamide-adenine dinucleotide (NAD)-dependent
adenosine disphosphate (ADP) ribosylation of the subunit of
many of the G proteins and inhibit their activity. Cholera toxin
ribosylates Gs and Gt, whereas Gi and Go are substrates for
pertussis toxin. Some G proteins (unclassified) are resistant
to both toxins.

The third group are receptors for steroid hormones, thyroid

hormones, vitamin D, and retinoic acids. These lipophilic
ligands penetrate the cell, where they bind to the nuclear DNA
to stimulate transcription. Response to hormones usually
occurs in hours. Receptors for steroid hormones have
previously been thought to be only intracellular since the
hormones act on nuclear DNA to alter gene expression. This
action is referred to as a genomic effect and usually takes
hours to days to be observed. These genomic actions include
the induction of neurotransmitter enzymes, receptors, and
dendritic spines. Recently, however, evidence has
accumulated that steroid receptors are also found on
membranes. The hormone's nongenomic effects, occurring in
seconds to minutes, act on the GABAA receptor to modulate
chloride flux (see Chapter 6) and modulate a variety of other
receptors (see Fig. 4-7). The major players in this activity are
progesterone and its metabolites, estrogens, testosterone,
and adrenal steroids. The neuroactive steroids, some of which
can be synthesized in the brain, may be sedative, anxiolytic,
antidepressant, or anticonvulsant. Recently, testosterone has
been shown to reduce the secretion of the -amyloid peptides
that characterize Alzheimer's disease.

The fourth group are the tyrosine kinase receptors, now

numbering around 100, differing from other receptors in that
the kinase activity is part of the receptor. Membrane-localized,
these receptors are activated by insulin and a number of
growth factors, including nerve growth factor. In the presence
of a ligand, the receptor dimerizes, activating the kinase to
autophosphorylate. The subsequent phosphotyrosine residues
provide acceptor sites for a variety of proteins.

Purinoreceptors have been classified as P1 (G protein-linked)

and P2, with adenosine as the endogenous ligand for P1 and
adenosine triphosphate (ATP) and other purine and pyrimidine
nucleotides activating P2 receptors. Adenosine is primarily
derived from the hydrolysis of ATP by ectonucleotidases. The
P1 family are G protein-coupled receptors, subdivided into A1,
A2a, A2b, and A3 subtypes; all have been cloned. Interest in
the P1 receptor stems from the fact that adenosine (which is
found at virtually every synapse that has been examined)
exhibits an extraordinary constellation of activities. Adenosine
(and its agonists) mainly inhibits the evoked release of
neurotransmitters, hyperpolarizes postsynaptic membranes,
decreases locomotor activity, possesses sedative and
anticonvulsant activities, increases coronary and cerebral
blood flow, produces bronchoconstriction, reduces
neurodegeneration that occurs with stroke, and at high doses
promotes catalepsy. Except for the A3 subclass, the P1 family
of purinoreceptors is blocked by xanthines, with the classic
antagonists being caffeine and theophylline.

The P2 receptor family with ATP as the major agonist can be

divided into two major classes, designated P2x and P2y, each
with seven members to date. The former is coupled to ligand-
gated ion channels and the latter is a G protein-linked family
with the usual seven transmembrane domains. ATP, acting on
P2X ion channel receptors, is considered to be a cotransmitter
in the peripheral nonadrenergic, noncholinergic nervous
system, where it is released with norepinephrine or ACh. ATP
also has been shown to mediate fast synaptic transmission in
mammalian neurons. Subclasses of the P2Y receptor are
activated by uridine triphosphate and ADP. The lack of
selective antagonists to the P2 receptor has hampered
understanding of the function of its subclasses.

Continued administration of agonists can cause many

receptors to desensitize and down regulate. Desensitization,
occurring on a time scale of minutes, is reflected by a
decreased response of the cell and is often related to receptor
phosphorylation. Downregulation of receptors is observed on
a time scale of hours after prolonged agonist exposure when
receptors are internalized and degraded. Conversely, and
predictably, continued administration of receptor antagonists
generally causes upregulation of receptors.

Many neuroactive agents act on receptors that are coupled to

the adenylyl cyclase system (see Fig. 4-8). The components
of this complex are the receptor, the catalytic portion of
adenylyl cyclase that converts ATP to cyclic adenosine
monophosphate, and two G proteins referred to as Gs and Gi
that are coupled to the catalytic unit of the enzyme. When the
receptor is stimulated (e.g., a 2-adrenergic receptor), the
coupling protein is Gs. Conversely, when the receptor is
inhibited (e.g., an 2-adrenergic receptor), Gi is the coupling
protein. Examples of adrenergic receptors whose activity is
linked to the adenylate cyclase complex are the receptors 2,
1, and 2 (but not 1, which may be coupled to phosphatidyl
inositide hydrolysis).

Finally, although not classified as receptors, another class of

proteins that deserves attention is the transport proteins.
With the exception of ACh, the action of all neurotransmitters
that are released is terminated primarily by reuptake into
their presynaptic terminals (amino acid transmitters can also
be taken up by glial cells). ACh is hydrolyzed by
acetylcholinesterase, and it is choline that is recaptured by
the cholinergic terminal. Neurotransmitter transport proteins
are relatively specific for each transmitter, are sodium-
dependent, exhibit high-affinity kinetics, and are dependent
on the membrane potential of the neuronal terminal. The
GABA and glutamate transporters have the ability to function
in reverse, transporting the transmitters out of the cell.
Cloned plasma membrane neurotransmitter transporters fall
into two families, one that includes transporters for
monoamines and GABA, containing 12 transmembrane
domains, and one for glutamate and aspartate, whose
topology is still uncertain but likely to contain at least eight
transmembrane domains. As discussed in the chapters on
each neurotransmitter, these reuptake systems are the basis
of the mechanism of action of many neuropharmacological
agents, particularly the antidepressant drugs and some drugs
of abuse such as cocaine.
The transporters operate at the plasma membrane of
terminals and are different from intracellular vesicular
transporters, which accumulate transmitters into synaptic
vesicles. Vesicular transporters, some of which have been
cloned, operate via a vacuolar type H+-pumping ATPase. This
proton pump generates an H+ electrochemical gradient
whereby the efflux of H+ is coupled to the reuptake of the
transmitter into the vesicles.

Because of space limitations, this chapter has not covered

such topics as the factors regulating synthesis and
degradation of receptors, subunit compositions of receptors,
complex kinetics, isolation techniques, or the chemical
compositions of these macromolecules. Students who are
interested in these subjects may consult a number of recent
papers and reviews listed below.

Figure 4-2. Ligand-receptor interactions plotted

in two ways, where [L] is the concentration of
ligand (drug, hormone, or neurotransmitter) and r
is the biological response, proportional to the
moles ligand bound per mole of protein. The
number of binding sites per molecule of protein is
designated by n.

Figure 4-3. Scatchard plots of the binding of a

ligand to a receptor. In (a) only one binding
affinity occurs, but in (b) both high- and low-
affinity binding sites are suggested.

Figure 4-4. Hill plots for receptor-ligand binding:

(a) noncooperative binding; (b) idealized plot of
cooperative binding with four sites; (c) a typical
Hill plot of multiple binding sites but less than four.
(Modified from Van Holde, 1971.)
Figure 4-5. Type I is the ionotropic or ligand-
gated receptor with a response time of
milliseconds. An example is the nicotinic
acetylcholine receptor. Type II is the metabotropic
or G protein-linked receptor, generally modulatory,
with a response time in minutes. Examples are the
-aminobutyric acidB (GABAB) and muscarinic
receptors. Type III is the steroid receptor localized
primarily on the nucleus rather than the cell
membrane. The response time is usually minutes
to hours, as exemplified by estrogen and thyroid
hormones. Type IV represents the tyrosine kinase
receptor, activated by insulin or growth factors,
with a response time of minutes.

Figure 4-6. Topographical representation of the

primary sequence of the human 2-adrenergic
receptor. The receptor protein is illustrated as
possessing seven hydrophobic regions, each
capable of spanning the plasma membrane, thus
creating extracellular and intracellular loops as
well as an extracellular terminus and a
cytoplasmic C-terminal region. (From Lefkowitz et
al., 1989.)

Figure 4-7. Nongenomic and genomic effects of

neuroactive steroids. The term neuroactive
steroids has been coined for steroids that interact
with neurotransmitter receptors. Modulation of
neuronal excitability by neuroactive steroids
occurs over a very short (milliseconds to a few
seconds) time period. The list in the upper left-
hand corner shows steroids that fulfill the criteria
for neuroactive steroids; the lower list gives
neurotransmitter receptors that are targets for
steroid modulation. The right-hand side of the
figure describes the classical model of steroid-
hormone action via the steroid-receptor cascade
 hormone action via the steroid-receptor cascade
at the genomic level, which takes place over
minutes to hours. The list on the right-hand side
gives the names of typical steroid hormones.
Certain steroids, such as 17 -estradiol and
progesterone, have to be defined both as steroid
hormones and as neuroactive steroids. BDZ,
benzodiazepine; DHEA-S, dehydroepiandrosterone
sulfate; ER, estrogen receptor; G, G protein; GR,
glucocorticoid receptor; HSP90, heat-shock protein
90; MR, mineralocorticoid receptor; PKA, protein
kinase A; PR, progesterone receptor; PS,
pregnenolone sulfate; R, receptor; THDOC,
tetrahydrodeoxycorticosterone; THP,
tetrahydroprogesterone. (From Rupprecht and
Holsboer, 1999.)

Figure 4-8. Components of a receptor-activated,

cyclic nucleotide-linked system. (From Lefkowitz
et al., 1984.)

SELECTED REFERENCES
Ariens, E. J. and A. J. Beld (1977). The receptor concept in
evolution. Biochem. Pharmacol. 26, 913.

Berman, D. M. and A. G. Gilman (1998). Mammalian RGS

proteins: barbarians at the gate. J. Biol. Chem. 273, 1269.

Birdsall, N. J. M., N. M. Nathanson, and R. D. Schwarz (2001).

Muscarinic receptors: it's a knockout. Trends Pharmacol. Sci.
22, 215.

Burnstock, G. (2001). Purine-mediated signalling in pain and

visceral perception. Trends Pharmacol. Sci. 22, 102.
Burnstock, G. (1999). Current status of purinergic signalling
in the nervous system. Prog. Brain Res. 120, 3.

Changeux, J. P., J. Thiery, Y. Tung, and C. Kittel (1976). On

the cooperativity of biological membranes. Proc. Natl. Acad.
Sci. U.S.A. 57, 335.

Cheng, Y.-C. and W. H. Prusoff (1973). Relationship between

the inhibition constant (Ki) and the concentration of inhibitor
which causes 50 percent inhibition (I50) of an enzymatic
reaction. Biochem. Pharmacol. 22, 3099.

Clapham, D. E. and E. J. Neer (1997). G protein beta gamma

subunits. Annu. Rev. Pharmacol. Toxicol. 37, 167.

Clark, A. J. (1937). General pharmacology. In Heffter's

Hanbuch d-exp. Pharmacol. Erg. Band 4. Springer, Berlin.

Girault, J. A. and P. Greengard (1999). Principles of Signal

Transduction. In Neurobiology of Mental Illness (D. S.
Charney, E. J. Nestler, and B. S. Bunney, eds.). Oxford
University Press, New York, p. 37.

Jonas, E. K. and L. K. Kaczmarek (1999). The inside story,

subcellular mechanisms of neuromodulation. In Beyond
Neurotransmission (P. S. Katz ed.). Oxford University Press,
New York, p. 83.

Karlin, A. (1967). On the application of a plausible model of

allosteric proteins to the receptor for acetylcholine. J. Theor.
Biol. 16, 306.

Kenakin, T. P. (1997). Agonist-specific receptor confirmations.

Trends Pharmacol. Sci. 18, 493.

Kenakin, T. P., R. A. Bond, and T. I. Bonner (1992). Definition

of pharmacological receptors. Pharmacol. Res. 44, 351.

Laduron, P. M. (1984). Criteria for receptor sites in binding

studies. Biochem. Pharmacol. 33, 833.

Lefkowitz, R. J., B. K. Kobilka, and M. G. Caron (1989). The

new biology of drug receptors. Biochem. Pharmacol. 38,
2941.

Lefkowitz, R. J., M. G. Caron, and G. L. Stiles (1984).

Mechanisms of membrane-receptor regulation. N. Engl. J.
Med. 310, 1570.

Lester, H. H., S. Mager, M. W. Quick, and J. L. Corey (1994).

Permeation properties of neurotransmitter transporters. Annu.
Rev. Pharmacol. Toxicol. 34, 219.

Monod J., J. Wyman, and J. P. Changeux (1960). On the

nature of allosteric transitions. J. Biol. Chem. 12, 88.

Morris, A. J. and C. C. Malbon (1999). Physiological regulation

of G protein-linked signalling. Physiol. Rev. 79, 1373.

Nelson, N. (1998). The family of Na+/Cl- neurotransmitter

transporters. J. Neurochem. 71, 1783.

Neubig, R. R. (1994). Membrane organization in G-protein

mechanisms. FASEB J. 8, 939.

Rakic, P., P. S. Goldman-Rakic, and D. Gallager (1988).

Quantitative autoradiography of major neurotransmitters in
monkey striate and extrastriate cortex. J. Neurosci. 8, 3670.

Rudnick, G. (1990). Bioenergetics of neurotransmitter

transport. J. Bioenerg. Biomembr. 30, 173.

Rupprecht, R. and A. F. Holsboer (1999). Neuroactive

steroids: mechanism of action and
neuropsychopharmacological perspectives. Trends Neurosci.
22, 410.

Seal, R.P. and S.G. Amara (1999). Excitatory amino acid

transporters: a family in flux. Annu. Rev. Pharmacol. Toxicol.
39, 431.

Singer, S. J. and G. L. Nicolson (1972). The fluid mosaic

model of the structure of cell membranes. Science 175, 720.

Strader, C. D., T. M. Fong, M. R. Tota, D. Underwood, and

R.A.F. Dixon (1994). Structure and function of G protein-
coupled receptors. Annu. Rev. Biochem. 63, 101.

Suenningsson, P., C. LeMoine, G. Fisone, and B. B. Fredholm

(1999). Distribution, biochemistry and function of striatal
adenosine A2A receptors. Prog. Neurobiol. 59, 355.

Takai, Y. (2001). Small GTP-binding proteins. Physiol. Rev. 81,

153.

Van Holde, K. E. (1971). Physical Biochemistry. Prentice-Hall,

Englewood Cliffs, NJ.

(+/-) Show / Hide Bibliography

Send Feedback Teton Server (5.0.0) - ©2007 Teton Data Title Updates
Customer Service Systems User
800.901.5494 Send Us Your Comments Responsibilities
Training Center
What's New
5. Modulation of Synaptic
Transmission
INTRODUCTION
Contrary to what one might assume, the more we learn about
intercellular communication in the nervous system, the more
complicated the situation appears. Up to about the mid-
1970s, synaptologists smugly focused on a straight point-to-
point transmission, where a presynaptically released
transmitter impinged on a postsynaptic cell. Gradually,
situations emerged where previously identified
neurotransmitters were observed to not act in this fashion
but, rather, to modulate transmission. These departures from
the previous norm and the continuing discovery of peptides
and small molecules such as adenosine, which exhibited
neuroactivity but did not appear to be transmitters in the
classic sense, supported the broader view of modulation of
synaptic transmission. In retrospect, it is a concept that
should have been apparent early on since it imparts to neural
circuitry an extraordinary degree of flexibility, which is
necessary when considering the mechanisms of behavioral
changes. Also, modulation is not a neurophysiological
property seen only in higher forms. In Cole Porter's words,
"Birds do it, bees do it, even educated fleas do it."

DEFINITIONS
Primarily via the activation of a receptor, synaptic
transmission may be modulated either presynaptically or
postsynaptically. With presynaptic modulation, regardless of
the mechanism, the ultimate effect is a change in the amount
of transmitter released. With postsynaptic modulation, the
ultimate effect is a change in the firing pattern of the
postsynaptic neuron or in the activity of a postsynaptic tissue
(e.g., blood vessel, gland, muscle). Because some confusion
has arisen as to the correct nomenclature of pre- and
postsynaptic receptors, an explanation is in order. What may
be classified as a presynaptic receptor on neuron B may be a
postsynaptic receptor of neuron A that is making an
axoaxonic contact with neuron B (see Fig. 2-3). Thus,
depending on which neuron is being investigated, the receptor
will be denoted as either pre- or postsynaptic.

Procedures to localize activity at the presynaptic receptor

level in a terminal include (1) the use of synaptosomes; (2)
the addition of tetrodotoxin to the preparation to block action
potentials in neighboring interneurons; (3) patch clamping
presynaptic terminals; (4) imaging techniques using
appropriate dyes; or (5) where feasible, either chemical
destruction of terminals or lesioning of the neuron and then
demonstration by ligand binding of loss of the receptor. To
complete the nomenclature on presynaptic receptors, an
autoreceptor is located on the terminal or somatic-dendritic
area of a neuron that is activated by the transmitter(s)
released from that neuron. A heteroreceptor is a presynaptic
receptor activated by a modulating agent that originates from
a different neuron or cell. As discussed in Chapter 1,
modulators differ from transmitters in that they have no
intrinsic activity but modulate ongoing neural activity.
However, a transmitter may modulate at a concentration that
is subthreshold for transmitter activity. As we will now detail,
there exists an extraordinary number of possibilities for
altering the point-to-point synaptic transmission mentioned
above.

Presynaptic modulation can be affected by the following:

1. Receptor activation of a presynaptic neuron causing the

following:

a. A change in the firing frequency in the presynaptic neuron.

This is probably the most common type of modulation,
particularly in the central nervous system (CNS), where it can
be assumed that the firing rate of virtually every neuron is
governed by inputs on dendrites, soma, or axons. The firing
rate determines the frequency of impulse conduction, hence
the spread of action potentials into terminals or varicosities
and the amount of transmitter that is liberated.

b. A change in the transport or reuptake of a transmitter or

precursor or in the synthesis, storage, release, or catabolism
of a transmitter. All of these possibilities will result in a
change in the concentration of a transmitter at the terminal.
In practice, it has been shown that presynaptic activation of
biogenic amine neurons promotes the phosphorylation of both
tyrosine and tryptophan hydroxylase, which increases the
synthesis of norepinephrine, dopamine, and serotonin.
Curiously, phosphorylation of the pyruvate dehydrogenase
complex causes a decrease in enzyme activity and in theory
would decrease acetylcholine levels of (ACh) and the amino
acid transmitters. To date, however, modulation of this
enzyme activity by presynaptic receptor activation has not
been observed.

c. An effect on ion conductances at the terminal. The three

ions and their respective channels that one might focus on
would be K+, Ca2+, and Cl-. Endogenous neuroactive agents
or drugs could alter transmitter release by opening or
blocking the channels or changing the kinetics of channel
open time via the possible mediation of protein
phosphorylation or other second messengers.

2. A direct effect of neuropharmacological agents on some

element of the release process. This could be an effect of the
modulating agent on vesicular apposition to a terminal,
fusion, or fission.

Postsynaptic modulation can be affected by the following:

1. A long-term change in the number of receptors. This is not
observed under normal physiological conditions. It is,
however, commonly noted pharmacologically where the
administration of a receptor agonist for a period of time will
result in downregulation of the receptor, and, conversely,
treatment with a receptor antagonist increases receptor
density.

2. A change in the affinity of a ligand for a receptor. The now

classic example is the salivary nerve of the cat, where both
ACh and vasoactive intestinal peptide (VIP) are co-localized.
When VIP is released upon electrical stimulation, it increases
the affinity of ACh for the muscarinic receptor on the salivary
gland up to 10,000-fold, with a consequent increase in
salivation.

3. An effect on ionic conductances. As discussed in 1a above,

this is frequency modulation. It is a postsynaptic effect on the
first neuron in a relay, but it would be classified as a
presynaptic effect on the second neuron. References to all of
the neurons and modulating agents that have been
investigated are given in the reviews by Starke et al. (1989),
Chesselet (1994), and Levitan and Kaczmarek (2002). These
reviews can be summarized by stating that virtually every
neuronal pathway is modulated and virtually every
endogenous neuroactive agent has been shown in one
preparation or another to be capable of affecting synaptic
transmission. All of this information is descriptive. The
question of the second-messenger systems that may be
involved in regulating synaptic activity is addressed in the
following section. Figure 5-1 depicts a major pathway for
modulation of synaptic transmission. Rapidly accumulating
evidence suggests that, in most cases of receptor-activated
inhibitory presynaptic modulation, the ultimate effect is to
open K channels. This hyperpolarizes terminals; less Ca2+
enters; and, as a consequence, less transmitter is released.
Another major presynaptic mechanism is the inhibition of a
calcium channel that would also decrease the evoked release
of a transmitter. With the less common receptor-activated
excitatory modulation, closing of a K channel has been
implicated. A well-documented postsynaptic modulatory
mechanism is that of a transmitter that inhibits a voltage-
dependent K current, causing a subsequent depolarizing
stimulus to produce an enhanced response.

Figure 2-3. Twelve steps in the synaptic

transmission process are indicated in this idealized
synaptic connection. Step 1 is transport down the
axon. Step 2 is the electrically excitable
membrane of the axon. Step 3 involves the
organelles and enzymes present in the nerve
terminal for synthesizing, storing, and releasing
the transmitter, as well as for the process of active
reuptake. Step 4 includes the enzymes present in
the extracellular space and within the glia for
catabolizing excess transmitter released from
nerve terminals. Step 5 is the postsynaptic
receptor that triggers the response of the
postsynaptic cell to the transmitter. Step 6 shows
the organelles within the postsynaptic cells which
respond to the receptor trigger. Step 7 is the
interaction between genetic expression of the
postsynaptic nerve cell and its influences on the
cytoplasmic organelles that respond to transmitter
action. Step 8 includes the possible "plastic" steps
modifiable by events at the specialized synaptic
contact zone. Step 9 includes the electrical portion
of the nerve cell membrane that, in response to
the various transmitters, is able to integrate the
postsynaptic potentials and produces an action
potential. Step 10 is the continuation of the
information transmission by which the
postsynaptic cell sends an action potential down
 its axon. Step 11, release of transmitter, is
subjected to modification by a presynaptic
(axoaxonic) synapse; in some cases, an analogous
control can be achieved between dendritic
elements. Step 12, release of the transmitter from
a nerve terminal or secreting dendritic site, may
be further subjected to modulation through
autoreceptors that respond to the transmitter
which the same secreting structure has released.
Glia (G) can accumulate (4) released transmitters.

Figure 5-1. Major pathways for modulation of

synaptic transmission. DG, diacylglycerol; IP3,
inositol trisphosphate; NO, nitric oxide; CO,
carbon monoxide; cGMP, cyclic guanosine
monophosphate; cAMP, cyclic adenosine
monophosphate.

SECOND MESSENGERS

Introduction
Three major biochemical cascades and two new gaseous
messengers, nitric oxide (NO) and carbon monoxide, have
been described:

1. Protein phosphorylation. Following the identification of

cyclic nucleotides by Sutherland and associates and the
implication that they act as a second-messenger system
preceded by an initial nerve impulse or hormonal signal,
Krebs and colleagues demonstrated a multistep sequence of
events that linked cyclic adenosine monophosphate (cAMP)
generation in muscle to the regulation of carbohydrate
metabolism. Since that time, the cyclic nucleotides have been
shown to regulate an enormous diversity of processes ranging
from axoplasmic transport to neuronal differentiation and
including transmitter synthesis and release and the
generation of postsynaptic potentials. All of these effects are
attributable to the cAMP- or cyclic guanosine monophosphate
(cGMP)-activating protein kinases and thus protein
phosphorylation. Second-messenger activity, achieved via
protein phosphorylation, can be mediated by cAMP- and
cGMP-dependent protein kinases as well as by calcium-
calmodulin-dependent protein kinase and calcium-
phosphatidylserine- or calcium-diacyglycerol-dependent
protein kinase (protein kinase C). The phosphoryl acceptor in
these proteins is the hydroxyl group of serine, threonine, or
tyrosine, with the first two being the most prominent. All
protein kinases can themselves be autophosphorylated, a
process that usually increases kinase activity. Although in
most instances biological activity results from kinase-
activated phosphorylation of a substrate protein, in some
cases it is phosphatase-activated dephosphorylation of a
phosphorylated protein that produces the biological response.
Currently, eight phosphoprotein phosphatases have been
identified (protein phosphatase-2B is also referred to as
calcineurin and is regulated by proteins referred to as
immunophilins). Phosphatases fall into two broad classes,
phosphoserine/phosphothreonine phosphatases and
phosphotyrosine phosphatases.

Yet another level of regulation, noted above, is suggested by

the finding that protein phosphatase activity can be inhibited
by other proteins, the most interesting of which is DARPP-32,
found in D1 dopaminoceptive neurons. DARPP-32 (dopamine-
and cAMP-regulated phosphoprotein of 32 kDa), by acting as
a protein phosphatase inhibitor when it is phosphorylated, can
regulate the postsynaptic effects of dopamine in
dopaminoceptive cells. Phosphorylated DARPP-32 is
inactivated by protein phosphatase-2B. In some instances,
Ca2+ acts as a second messenger without the participation of
protein phosphorylation. The diversity of signals that are
coupled to protein phosphorylation is depicted in Figure 5-2.
(For a molecular illustration of receptor coupling, see Fig. 4-
6.)

Despite the vast number of systems in which protein

phosphorylation is implicated, around 500, there are currently
only about a dozen cases where direct evidence links this
process to modulation of synaptic transmission, either pre- or
postsynaptically. It is not yet known if the proteins that make
up the channels are phosphorylated or if the phosphorylated
proteins are morphologically associated with the channels. At
any rate, with over four dozen proteins in the brain that are
known to be phosphorylated (Table 5-1), the story is far from
complete. Questions that remain to be answered include the
regulation of the enormous number of steps in the cascade
and the substrate specificity of the phosphodiesterases,
protein kinases, and phosphoprotein phosphatases. What
cannot be overemphasized is the involvement of
phosphorylation in virtually every aspect of neuronal function,
including transmitter release, synthesis, and reuptake; ion
channel activity; regulation of receptors; cytoskeleton
proteins; gene expression; and possibly short-term memory
(Nestler Greengard, 1999).

2. Phosphoinositide hydrolysis. In 1953, Hokin and Hokin

showed that the incorporation of inorganic phosphate (Pi) into
phosphatidylinositol (PI) and phosphatidic acid (PA) in
pancreatic slices was stimulated by ACh and ultimately
resulted in the release of amylase. This receptor-activated
hydrolysis of phosphoinositides is referred to as the
phosphatidylinositol effect. For nearly 30 years after this
report, the literature on this effect was replete with the
traditional scientific jargon "it is tempting to speculate that
...," with no one having solid evidence as to whether the
phosphoinositides or the inositol phosphates were the
message and, if so, exactly what was the medium for the
exchange. That this situation has now dramatically improved
is shown in Figure 5-3.

The signals that initiate this transduction process in neuronal

systems include ACh, norepinephrine, serotonin, histamine,
glutamate bradykinins, substance P, vasopressin, thyrotropin-
releasing hormone, neurotensin, VIP, nerve growth factor, and
angiotensin acting on brain, sympathetic ganglia, salivary
glands, iris smooth muscle, adrenal cortex, and neuronal
tumor cells. Specific receptors that have been implicated are
muscarinic cholinergic receptors, 1-adrenergic receptors, the
H1 histaminergic receptor, substance P, and the V1
vasopressin receptor. In each case, Ca2+ appears to be the
intracellular second messenger that activates
phosphoinositide hydrolysis. Like the specific guanosine
triphosphate (GTP)-binding proteins that link receptors to the
adenylate cyclase system discussed earlier, a specific GTP-
binding protein is also coupled to the phosphodiesterase that
catalyzes the hydrolysis of phosphatidylinositol 4,5-
bisphosphate. In addition to the G protein-linked receptor, it
is now known that a tyrosine kinase-linked receptor is coupled
to a specific phosphodiesterase (phospholipase C), which
yields inositol 1,4,5-trisphosphate (InsP3) and diacylglycerol.

Thus, as noted in Figure 5-3, the key participants in the

transduction process are a G protein-linked receptor, a
tyrosine kinase-linked receptor, and phosphodiesterases,
yielding two separate second messengers, the water-soluble
InsP3 and the lipid-soluble diacylglycerol. The former
mobilizes calcium (released in a wave form, i.e., oscillatory),
which can act through calmodulin to phosphorylate specific
proteins, and the latter, by activating protein kinase C, a
calcium-phosphatidylserine-dependent family of protein
kinases, also phosphorylates specific proteins. The InsP3
receptor, associated with the smooth endoplasmic recticulum,
is a tetramer of identical subunits that is regulated by
adenosine triphosphate (ATP) and cAMP. Since diacylglycerol
can activate guanylate cyclase to produce cGMP, cGMP-
dependent activity (protein kinase or otherwise) must be
considered. With an assumed ambidexterity of neuronal cells,
these two arms could function singly, cooperatively, or
antagonistically, depending on the situation, thus providing
subtle variations on the modulatory mechanism. In addition,
as noted in Figure 5-3, calcium may produce a physiological
response directly without invoking activation of calmodulin, so
yet another control is indicated. On the subject of control, the
activities of the various kinases, esterases, and
phospholipases in the PI cycle would be expected to be vital
control points. For example, five isoforms of phospholipase C
have now been identified, of which some are enriched in
specific brain areas. Free inositol in the brain must be derived
from glycolysis since plasma inositol cannot pass the blood-
brain barrier to any significant degree. Glycolysis, therefore,
would be another regulatory factor in the response
mechanism.

Finally, although the origin of the mobilized calcium is now

clear (it is released from endoplasmic reticulum and
mitochondria), some controversy remains as to whether
phosphoinositide hydrolysis releases only internal calcium or
whether external calcium influx is also invoked. Current
evidence suggests that for neuronal modulation both or either
may be involved, depending on the preparation. The same
answer can also be given as to whether the PI effect is
presynaptic or postsynaptic. A complication in the PI cycle
that has recently surfaced is the finding that the inositol
trisphosphate that is produced is not always or only
Ins(1,4,5)P3 but may be Ins(1,3,4)P3 as well as InsP4, InsP5,
and InsP6. The physiological role of these polyphospate
compounds is unknown, but they may function as phosphate
donors.

3. Eicosanoids (arachidonic acid metabolites). Arachidonic

acid, synthesized from dietary linoleic acid, derived on
demand by either a G protein-regulated phospholipase A2 or
diglyceride lipase activation (Fig. 5-4), yields a bewildering
array of bioactive metabolites, as shown in Fig. 5-5. The three
major groups are prostaglandins, thromboxanes, and
leukotrienes.

It is well known that the eicosanoids, particularly the

prostaglandin series, play an important modulatory role in
nervous tissue, but it is difficult to write a lucid account of
specifically how and where they act. This is primarily due to
the fact that they are not stored in tissue but synthesized on
demand, particularly in pathophysiological conditions; they
act briefly (some with a half-life of seconds) and at extremely
low concentrations (10-10 M). Although indomethacin is a
good inhibitor of cyclooxygenase, blocking the conversion of
arachidonic acid to prostaglandins, there are few specific
inhibitors available to block lipoxygenase and epoxygenase.
Thus, although it had been postulated that the E series of
prostaglandins modulates noradrenergic release; blocks the
convulsant activity of pentylenetetrazol, strychnine, and
picrotoxin (possibly by increasing the level of -aminobutyric
acid in the brain); and increases the level of cAMP in cortical
and hypothalamic slices, these effects were noted in vitro with
the addition of substantial amounts of the prostaglandins.
There was very little evidence in intact animals to support
these neuronal findings, and since we all believe in "in vivo
veritas," the physiological relevance of the effect was in
doubt.

Recently, however, direct evidence has implicated arachidonic

acid and lipoxygenases as second messengers. The cascade
begins with the binding of a neuroactive agent to its receptor.
Then, according to findings from the Axelrod laboratory, the
receptor is coupled to G proteins, which may either activate
or inhibit phospholipase A2, although this has not been
conclusively established for all neural tissues. The activated
enzyme promotes the release of arachidonic acid, which will
then act intracellularly as a second messenger. Arachidonic
acid and its metabolites can also leave the cell to act
extracellularly as first messengers on neighboring cells.
Eicosanoids have been shown to mediate the somatostatin-
induced opening of an M channel in hippocampal pyramidal
cells and the release of VIP in mouse cerebral cortical slices.
It is thus becoming clear, despite enormous technological
difficulties in assaying eicosanoids, that these agents are
major messengers. Before we leave the arachidonic acid
story, an exciting new finding should be mentioned.
Cannabinoids are a group of psychoactive compounds found
in marijuana, of which the principal component is 9-
tetrahydrocannabinol. In binding studies, two cannabinoid
receptors were isolated and cloned, referred to as CB1 and
CB2 receptors. The CB1 receptor is found in the brain and
peripheral nervous system, while the CB2 receptor is
expressed mainly in the immune system. Both receptors are
G protein-coupled. Subsequently, several cannabinoid ligands
were identified, with the major ones in brain being
arachidonylethanolamide (a.k.a., anandamide) and 2-
arachidonylycerol. Recent evidence suggests that anandamide
may react with G protein-coupled receptors that are not CB1
or CB2. Anandamide is synthesized by the transfer of an
arachidonic acid-containing phospholipid to the amine group
of phosphatidylethanolamine to form n-arachidonoyl-
phosphatidylethanolamine, followed by phospholipase and
hydrolysis to yield arachidonylethanolamide (anandamide)
and phosphatidic acid. Cannabinoids are hydrolyzed by fatty
acid amide hydrolase. Cannabinoid agonists appear to be
involved in neuroprotection, as appetite stimulants,
antiemetics, and analgesics. The naturally occurring
canabinoid 2-arachidonylglycerol appears to have hypotensive
activity.

Figure 4-6. Topographical representation of the

primary sequence of the human 2-adrenergic
receptor. The receptor protein is illustrated as
possessing seven hydrophobic regions, each
capable of spanning the plasma membrane, thus
creating extracellular and intracellular loops as
well as an extracellular terminus and a
cytoplasmic C-terminal region. (From Lefkowitz et
al., 1989.)

Figure 5-2. Schematic diagram of the role played

by protein phosphorylation in mediating some of
the biological effects of a variety of regulatory
agents. Many of these agents regulate protein
phosphorylation by altering intracellular levels of a
second messenger, cyclic adenosine
monophosphate (AMP), cyclic guanosine
monophosphate (GMP), or Ca2+. Other agents
appear to regulate protein phosphorylation
through mechanisms that do not involve these
second messengers. Most drugs regulate protein
phosphorylation by affecting the ability of first
messengers to alter second-messenger levels
(curved arrows). A small number of drugs (e.g.,
phosphodiesterase inhibitors, Ca2+ channel
blockers) regulate protein phosphorylation by
directly altering second-messenger levels (straight
arrows). EGF, epidermal growth factor; PDGF,
platelet-derived growth factor; NGF, nerve growth
factor; FGF, fibroblast growth factor; TGF,
transforming growth factor; SC, somatomedin C.
(Modified from Nestler and Greengard, 1989.)

Figure 5-3. Receptor-activated phosphoinositide

metabolism. The binding of an agonist to a
receptor on the plasma membrane stimulates the
hydrolysis of phosphatidylinositol 4,5-bis-
phosphate [Ptdins (4,5)P2] by a
phosphodiesterase (PDE, phosphoinositidase,
phospholipase C), a specific phospholipase whose
activity is controlled by a guanine nucleotide
regulatory protein to form inositol 1,4,5-
trisphosphate (InsP3) and diacylglycerol (DG). The
former binds to a receptor (R2) on the
endoplasmic reticulum to release calcium, which
can directly produce a biological response or can
activate calmodulin kinase to promote protein
phosphorylation and a subsequent biological
response. In some cells (e.g., mouse atria,
neuroblastoma, and glioma hybrid NG108-15),
receptor-activated production of InsP3 requires
extracellular Ca2+. The latter parallel arm of the
cycle, diacylglycerol, can also promote a biological
response via the production of prostaglandins,
thromboxanes, and leukotrienes from released
arachidonic acid (arachidonic acid has also been
reported to stimulate guanylate cyclase to
generate cyclic guanosine monophosphate) or via
stimulation of protein kinase C (C-kinase) and
subsequent protein phosphorylation.
Diacylglycerol has also been reported to promote
fusion of synaptic vesicles to terminal membrane.
Diacylglycerol can also be generated from
phosphatidylcholine via phospholipase D followed
by phosphatidic acid phosphatase. The
phosphoinositides are synthesized from inositol
with cytidine diphosphate:diacylglycerol (CDP-DG)
as intermediary and the stepwise phosphorylation
by kinases (a and b). As shown in the figure,
lithium blocks the cycle by inhibiting inositol-1-
phosphatase. Although the antimanic activity of
lithium has been ascribed to this inhibitory effect,
the evidence is not compelling. ATP, adenosine
 triphosphate; GTP, guanosine triphosphate; GDP,
guanosine diphosphate (Modified from Berridge
and Irvine, 1984.)

Figure 5-4. Pathways for the generation and

metabolism of arachidonic acid. Arachidonate can
arise directly from phospholipids through the
action of phospholipase A2 or prior action of
phospholipase C, followed by the action of
diglyceride lipase. Alternatively, the diglyceride
may be phosphorylated to phosphatidic acid by
the action of diglyceride kinase, and arachidonate
then can be released through the action of
phospholipase A2. The released arachidonate may
then be metabolized by lipoxygenase,
cyclooxygenase, or epoxygenase enzymes to form
leukotrienes (LTs), hydroxyeicosatetraenoic acids
(HETEs), prostaglandins (PGs), thromboxanes
(TBXs), and epoxides. (Modified from Axelrod et
al., 1988.)

Figure 5-5. Metabolism of arachidonic acid (AA)

by cyclooxygenase, lipoxygenases (LO),
epoxygenase, and corresponding end products
(prostanoids). PGI2, prostaglandin I2; HETE,
hydroxyeicosatetraenoic acid. (From Schaad et al.,
1991.)

Nitric Oxide
In 1980, Furchgott and Zawadzki observed that stimulation of
the endothelium released a factor that relaxed blood vessels.
This factor, referred to as the endothelium-derived relaxing
factor, was subsequently identified by the Moncada laboratory
as NO. Acting as a second messenger, NO is now known to be
involved in an incredible number of systems. Aside from its
role in the nervous system, NO is a mediator in the
cardiovascular, renal, pulmonary, endocrine, and immune
systems.

NO is synthesized from arginine via the enzyme NO synthase,

a flavin adenine dinucleotide and flavin mononucleotide
enzyme, requiring molecular O2 and with reduced
nicotinamide-adenine dinucleotide phosphate as coenzyme
and tetrahydrobiopterin as cofactor. The neuronal and
endothelial enzyme that is constitutively expressed is
activated by Ca2+ and calmodulin, whereas the macrophage
enzyme that is inducible by cytokines is not. The synthetic
reaction sequence in the brain is shown in Figure 5-6.

NO synthase is inhibited by a variety of arginine analogs, and

this finding has proved to be invaluable in delineating the
functions of NO in vivo. The constituitive neuronal NO
synthase is regulated by phosphorylation. A number of
protein kinases can phosphorylate the enzyme, and this
process decreases enzyme activity. NO itself is destroyed by
reacting with hemoglobin and other iron-containing
compounds. Interestingly, vasodilators such as nitroprusside
and nitroglycerin produce NO, and this is considered to be
their mechanism of action.

NADPH diaphorase immunoreactivity is co-localized in most,

but not all, cells with NO synthase activity; and this has been
used to map the distribution of neurons that release NO.
Currently, three isoforms of NO synthase have been reported
that arise from distinct genes and share approximately 50%
amino acid identity. An unusual feature of this enzyme is that
it is a member of the family of cytochrome P-450 proteins,
which catalyze the hydroxylation of a variety of metabolites
as well as drugs.

So much for the background. Now to the neuronal effects of

NO. First, it should be noted that NO is not present in
synaptic vesicles, is not released from terminals by
exocytosis, and is not stored in any reservoir. Rather, it is
released upon stimulation and diffuses from a neuronal (and
glial) population to act on enzymes and other elements. The
first and primary action of this gaseous molecule is to
complex the iron of soluble guanylyl cyclase to stimulate the
enzyme and increase the concentration of cGMP. Another
target, aside from guanylyl cylase, is cytosolic adenosine
diphosphate (ADP)-ribosyl transferase. The consequence of
this latter activity is unknown. The rise in cGMP will activate
cGMP-dependent protein kinases, which catalyze the
phosphorylation of substrate proteins, largely unidentified,
which then give rise to myriad effects. The first indication of a
role of NO in neuronal systems came from the Garthwaite
laboratory, when investigators found that in cerebellar slices
excitatory amino acids led to NO release, as did stimulation of
nonadrenergic, noncholinergic peripheral neurons.
Subsequent research has provided evidence that NO synthase
inhibitors block N-methyl-D-aspartate (NMDA) receptor
activation, preventing the well-documented NMDA-induced
neurotoxicity. However, the situation is confusing because
prolonged glutamate stimulation produces high NO levels that
can kill neuronal cells. In addition to this activity and its
relaxation effect in stimulating peripheral nonadrenergic
noncholinergic neurons, NO is thought to play a role in both
long-term potentiation, where it acts as a retrograde
messenger, and long-term depression. These functions
concerning synaptic plasticity are complex and remain to be
resolved. Clearly, the role of NO in both physiological and
pathophysiological processes is still evolving.

Figure 5-6. Synthesis of nitric oxide.

Carbon Monoxide
Another gaseous molecule has recently surfaced as a
neuromodulating second messenger that also activates
guanylyl cyclase. This is carbon monoxide (CO). CO is
generated via heme oxygenase, which catalyzes the
conversion of heme to biliverdin with the liberation of CO. Two
heme oxygenases, expressed by separate genes, have been
described. Heme oxygenase-1 is inducible and localized
mainly in the spleen and liver, with a small (but inducible)
concentration in glial cells and in a few neurons. In contrast,
heme oxygenase-2 is constitutive, not inducible, and found in
high concentrations in the brain, particularly cerebellum,
olfactory bulb, and hippocampus. It is not present in glial cells
under normal conditions but can be induced there. It is
activated by protein kinase C.

As noted earlier, CO, like NO, raises the level of cGMP. While
the localization of NO synthase does not always correlate with
the localization of guanylyl cyclase, the localization of heme
oxygenase and guanylyl cyclase is virtually identical and
coincides with the presence of cytochrome P-450 reductase, a
necessary electron donor for heme oxygenase as well as NO
synthase.

Two other possible second messengers are protein carboxyl

methylation and phospholipid methylation. Both processes
involve S-adenosylmethionine as the methyl donor. Although
it has been shown that protein carboxyl methylation inhibits
calmodulin-linked enzymes and that phospholipid methylation
alters terminal membrane viscosity, these activities currently
are not seen as playing an important role in modulation.

Among endogenous modulators, the neuroactive peptides are

by far the most mysterious with respect to their widespread
activity. These agents can be released presynaptically,
postsynaptically in response to presynaptic receptor
activation, from glial cells, or from blood vessels. Another
mysterious modulator is adenosine, which is found at virtually
every synapse that has been examined. Electrophysiologically,
it tends to inhibit the evoked release of transmitters; but it
also acts postsynaptically, exhibiting a variety of behavioral
effects ranging from evoking premature arousal in hibernating
ground squirrels to anticonvulsant activity, increasing cerebral
blood flow, and interacting with the benzodiazepine receptor.
Finally, recent evidence implicates D-serine, synthesized in
astrocytes, as a neuromodulator of NMDA receptors (see
Chapter 6).

Reflecting on the mechanisms available to nervous tissue to

modulate synaptic transmission, one cannot fail to be
overwhelmed by the almost infinite possibilities that provide
the fine tuning for behavioral changes. It may well turn out
that the key player in this scenario is calcium. Since Ca2+ is
released as a wave, its concentration as well as its
translocation at discrete sites might dictate the modulatory
effect of the attachment of a ligand to its receptor. It may also
be the key to nonsynaptic cellular effects, such as cell
movement or proliferation, gene expression, and metabolism,
that involve second-messenger systems. At any rate, one
should be aware that the ultimate effect of cascades of
second-messenger systems, transcription factors,
neurotrophic factors, transport factors, gene expression, or
the endless numbers of signaling factors (irritatingly referred
to by acronyms) is to cause a neuron to fire at a certain
frequency or not to fire. It is that simple except that a neuron
may have 10,000 inputs dictating its ultimate effect on other
neurons to which it is coupled.

Somewhere over the rainbow, one hopes that one day

somebody will reveal the grand design in neuronal
communication that leads to behavioral changes ranging from
the subtle to the dramatic.
SELECTED REFERENCES
Arancio, O., M. Kiebler, C. J. Lee, V. Lev-Ram, R. Y. Tsien, E.
R. Kandel, and R. D. Hawkins (1996). Nitric oxide acts directly
in the presynaptic neuron to provide long-term potentiation in
cultured hippocampal neurons. Cell 87, 1025.

Axelrod, J., R. M. Burch, and C. L. Jelsema (1988). Receptor-

mediated activation of phospholipase A2 via GTP-binding
proteins: arachidonic acid and its metabolites as second
messengers. Trends Neurosci. 11, 117.

Baranano, D. E., C. D. Ferris, and S. H. Snyder (2001).

Atypical neural messengers. Trends Neurosci. 24, 99.

Berridge, M. J. (1993). Inositol trisphosphate and calcium

signaling. Nature 361, 315.

Berridge, M. J., P. Lipp, and M. D. Bootman (2000). The

versatility and unversatility of calcium signalling. Nat. Rev.
Mol. Cell. Biol. vol 1, p. 11.

Buttner, N., S. A. Siegelbaum, and A. Volterra (1989). Direct

modulation of Aplysia S-K+ channels by a 12-lipoxygenase
metabolite of arachidonic acid. Nature 342, 553.

Chesselet, M.-F. (1994). Presynaptic regulation of

neurotransmitter release in the brain. Neuroscience 12, 347.

Clapham, D. E. and E. J. Neer (1993). New roles for G-protein

dimers in transmembrane signaling. Nature 365, 403.

deMendonca, A., A. M. Sebastiao, and J. A. Ribeiro (2000).

Adenosine: does it have a neuroprotective role after all. Brain
Res. Rev. 33, 258.
Dennis, E. A., S. G. Rhee, M. M. Billah, and Y. A. Hannun
(1991). Role of phospholipases in generating lipid second
messengers in signal transduction. FASEB J. 5, 2068.

Dunwiddie, T.V. and S. A. Masino (2001). The role and

regulation of adenosine in the central nervous system. Annu.
Rev. Neurosci. 24, 31.

Fisher, S. K., A. M. Heacock, and B. W. Agranoff (1992).

Inositol lipids and signal transduction in the nervous system.
an update. J. Neurochem. 58, 18.

Furchgott, R. F. and J. V. Zawadzki (1980). The obligatory role

of endothelial cells in the relaxation of arterial smooth muscle
by acetylcholine. Nature 288, 373.

Garthwaite, J. (1991). Glutamate, nitric oxide, and cell-cell

signaling in the nervous system. Trends Neurosci. 14, 60.

Halushka, P. V., D. E. Mais, P. R. Mayeux, and T. A. Morinelli

(1989). Thromboxane, prostaglandin, and leukotriene
receptors. Annu. Rev. Pharmacol. Toxicol. 29, 213.

Hokin, L. E. (1985). Receptors and phosphoinositide-

generated second messengers. Annu. Rev. Biochem. 54, 205.

Kerwin, J. R., Jr. and M. Heller (1994). The arginine-nitric

oxide pathway: a target for new drugs. Med. Res. Rev. 14,
23.

Kruszka, K. K. and R. W. Gross (1994). The ATP- and CoA-

independent synthesis of arachidonoylethanolamide. J. Biol.
Chem. 269, 14345.

Levitan, I. B. (1994). Modulation of ion channels by protein

phosphorylation and dephosphorylation. Annu. Rev. Physiol.
56, 193.
Levitan, I. B. and L. K. Kaczmarek (2002). The Neuron: Cell
and Molecular Biology, 2nd ed. Oxford University Press, New
York.

Maines, M. D., J. A. Mark, and J. F. Ewing (1993). Heme

oxygenase, a likely regulator of cGMP production in the brain:
induction in vivo of HO-1 compensates for depression in NO
synthase activity. Mol. Cell. Neurosci. 4, 398.

Mechoulam, R., L. Hanus, and B. R. Martin (1994). Search for

endogenous ligands of the cannabinoid receptor. Biochem.
Pharmacol. 48, 1537.

Moncada, S. and A. Higgs (1993). The L-arginine-nitric oxide

pathway. N. Engl. J. Med. 329, 2002.

Nestler, E. J. and P. Greengard (1999). Serine and threonine

phosphorylation. In Basic Neurochemistry: Molecular, Cellular
and Medical Aspects, 6th ed. (G. J. Siegel, et al., eds.). Raven
Press, New York. Chapter 24, p. 471.

Nestler, E. J., S. E. Hyman and R. C. Malenka (2001).

Molecular Neuropharmacology. McGraw-Hill, New York.

Nicoll, R. A. (1988). The coupling of neurotransmitter

receptors to ion channels in the brain. Science 241, 545.

Piomelli, D. (1994). Eicosanoids in synaptic transmission. Crit.

Rev. Neurobiol. 8, 65.

Piomelli, D., E. Shapiro, R. Zipkin, J. H. Schwartz, and S. J.

Feinmark (1989). Formation and action of 8-hydroxy-ii, 12-
epoxy-5,9,14-icosatrienoic acid in Aplysia: a possible second
messenger. Proc. Natl. Acad. Sci. U.S.A. 86, 1721.

Popoli, M., N. Brunello, J. Perez, and G. Racagni (2000).

Second messenger-regulated protein kinases in the brain:
their functional role and the action of antidepressant drugs. J.
Neurochem. 74, 21.

Rottingen, J. A. and J. G. Iversen (2000). Ruled by waves?

Intracellular and intercellular calcium signalling. Acta Physiol.
Scand. 169, 203.

Schaad, N. C., P. J. Magistretti, and M. Schorderet (1991).

Prostanoids and their role in cell-cell interactions in the
central nervous system. Neurochem. Int. 18, 303.

Schimizu, T. and L. S. Wolfe (1990). Arachidonic cascade and

signal transduction. J. Neurochem. 55, 1.

Schuman, E. M. and D. V. Madison (1994). Nitric oxide and

synaptic function. Annu. Rev. Neurosci. 17, 153.

Sessa, W. C. (1994). The nitric oxide synthase family of

proteins. J. Vasc. Res. 31, 131.

Snider, S. H., S. R. Jaffrey, and R. Zakhary (1998). Nitric

oxide and carbon monoxide: parallel roles as neural
messengers. Brain Res. Rev. 28, 167.

Starke, K., M. Gothert, and H. Kilbinger (1989). Modulation of

neurotransmitter release by presynaptic autoreceptors.
Physiol. Rev. 69, 864.

Vogel, Z., J. Barg, R. Levy, D. Saya, E. Heldman, and R.

Mechoulam (1993). Anandamide, a brain endogenous
compound, interacts specifically with cannabinoid receptors
and inhibits adenylate cyclase. J. Neurochem. 61, 352.

Weng, G., V. S. Bhalla, and R. Inengar (1999). Complexity in

biological signaling systems. Science 784, 92.
Williams, J. H. and T. V. Bliss (1989). An invitro study of the
effect of lipoxygenase and cyclooxygenase inhibitors of
arachidonic acid on the induction and maintenance of long-
term potentiation in the hippocampus. Neurosci. Lett. 107,
301.

Zhang, J. and S. H. Snyder (1995). Nitric oxide in the nervous

system. Annu. Rev. Pharmacol. Toxicol. 35, 213.

(+/-) Show / Hide Bibliography

Send Feedback Teton Server (5.0.0) - ©2007 Teton Data Title Updates
Customer Service Systems User
800.901.5494 Send Us Your Comments Responsibilities
Training Center
What's New
6. Amino Acid Transmitters
INTRODUCTION
On the basis of their functional actions, amino acid
transmitters have been divided into two general categories:
excitatory amino acid transmitters (glutamate [Glu],
aspartate [Asp], cysteate, and homocysteate), which
depolarize neurons in the mammalian central nervous system
(CNS), and inhibitory amino acid transmitters ( -aminobutyric
acid [GABA], glycine [Gly], taurine, and -alanine), which
hyperpolarize mammalian neurons. A few amino acids have
been demonstrated to fulfill most of the criteria for
neurotransmitter candidates in the mammalian CNS. Among
them are GABA, the major inhibitory transmitter in the brain;
Glu, the major excitatory transmitter in the brain; and Gly,
another important inhibitory transmitter in the brain stem and
spinal cord. This has not been an easy task since many amino
acids are also involved in intermediary metabolism and
obviously in protein synthesis, which makes it difficult to
separate their biochemical role from their transmitter role.
Agmatine, the decarboxylation product of arginine, has
fulfilled many of these transmitter criteria and is the most
recent addition to this family of transmitters. From a
quantitative standpoint, the amino acids are probably the
major transmitters in the mammalian CNS, while the better-
known transmitters discussed in other chapters
(acetylcholine, norepinephrine, dopamine, histamine, and 5-
hydroxytryptamine) probably account for transmission at only
a small percentage of central synaptic sites.

GABA

Neurotransmitter Role in the Mammalian

Central Nervous System
GABA was identified as the neurotransmitter for an inhibitory
neuromuscular junction in the walking leg of the lobster. Since
its discovery over 50 years ago, numerous biochemical and
neurophysiological observations have been made about brain
GABA and GABA systems that make a strong case for its
neurotransmitter role in mammalian brain.

Like other neurotransmitters or neurotransmitter candidates,

GABA and its biosynthetic enzyme glutamic acid
decarboxylase (GAD) have a discrete, nonuniform distribution
in the brain. The brain contains a high-affinity, sodium-
dependent transport system that serves to terminate GABA
action. Storage of GABA can be demonstrated in selected
synaptosomal populations, and a vesicular GABA transporter
has been cloned and sequenced. Release of endogenous or
radioactively labeled, exogenously accumulated GABA can be
evoked by the appropriate experimental conditions. The
presence of GABA-containing neurons has been verified, and
the anatomical distribution of GABAergic neurons has been
mapped out, using in situ hybridization for GAD mRNA and
immunocytochemical detection of GAD. However, the most
compelling evidence that GABA plays a neurotransmitter role
in mammalian brain has emerged from intracellular recording
studies, which show that GABA causes a hyperpolarization of
neurons similar to that evoked by the naturally occurring
transmitter substance and that these inhibitory actions of
synthetic GABA and of GABA-containing pathways can be
antagonized by drugs selective for the GABA receptor, such as
bicuculline.

Distribution
Synthesized in 1883, GABA was known for many years as a
product of microbial and plant metabolism. Not until 1950,
however, did investigators identify GABA as a normal
constituent of the mammalian CNS. Moreover, no other
mammalian tissue, with the exception of the retina, contains
more than a mere trace of this material. Obviously, a
substance with such an unusual enrichment in the brain must
have some specific physiological effects that would make it
important for the function of the CNS. Much evidence has now
accumulated in support of the hypothesis that the major
share of GABA found in the brain functions as an inhibitory
transmitter. The probability that GABA functions as an
inhibitory transmitter in the brain has spurred a prodigious
research effort to implicate GABA in the etiology of a host of
neurological and psychiatric disorders. Although the present
evidence is not overwhelming, GABA has been most
convincingly implicated, both directly and indirectly, in the
pathogenesis of epilepsy.

In mammals, GABA is found in high concentrations in the

brain and spinal cord but is absent or present only in trace
amounts in peripheral nerve tissue, such as sciatic nerve,
splenic nerve, and sympathetic ganglia, or in any other
peripheral tissue, such as liver, spleen, and heart. These
findings give some idea of the uniqueness of the occurrence
of GABA in the mammalian CNS. Like the monoamines, GABA
appears to have a discrete distribution within the CNS.
However, unlike the monoamines, the concentration of GABA
found in the CNS is on the order of millimoles per gram rather
than nanomoles per gram. The brain also contains large
amounts of glutamic acid (8-13 mmole/g), which is the main
precursor of GABA and itself a neurotransmitter candidate
(see Glutamic Acid).

Since GABA does not easily penetrate the blood-brain barrier,

brain concentrations of GABA cannot be increased by systemic
administration unless one opens the blood-brain barrier.
GABA-lactam (2-pyrrolidinone), a less polar and more lipid-
soluble compound, can reach the CNS but is not significantly
hydrolyzed to GABA. A more successful approach has been
use of the drug progabide, which not only penetrates the
blood-brain barrier but is subsequently metabolized into
GABA.

Metabolism
Three primary enzymes are involved in the catabolism of
GABA before its final metabolite, succinate, enters the Krebs
cycle. The relative activity of enzymes involved in the
degradation of GABA suggests that, as with monoamines,
they play only a minor role in the termination of the action of
any neurally released GABA.

Figure 6-1 outlines the metabolism of GABA and its

relationship to the Krebs cycle and carbohydrate metabolism.
As mentioned previously, GABA is formed by the -
decarboxylation of L-glutamic acid, a reaction catalyzed by
GAD, an enzyme that occurs uniquely in the mammalian CNS
and retinal tissue. The precursor of GABA, L-glutamic acid,
can be formed from -oxoglutarate by transamination or
reaction with ammonia. GABA is intimately related to the
oxidative metabolism of carbohydrates in the CNS by means
of a "shunt," involving its production from glutamate, its
transamination with -oxoglutarate by GABA- -oxoglutarate
transaminase (GABA-T) yielding succinic semialdehyde and
regenerating Glu, and finally its entry into the Krebs cycle as
succinic acid via the oxidation of succinic semialdehyde by
succinic semialdehyde dehydrogenase. In essence, then, the
shunt bypasses the normal oxidative metabolism involving the
enzymes -oxoglutarate dehydrogenase and succinyl
thiokinase.

From a metabolic standpoint, the significance of the shunt is

unknown; energetically, at least, it is less efficient than direct
oxidation through the Krebs cycle (3 adenosine triphosphate
[ATP] equivalents versus 3 ATP + 1 guanosine triphosphate
[GTP] for the Krebs cycle). The most comon precursor for
GABA formation is glucose, although pyruvate can also act as
a precursor.

Figure 6-1. Interrelationship between -

aminobutyric acid and carbohydrate metabolism.

Glutamic Acid Decarboxylase

GAD is the only synthetic enzyme responsible for the
conversion of L-glutamic acid to GABA, and the reaction is
irreversible. In mammals, this relatively specific
decarboxylase is found primarily in the CNS, where it occurs
in higher concentrations in the gray matter. In general, the
localization of this enzyme in mammalian brain correlates
quite well with the GABA content. The brain enzyme has been
purified to homogeneity and its properties studied in detail. It
has a pH optimum of about 6.5 and requires pyridoxal
phosphate, a form of vitamin B6 as a coenzyme. This purified
enzyme is inhibited by structural analogs of Glu, carbonyl
(pyridoxal phosphate [PLP])-trapping agents, sulfhydryl
reagents, thiol compounds, and anions such as chloride. Two
isoforms of GAD have been identified, which are encoded by
two distinct genes. These two isoforms, designated GAD65
and GAD67 in accordance with their molecular weights, differ
in animo acid sequence, antigenicity, cellular and subcellular
location, and interaction with the GAD cofactor PLP. Their
different intracellular distributions suggest that the two GAD
forms may be regulated in different ways. GAD65 and GAD67
differ significantly in their affinity for the pyridoxal cofactor:
GAD65 shows a relatively high affinity for the cofactor,
whereas the larger GAD isoform does not. The affinity of
GAD65 for the cofactor results in the ability of GAD65 enzyme
activity to be efficiently and quickly regulated. In contrast, the
activity of GAD67 is determined through induction of new
enzyme protein rather than through posttranslational
mechanisms.

GABA-Transaminase
GABA-T, unlike the decarboxylase, has a wide tissue
distribution. Therefore, although GABA cannot be formed to
any extent outside the CNS, exogenous GABA can be rapidly
metabolized by both central and peripheral tissue. However,
since endogenous GABA is present only in nanomolar
amounts in cerebrospinal fluid, it is unlikely that a significant
amount of endogenous GABA leaves the brain intact. The
brain transaminase has a pH optimum of 8.2 and requires
PLP. It appears that the coenzyme is more tightly bound to
this enzyme than it is to GAD. The brain ratio of GABA-T/GAD
activity is almost always greater than 1. Sulfhydryl reagents
tend to decrease GABA-T activity, suggesting that this
enzyme requires the integrity of one or more sulfhydryl
groups for optimal activity. Transamination of GABA catalyzed
by GABA-T is a reversible reaction, so if a metabolic source of
succinic semialdehyde were made available, it would be
theoretically possible to form GABA by the reversal of this
reaction. However, as indicated below, this does not appear to
be the case in vivo under normal or experimental conditions,
at least those investigated so far.

Recent studies with more sophisticated cell fractionation

techniques and electron microscopic monitoring of the
fractions obtained have borne out the original claims that
both GAD and GABA-T are particulate to some extent. GAD
was associated with the synaptosome fraction, whereas
GABA-T was largely associated with mitochondria. Further
studies on the mitochondrial distribution of GABA-T have
suggested that the mitochondria released from synaptosomes
have less activity than the crude, unpurified mitochondrial
fraction, and it has been inferred that the mitochondria within
nerve endings have little GABA-T activity. This has led to the
speculation that GABA is metabolized largely at extraneuronal
intercellular sites or in the postsynaptic neurons. Gabaculine
is the most potent GABA-T inhibitor available. Similar to -
acetylenic and -vinyl GABA, this agent is a catalytic inhibitor
of GABA-T and, unfortunately, will inhibit GAD. However,
gabaculine has a fair degree of specificity since it is about
1000-fold less effective as a GAD inhibitor than as a GABA-T
inhibitor.

Succinic Semialdehyde Dehydrogenase

Brain succinic semialdehyde dehydrogenase (SSADH) has a
high substrate specificity and can be distinguished from the
nonspecific aldehyde dehydrogenase found in the brain. The
enzyme purified from human brain has a pH optimum of
about 9.2 and a Km for succinic semialdehyde of 5.3 10-6
and a Km for NAD of 3 10-5. SSADH from rat brain has a
similarly low Km for succinic semialdehyde of 7.8 10-5 and
for NAD of 5 10-5. The high activity of this enzyme and the
low Michaelis constant, which allow the enzyme to function
effectively at low substrate concentrations, probably account
for the fact that succinic semialdehyde has not even been
detected as an endogenous metabolite in neural tissue,
despite the active metabolism of GABA in vivo.

Since GABA's rise to popularity, the literature has been

inundated with claims that many pharmacological and
physiological effects can be ascribed to and correlated well
with changes in GABA levels in the brain. Since both GAD and
GABA-T are dependent on the coenzyme PLP, it is not
surprising that pharmacological agents or pathological
conditions affecting this coenzyme can cause alterations in
the GABA content of the brain. Epileptiform seizure can be
produced by a lack of this coenzyme or by its inactivation.
Conditions of this sort also lead to a reduction in GABA levels,
since GAD appears to be preferentially inhibited over the
transaminase, presumably due to the fact that GAD has a
lower affinity for the coenzyme than does GABA-T. A diet
deficient in vitamin B6 in infants can lead to seizures that
respond successfully to treatment consisting of addition of
pyridoxine to the diet. However, many other enzymes,
including some of those involved in the biosynthesis of other
bioactive substances, are also pyridoxal-dependent.

A number of observations indicate that there is no simple

correlation between GABA content and convulsive activity.
Administration to experimental animals of a variety of
hydrazides, such as thiosemicarbozide, has uniformly resulted
in the production of repetitive seizures following a rather
prolonged latent period. The finding that hydrazide-induced
seizures could be prevented by parenteral administration of
various forms of vitamin B6 led to the suggestion that some
enzyme system requiring PLP as a coenzyme was being
inhibited and that the decrease in the activity of this enzyme
was somehow related to the production of the seizures
observed. At this time, attention focused on GABA and GAD
because of their unique occurrence in the CNS and because
GAD had been shown to be inhibited by carbonyl-trapping
agents in vitro. The hydrazide-induced seizures were
accompanied by substantial decreases in GABA levels and
reductions in GAD activity in various areas of the brain.
Subsequently, it was recognized that these decreases were
underestimated by the agonal increases in GABA that
accompany terminal ischemia. However, even preferential
inhibition of GABA-T with carbonyl reagents such as
hydroxylamine (NH2OH) or amino-oxyacetic acid, which
increased GABA levels (up to 500% of control) in the CNS,
gave no protective effect against the hydrazide-induced
convulsions, even though they prevented the depletion of
GABA. Administration of very high doses of only amino-
oxyacetic acid, instead of producing the normally observed
sedation, caused some seizure activity in spite of the
extremely high brain levels of GABA.

Administration of amino-oxyacetic acid to a strain of

genetically spastic mice in a single dose of 5-15 mg/kg
resulted in marked improvement of their symptomatology for
12-24 hr. Although this improvement was associated with an
increase in GABA levels, the GABA level increased with a
similar time course and to the same extent as in normal
control mice, making the improved spasticity hard to attribute
to GABA. All studies to date indicate that the principal genetic
defect in these mice is not in the operation of the GABA
system. Perhaps, the drug-induced increase in GABA quells an
excess of excitatory input in some unknown area of the CNS.

Alternate Metabolic Pathways

In addition to undergoing transamination and subsequently
entering the Krebs cycle, GABA can apparently undergo
various other transformations in the CNS, forming a number
of compounds whose importance, and in some cases natural
occurrence, has not been conclusively established. Figure 6-2
depicts a variety of derivatives for which GABA may serve as
a precursor. Perhaps the simplest of these metabolic
conversions is the reduction of succinic semialdehyde (a
product of GABA transamination) to -hydroxybutyrate (GHB).
The transformation of GABA to GHB has been demonstrated in
rat brain both in vivo and in vitro. Recent studies have
demonstrated that GHB administered in physiologically
relevant concentrations can also be converted to GABA by
transamination. GHB aciduria, a rare inborn error in the
metabolism of GABA, has been reported in children and
appears to be the result of a deficiency of SSADH, the enzyme
that oxidizes succinic semialdehyde to succinic acid (see Fig.
6-1). GHB levels are increased in urine, plasma, and
cerebrospinal fluid; but it is unclear whether the main clinical
symptoms of motor and mental retardation, muscular
hypotonia, and ataxia are related to the elevated levels of
GHB.

GHB has recently achieved notoriety as a "date rape drug."

Ingested in doses as low as 10 mg/kg, GHB or its lactone
precursor -butyrolactone produces euphoria, impairment of
judgment, anxiolysis, and amnestic deficits in short-term
memory. Higher doses lead to unconsciousness, seizures,
respiratory depression, and coma; several deaths have been
attributed to GHB.

Figure 6-2. Possible alternate metabolic pathways

for -aminobutyric acid.

Storage
Like most classical neurotransmitters, GABA is packaged and
stored in vesicles in the presynaptic terminals, from which it
is released into the synaptic cleft. A vesicular transporter that
accumulates GABA has been identified in GABAergic cells.
This transporter was cloned on the basis of homology to unc-
47 in Caenorhabditis elegans, a strategy of moving from the
worm to the mammalian species that has proven to be very
useful for identifying a variety of mammalian transmitter-
related genes. The vesicular GABA transporter (GAT) differs
from the two vesicular monoamine transporters (VMATs, see
Chapters 8, 9, and 10) by having 10 rather than 12
presumptive transmembrane domains and a very large
cytoplasmic N terminus of approximately 130 amino acids.
Like the vesicular Glu transporter, GAT is highly dependent on
the electrical potential across the vesicular membrane and
differs from VMATs in terms of this bioenergetic dependence.
Specific inhibitors of the vesicular inhibitory GAT have not
been identified.

However, the vesicular GAT shares with the VMATs a lack of

substrate specificity and will transport the inhibitory
transmitter Gly as well as GABA. Consistent with this
pharmacology, the vesicular GAT has been found in Gly- as
well as GABA-containing neurons. Accordingly, some have
suggested that it can be more accurately designated a
vesicular inhibitory amino acid transporter. Interestingly, a
limited number of GABAergic neurons appear to lack this
transporter, raising the suspicion that another (related)
transporter may exist in these neurons or that this small
population has an alternate mechanism of storage.

Release and Reuptake

The arrival of an action potential or other depolarizing
stimulus in the presynaptic GABAergic terminal initiates a
sequence of events that ultimately results in vesicular fusion
and release of GABA into the synaptic cleft, as is presumed to
be the case for all other neurotransmitters. After release, the
action of GABA is terminated largely by removal from the
synaptic cleft by the actions of several types of plasma
membrane transporter (see below). The uptake of several
transmitters by both glia and neurons has been reported. This
dual glial-neuronal reuptake is a common property in neurons
using amino acid transmitters, probably because amino acids
can play dual roles as both transmitters and metabolic
intermediates.

The ability of glia to avidly accumulate GABA and other amino

acids distinguishes amino acid transmitters from other
classical transmitters. Reuptake is the primary mode of
inactivation of GABA that is released from neurons. Molecular
cloning techniques have suggested greater heterogeneity in
the GATs than previously suspected, with the genes for four
distinct GATs being detected. At least three specific GAT
proteins are expressed in the CNS. In addition, a betaine
transporter that accumulates GABA has been cloned. All
known GATs are expressed in both neurons and glia. There is
as yet no obvious answer to the question of why are there are
multiple transporters for GABA?. GATs are expressed in both
GABAergic neurons and non-GABAergic cells (presumably
cells that receive GABA innervation) as well as glia. The
presence of multiple transporter proteins for the same
transmitter, localized in neurons as well as glia, differs from
the situation for catecholamine transmitters, in which a single
membrane-associated transporter protein with relatively
selective substrate specificity is found in a neuron defining its
chemical identity. One possibility is that the cloned GATs may
be cotransporters for other amino acids. For example, no
transporters for -alanine and taurine have yet been cloned,
but these amino acids are accumulated by GATs. Finally, it is
possible that one or more of these transporters may have the
ability to function in the outward direction, serving as a
paradoxical mechanism for the release, rather than removal,
of GABA.

The fact that the GATs transporters are not uniquely

concentrated in the plasma membrane of the presynaptic
GABA terminals has important functional ramifications. The
GABA that is taken up in glia or non-GABAergic postsynptic
cells will not be available to recycle for another round of
exocytotic release. This lost GABA will have to be replaced by
de novo synthesis, placing enhanced demands on the
synthetic capacity of the GABA-ergic neuron.

GABA Receptors
In vertebrate species, GABA receptors are found primarily in
nerve cell membranes and are sufficiently widespread that
most neurons in the CNS possess them. However, GABA
receptors are not exclusively associated with neurons. They
are also expressed by astrocytes, where they appear to be
involved in the regulation of chloride channels. Interestingly,
GABA receptors are also found outside the CNS on neurons of
the autonomic nervous system.
In vertebrates, there are two major types of GABA receptor:
the inotropic GABAA receptor and the metabotropic GABAB
receptor. GABA receptors were initially subdivided into these
two groups based on pharmacological evidence. However, the
functional separation also extends to second-messenger
mechanisms, differences in the location of these receptor
subtypes in the mammalian CNS, and their molecular
composition. In addition, both receptor subtypes have pre-
and postsynaptic locations and are thought to participate
independently in synaptic transmission.

Autoreceptor Regulation of GABA Release

Pharmacological studies indicate that autoreceptor-mediated
regulation of GABA neurons takes place predominantly
through GABAB receptors located on GABAergic nerve
terminals (see GABAB Receptor). Immunohistochemical
studies have revealed that both GABAB and GABAA receptors
are present on postsynaptic non-GABAergic neurons.
Presumably, these GABAA postsynaptic receptors respond to
GABA released from a presynaptic GABA-ergic neuron. An
anatomical arrangement of one GABA neuron terminating on
another GABA cell would have the same functional
consequence as an autoreceptor (decreasing subsequent
transmitter release), making it difficult to distinguish between
true autoreceptor and heteroceptor regulation of GABA
release.

GABAA Receptor

GABAA receptors are the major inhibitory neurotransmitter

receptors in the brain and the site of action of many clinically
important drugs (Fig. 6-3). These receptors are believed to be
involved in mediating anxiolytic, sedative, anticonvulsant,
muscle-relaxant, and amnesic activity.

The inotropic GABAA receptor is by far the most prevalent of

the two known GABA receptor types in mammalian CNS and
has been extensively studied and characterized. Like the
nicotinic acetylcholine receptor (nAChR), the GABAA receptor
is composed of four subunits comprising an integral
transmembrane ion channel that is gated by the binding of
two agonist molecules. However, unlike the nAChR, the
receptor-associated GABA channel predominantly conducts
chloride ions. Since the Cl- equilibrium potential is near the
resting potential in most neurons, increasing chloride
permeability hyperpolarizes the neuron and thereby
decreases the depolarizing effects of an excitatory input, thus
depressing excitability.

The GABAA receptor, a multisubunit receptor-channel

complex, can be allosterically modulated by two important
classes of drugs: the benzodiazepines and the barbiturates.
The primary structure of the GABAA receptor, described in
1987, revealed that it is a member of a large superfamily of
ligand-gated ion channels that includes the nicotinic-
cholinergic, inotropic Glu, Gly, and 5-hydroxytryptamine3
(5HT3) receptors. The GABAA receptor-ion channel complex is
believed to be a heteropentameric glycoprotein of
approximately 275 kDa composed of a combination of
multiple polypeptide subunits (cf. Fig. 6-4). Seven distinct
classes of polypeptide subunits ( , , , , , , and ) have
been cloned, and multiple isoforms of each have been shown
to exist so that the total number of identified subunits now
stands at 18. The existence of a large family of genes coding
for diverse subunits ( 1-6, 1-4, 1-3, , , , 1-2) provides
the basis for the extraordinary structural diversity of GABAA
receptors.
The subunit composition of the GABAA receptors appears to
vary from one brain region to another and even between
neurons in a given region, but the exact composition of most
native GABAA receptors is unknown. The distribution of mRNA
in the CNS determined by in situ hybridization is very
different for each subunit subtype. A recently cloned subunit
( 6), which confers a unique pharmacology (binding of the
partial inverse agonist RO-15-4513, a putative ethanol
antagonist) to a recombinantly expressed GABAA receptor, is
expressed in only a single type of neuron, the cerebellar
granule neuron. Thus, it is now becoming clear that the
heterogeneity of the GABAA receptor subunit isoforms confers
a diversity of pharmacological and perhaps physiological
response characteristics upon the GABAA receptor. For
example, coexpression of an additional subunit is necessary
for the potentiation of GABA responses by benzodiazepines.
In addition, coexpression of individual -subunit variants ( 1
2, or 3) with and subunits results in varying degrees of
modulation by benzodiazepine receptor ligands (agonist,
antagonist, inverse agonist). Photoaffinity-labeling studies
have further suggested that the benzodiazepine-binding site
resides on the subunit, while the GABA-binding site itself
resides on the subunit. Finally, it appears that the -subunit
heterogeneity determines the diversity of physiological and
pharmacological response characteristics of native GABAA
receptors, even though expression of the subunits is
essential for conferring the modulatory actions of
benzodiazepines on GABAA receptors. Thus, when
coexpressed with 1, the 1 subunit yields a receptor with a
relatively high affinity for GABA. By contrast, coexpression of
the 2 or 3 subunit with the 1 subunit results in GABAA
receptors with far lower affinity for GABA. Thus, the subunit
composition of a given receptor may actually determine the
local response of the GABAA receptor to synaptically released
GABA. These subtle differences in subunit organization may
result in subpopulations of GABAA receptors that have
different regional and cellular locations, each with differential
sensitivity to GABA and allosteric modulators.

This extraordinary heterogeneity of GABAA receptors clearly

provides a hitherto unexplored diversity in the function of
receptor subtypes affecting their sensitivity to GABA,
modulation by allosteric effectors, adaptation to stimulus
conditions, distribution within a neuron and between neurons,
ontogenetic development, and alterations in pathological
states. An exciting new pharmacology is emerging from the
recognition of the functional relevance of GABAA receptor
subtypes, providing a rational basis for the development of
subtype-specific ligands Several studies have suggested that
phosphorylation of the GABAA receptor channels may also be
of importance for both short-term and long-term regulation of
GABAA receptor function and expression. At present,
however, the physiological significance and specific
consequences of the phosphorylation of GABAA receptor
channels are unknown.

An increased understanding of the benzodiazepine GABA

receptor chloride channel complex has led to the development
of selective anxiolytic and anticonvulsant agents that lack
significant sedative and muscle-relaxant action, properties
that often limit the usefulness of traditional agents such as
benzodiazepines and barbiturates. A better understanding of
the molecular characteristics and regulation of the multiple
allosteric sites of the supramolecular complex and the
endogenous substances that may physiologically subserve
these sites should not only contribute to our understanding of
the possible etiology of anxiety and seizure disorders but also
aid in the development of more effective and specific
therapeutic agents. Once the functional properties of the
GABAA subunits and their subtypes are more clearly defined,
it should be possible to use this knowledge in the rational
screening and/or design of new, clinically useful subtype-
specific agents. The generation of animal models in which
particular GABAA receptor subunits are either inactivated
(knockout strategy) or selectively point-mutated (knockin
strategy) should help define the functional properties of
GABAA subunits and their subtypes. These animals will also
accelerate the recognition of the role of these receptor
subtypes as potential drug targets.

Figure 6-3. Schematic illustration of the GABAA

receptor complex and the sites of action of
different agents on the receptor. BDZ,
benzodiazepine.

Figure 6-4. Schematic illustration of the GABAA

receptor structure containing two and subunits
and a single subunit to form an intrinsic Cl- ion
channel. Putative ligands and drugs known to
interact at one of the major sites associated with
the GABAA receptors and to either positively or
negatively modulate GABA-gated Cl- ion
conductance are also illustrated. TBPS, t-
butylbicyclophosphothianate; DBI, diazepam-
binding inhibitor. (From Paul, 1995.)

GABAB Receptor

GABAB receptors belong to the superfamily of G protein-

coupled receptors and are classified as metabotropic
receptors. Their ligand-binding domain is not directly
associated with their ion channel effector. The GABAB
receptor, as mentioned above, is present at lower levels in the
CNS than the GABAA receptor and is not linked to a chloride
channel.

Since its pharmacological discovery in 1980, much progress

has been made. Selective agonists and antagonists have been
developed, and a functional role for this receptor as a
mediator of slow inhibitory postsynaptic potentials in many
brain regions has emerged. GABAB receptor activation also
plays a role in attenuating the release of biogenic amines,
acetylcholine, excitatory amino acids, neuropeptides,
hormones, as well as GABA via an interaction with
autoreceptors. Whereas GABAA receptors are directly
associated with a Cl- channel, GABAB receptors seem to be
coupled to Ca2+ or K+ channels via second-messenger
systems. The inhibitory hyperpolarizing action of GABAB
receptor activation appears to be mediated through either
increases in potassium conductance or decreases in calcium
conductance.

Molecular cloning studies have revealed that GABAB

receptors, like the metabotropic glutamate receptors
(mGluRs, see EAA Receptors), are members of the G protein-
coupled receptor superfamily and contain seven presumptive
transmembrane domains. Two major GABAB subunits have
been cloned (GABABR1a and -R1b) and a novel GABAB
receptor subunit has been identified (GABABR2). These Gi-
coupled GABAB receptors are larger than most G protein-
coupled receptors, being comprised of 850-960 amino acids.
The GABAB and mGluRs receptors can be distinguished from
most other G protein-coupled receptors by their large
exracellular N-terminal domains.

GABAB receptors are expressed on both pre- and postsynaptic

membranes, where, as mentioned earlier, they decrease Ca2+
conductance, open K+ channels, and inhibit adenylyl cyclase.
In contrast to GABAA receptors, postsynaptic GABAB
receptors elicit a slower, longer-lasting form of inhibition, an
effect that is attributed to the opening of inwardly rectifying
K+ channels. The GABAB receptor can be distinguished
pharmacologically from the GABAA receptor by its selective
affinity for the agonist baclofen and its lack of affinity for
muscimol and bicuculline (see Table 6-1). The GABAB
receptor is believed to be linked through GTP-sensitive
proteins to a calcium channel. Activation of the GABAB
presynaptic receptors by baclofen decreases calcium
conductance and transmitter release. Postsynaptic GABAB
receptors are indirectly coupled to K+ channels via G
proteins, and they mediate late inhibitory postsynaptic
potentials. Unlike the GABAA receptor, the GABAB receptor is
not modulated by the benzodiazepines or barbiturates.
Pharmacological studies have demonstrated that blockade of
GABAB receptors produces none of the profound behavioral
sequelae observed following administration of GABAA
antagonist (e.g., seizures). These observations suggest that,
unlike GABAA receptors, which are believed to be in a
continuous tonically activated state, GABAB receptors may be
activated only under certain physiological conditions.

With regard to the functions of the GABAB receptor in the

brain, it seems premature to assign a physiological or
pathological role. However, the discovery of selective GABAB
antagonists that cross the blood-brain barrier has aided in
evaluating the functions of this receptor (see Fig. 6-5). With
the development of a potent, orally effective GABAB
antagonist, CGP-54626, it became possible to evaluate better
the physiological role of this receptor. In vivo and in vitro
studies clearly demonstrated that blockade of GABAB
receptors with this agent increased neurotransmitter (GABA
and Glu) release, reduced late inhibitory postsynaptic
potentials of CA1 hippocampal pyramidal neurons, and led to
an increase in neuronal excitability. Behavioral studies in
several species have suggested that GABAB receptor blockade
can improve cognition in rats (social learning), mice (passive
avoidance), and rhesus monkeys (conditional spatial color
test). However, baclofen is the only drug in clinical use that
interacts with GABAB receptors. This drug is used as a
muscle-relaxant to decrease spasticity in a diversity of
neurological disorders.

Figure 6-5. Chemical structure of GABAB receptor

antagonists.

Pharmacology of GABAergic Neurons

Drugs can influence GABAergic function by interacting at
many different sites, both pre- and postsynaptic (Fig. 6-6).
Drugs can influence presynaptic events and modify the
amount of GABA that ultimately reaches and interacts with
postsynaptic GABA receptors. In most cases, presynaptic drug
effects do not involve an interaction with GABA receptors. The
most extensively studied presynaptic drug actions involved
inhibitory effects exerted on enzymes involved in GABA
synthesis (GAD) and degradation (GABA-T) and the neuronal
reuptake of GABA. The major exception is the interaction of
drugs with GABAergic autoreceptors to modulate both the
physiological activity of GABA neurons and the release and
synthesis of GABA in a manner analogous to the role played
by dopamine autoreceptors in the regulation of dopaminergic
function.
A great deal of emphasis has been directed recently to the
study of drug interactions with GABA receptors. Drugs
interacting at the level of GABA receptors can be classified
into two general categories: GABA antagonists and GABA
agonists. Figure 6-6 depicts possible sites of drug interaction
in a hypothetical GABAergic synapse. (The structures of
compounds that act at GABAergic synapses are depicted in
Figs. 6-7 and 6-8.) Picrotoxinin is the active component of
picrotoxin.

Figure 6-6. Schematic illustration of a GABAergic

neuron indicating possible sites of drug action.
Site 1: Enzymatic synthesis. Glutamic acid
decarboxylase (GAD-1) is inhibited by a number of
various hydrazines. These agents appear to act
primarily as pyridoxal antagonists and are
therefore very nonspecific inhibitors. L-glutamate-
-hydrazide and allylglycine are more selective
inhibitors of GAD-1, but these agents are also not
entirely specific in their effects. Site 2: Release.
GABA release appears to be calcium-dependent.
At present, no selective inhibitors of GABA release
have been found. Site 3: Interaction with
postsynaptic receptor. Bicuculline and picrotoxin
block the action of GABA at postsynaptic
receptors. 3-Aminopropane sulfonic acid and the
hallucinogenic isoxazole derivative muscimol
appear to be effective GABA agonists at
postsynaptic receptors and autoreceptors. THIP,
tetrahydroisoxazolopyridinol. Site 3a: Presynaptic
autoreceptors. Possible involvement in the control
of GABA release. Site 4: Reuptake. In the brain,
GABA appears to be actively taken up into
presynaptic endings by a sodium-dependent
mechanism. A number of compounds will inhibit
this uptake mechanism, such as 4-methyl-GABA
 and 2-hydroxy-GABA, but these agents are not
completely specific in their inhibitory effects. Site
5: Metabolism. GABA is metabolized primarily by
transamination by GABA-transaminase (GABA-T),
which appears to be localized primarily in
mitochondria. Amino-oxyacetic acid, gabaculline,
and acetylenic GABA are effective inhibitors of
GABA-T.

Figure 6-7. Structures of compounds that act at

GABAergic synapses.

Figure 6-8. Structures of compounds that act at

GABAergic synapses.

GABA Antagonists
The action of GABA at the receptor-ionophore complex may
be antagonized by GABA antagonists either directly, by
competition with GABA for its receptor, or indirectly, by
modification of the receptor or inhibition of the GABA-
activated ionophore. The two classic GABA antagonists (Fig.
6-6) bicuculline and picrotoxin appear to act by different
means. Bicuculline acts as a direct competitive antagonist of
GABA at the receptor level, while picrotoxin acts as a
noncompetitive antagonist, presumably due to its ability to
block GABA-activated ionophores. Although early studies
raised some doubts about the usefulness of bicuculline as a
selective GABA antagonist, this skepticism has been largely
resolved and appears to be primarily related to the instability
of bicuculline at 37 C and physiological pH. Under normal
physiological conditions, bicuculline is hydrolyzed to bicucine,
a relatively inactive GABA antagonist with a short half-life of
several minutes. The quaternary salts now used for most
electrophysiological experiments (bicuculline methiodide and
bicuculline methochloride) are much more water-soluble and
stable over a broad pH range of 2-8. However, these
quaternary salts are not suitable for systemic administration
because of their poor penetration into the CNS.

GABA Agonists
Electrophysiological studies have demonstrated a wide variety
of compounds that are capable of directly activating
bicuculline-sensitive GABA receptors. These agonists can be
readily subdivided into two groups based on their ability to
penetrate the blood-brain barrier, dictating whether they will
be active or inactive following systemic administration. Agents
such as 3-aminopropane-sulfonic acid, -guanidinoproprionic
acid, 4-aminotetrolic acid, trans-4-aminocrotonic acid, and
trans-3-aminocyclopentane-1-carboxylic acid are effective
direct-acting GABA agonists. However, entry of these agents
into the brain following systemic administration is minimal. In
addition, compounds such as trans-4-aminotetrolic acid and
4-aminocrotonic acid also inhibit GABA-T and GABA uptake;
therefore, their action is not totally attributable to their direct
agonist properties.

In contrast to this class of direct-acting GABA agonists, other

GABA agonists readily pass the blood-brain barrier and are
active following systemic administration. Muscimol (3-
hydroxy-5-aminomethylisoxazole) is the agent in this group
that has been most extensively studied. Some other agents in
this group include (5)-(2)-5-(1-aminoethyl)-3-isoxazole,
tetrahydroisoxazolopyridinol (THIP, a bicyclic muscimol
analog), SL-76002 ( [chloro-4-phenyl]fluro-5-hydroxy-2
benzilide-neamino-4H butyramide), and kojic amine (2-
aminomethyl-3-hydroxy-4H-pyran-4-one). In addition,
GABAergic substances may be further categorized as direct or
indirect GABA receptor activators. For example, muscimol,
isoguvacine, and THIP are GABA mimetic agents that interact
directly with GABA receptors. Indirectly acting GABA mimetics
facilitate GABAergic transmission by increasing the amount of
endogenous GABA that reaches the receptor or by altering in
some manner the coupling of the GABA receptor-mediated
change in chloride permeability. Thus, many drugs often
classified as indirect GABA agonists act presynaptically to
modify GABA release and metabolism rather than by
interacting directly with GABA receptors. For this reason,
drugs like gabaculine (a GABA-T inhibitor), nipecotic acid (a
GABA uptake inhibitor), and baclofen (an agent that, in
addition to many other actions, causes release of GABA from
intracellular stores) are often classified incorrectly as GABA
agonists. The benzodiazepines mentioned earlier also appear
to potentiate the action of tonically released GABA at the
receptor by displacement of an endogenous inhibitor of GABA
receptor binding, allowing more endogenous GABA to reach
and bind receptors. Thus, benzodiazepines are sometimes
also classified as GABA agonists. A GABA-like action can also
be elicited by agents that bypass GABA receptors and
influence GABA ionophores. It has been suggested that
pentobarbital acts at the level of the GABA ionophore, but it is
unclear whether its CNS-depressant effects are explained by
this action.

The structures of some of the more potent and widely used

direct-acting GABA agonists are illustrated in Figures 6-7 and
6-8. Included are muscimol, isoguvacine, THIP, and (1)-trans-
3-aminocyclopentane carboxylic acid. Useful therapeutic
effects have not yet been obtained by use of agents of this
sort, which have direct GABA mimetic effects (e.g.,
muscimol), inhibit the active reuptake of GABA (e.g., nipecotic
acid), or alter the rate of synthesis or degradation of GABA
(e.g., amino-oxyacetic acid and gabaculine). However, useful
therapeutic effects are achieved with the anxiolytic
benzodiazepines (e.g., diazepam [Valium] and
chlordiazepoxide [Librium]), which may exert their actions by
facilitating GABAergic transmission.

If the anatomical distribution and functional properties of

GABAA receptor subtypes and subunits become clearly
defined, this knowledge may enable the development of
therapeutically useful subtype-specific agonists that can be
directed to modify GABAergic function in selective brain areas.

Endogenous Modulators
The large number of recognition sites associated with GABAA
receptors has led to speculation that a host of endogenous
regulatory factors exists. A number of candidates have been
identified, but with the exception of the neurosteroids and the
endogenous diazepam-binding inhibitor (DBI), there is little
compelling evidence that any play an important role in
modulating GABAA receptor function in vivo. DBI is an
endogenous peptide that has been purified to homogeneity
from rat and human brain and its structure determined by
recombinant DNA technology. Several lines of evidence
suggest that DBI functions as a precursor of a family of
allosteric modulatory peptides of the GABAA receptor, causing
negative modulation of the GABA-operated Cl- ion fluxes (see
Chapter 12). Consistent with this action, DBI appears to have
anxiogenic properties similar to those associated with inverse
benzodiazepine agonists in experimental animals. A similar
endogenous peptide has been found in human brain and
cerebrospinal fluid. The amount of human DBI
immunoreactivity is elevated in the cerebrospinal fluid of
severely depressed patients.

The other postulated endogenous ligands for the GABAA

receptor include two naturally occuring reduced steroid
metabolites of deoxycorticosterone and progesterone
(allotetrahydro-deoxycorticosterone (DOC) and
allopregnanolone, respectively). These neurosteroids are
formed in the brain and bind with high affinity to GABAA
receptors, eliciting a barbiturate-like action that potentiates
GABA-elicited Cl- conductance. These are among the most
potent known endogenous ligands of GABAA receptors found
in the CNS. Their plasma and brain (cortex and
hypothalamus) levels are increased dramatically in rats
following exposure to stress. Plasma levels of
allopregnanolone are also elevated during the third trimester
of pregnancy and decrease dramatically following parturition.
The observations that brain and plasma levels of
allopregnanolone and allotetrahydro-DOC increase rapidly in
the brain after stress (4- to 20-fold in less than 5 minutes)
suggest that these neurosteroids may have a physiological
role in stress and anxiety. In addition, conditions that may
lead to large increases in neuroactive steroid levels, such as
puberty, pregnancy, or the menstrual cycle, could also alter
neurochemical and behavioral adaptations to stress. To date,
however, none of these putative natural ligands
(neurosteroids or DBIs) have been unequivocally
demonstrated to subserve a physiological action or to
modulate a pathological state.

GLYCINE

As an Inhibitory Transmitter
Structurally, Gly is the simplest amino acid. It is found in all
mammalian body fluids and tissue proteins in substantial
amounts. Although Gly is not an essential amino acid, it is an
essential intermediate in the metabolism of protein, peptides,
one-carbon fragments, nucleic acids, porphyrins, and bile
salts. It is also considered to be an established inhibitory
neurotransmitter, enriched in the medulla, spinal cord, and
retina. Thus, Gly appears to have a more circumscribed
function in the CNS than the more ubiquitously distributed
GABA. As with the other major amino acid transmitters,
numerous neurochemical studies have attempted to separate
and distinguish between the general metabolic and
transmitter functions of Gly within the CNS. Gly also appears
to be an exclusively vertebrate transmitter, making it unique
among the transmitter substances.

Glycinergic neurons appear to respond to activation as other

chemically defined neurons do. Arrival of an action potential
in the presynaptic nerve terminal initiates a calcium-
dependent cascade of events, which ultimately involves fusion
of the presynaptic membrane and release of Gly into the
synaptic cleft. Gly is removed from the synaptic cleft by
uptake transporters located on glial cells and on the
presynaptic terminals of the glycinergic neurons. However, in
the last several years, very little progress has been achieved
in developing pharmacological tools that act selectively on Gly
systems or in generating more information concerning Gly
metabolism in neuronal tissue.

In the spinal cord and brain stem, specific uptake of Gly has
been demonstrated in regions exhibiting high densities of
inhibitory Gly receptors. Two Gly transporter proteins have
been cloned and shown to be expressed in brain as well as in
peripheral tissues. Both are members of the large family of
Na+/Cl--dependent neurotransmitter transporters (see
Chapter 10) and share approximately 50% sequence identity
with the GABA transporter. The Gly transporters have been
named GLYT-1 and GLYT-2, in the order in which they were
reported. These transporters have very similar kinetics and
pharmacological properties but differ in the distribution of
their transcripts, measured by in situ hybridization. The
distribution of GLYT-1 mRNA closely parallels the distribution
of the Gly receptor, suggesting that GLYT-1 is primarily a glial
transporter and GLYT-2 is associated primarily with neurons.
GLYT-1 exist in three isoforms, which are probably generated
by alternate splicing. These isoforms do not exhibit any
known variation in their uptake properties but do possess
distinct patterns of expression in the CNS. GLYT-1 is
expressed in both astrocytes and neurons, whereas GLYT-2 is
localized on axons and the terminal boutons of neurons that
contain vesicular Gly. The GLYT-1 isoforms can be
distinguished pharmacologically from GLYT-2 since they are
sensitive to the effects of sarcosine (N-methylglycine).

Both GLYT-1 and GLYT-2 are expressed in the brain stem and
spinal cord, a location consistent with their role in terminating
glycinergic transmission. However, GLYT-1 is also expressed in
several regions of the forebrain that are devoid of glycinergic
neurotransmission. Thus, GLYT-1 may regulate N-methyl-D-
aspartate (NMDA) Glu receptor function in these areas by
controlling the levels of extracellular Gly available to
allosterically modulate the activity of these receptors (see Gly
as modulator or NMDA receptor). If that is the case, then
GLYT-1 inhibitors may prove to be useful clinically to augment
NMDA receptor function.

The strychnine-sensitive Gly receptor has also been cloned

and expressed and appears to be structurally quite
homologous to the multimeric subunits of other ligand-gated
ion channels, including the GABAA receptor. The native Gly
receptor appears to be a pentameric structure, and
photoaffinity labeling reveals that both Gly- and strychnine-
binding sites are located on the subunit. Several Gly
receptor -subunit variants have been identified ( 1-4) and
shown to differ in their pharmacological properties and levels
of expression. Expression of 1 and 2 subunits is
developmentally regulated, with a switch from the neonatal
2 subunit (strychnine-insensitive) to the adult 1 form
(strychnine-sensitive) at about 2 weeks postnatally in the
mouse. It is interesting that the timing of this switch
corresponds with the development of spasticity in the mutant
spastic mouse, prompting the speculation that insufficient
expression of the adult strychnine-sensitive isoform may
underlie some forms of spasticity.

To date, very little is known about the factors controlling the

release of Gly from the spinal cord. Again, as with GABA, the
efficient uptake process may explain why it is difficult to
detect Gly release from the CNS. The main problem (as with
GABA, Glu, etc.) is that there is no distinct neuronal pathway
that may be isolated and stimulated; thus, all of the induced
activity is very generalized, making the significance of any
demonstrable release (metabolite or excess transmitter) very
difficult to interpret.

Despite these well-characterized functional properties, our

understanding of some aspects of the metabolism of Gly in
nervous tissue remains rudimentary. For example, we still do
not know whether biosynthesis is important for the
maintenance of Gly levels in the spinal cord or whether the
neurons depend on the uptake and accumulation of
preformed Gly. As indicated in Figure 6-9, Gly can be formed
from serine by a reversible folate-dependent reaction
catalyzed by the enzyme serine trans-hydroxymethylase.
Serine itself can also be formed in nerve tissue from glucose
via the intermediates 3-phosphoglycerate and 3-
phosphoserine. It is also conceivable that Gly might be
formed from glyoxylate via a transaminase reaction with Glu.
Although not established definitively, it appears likely that
serine serves as the major precursor of Gly in the CNS and
that serine hydroxymethyltransferase and D-glycerate
dehydrogenase are the best candidates for the rate-limiting
enzymes involved in the biosynthesis of Gly. Not only is our
knowledge of the metabolism of Gly in nervous tissue
minimal, but at the present time only scanty information is
available on the factors regulating the concentration of Gly in
the CNS. Gly in the spinal cord is labeled only slowly from
radioactive glucose via one of the glycolytic intermediates
indicated in Figure 6-9. It is also possible to label Gly by
administration of labeled glyoxylate, which is readily
transmitted to Gly by nervous tissue.

In summary, probably the most critical missing piece of

evidence to establish the inhibitory role of Gly in the spinal
cord is the demonstration that it is the main substance
contained in the terminals of the interneurons that synapse
on the motoneurons and that it is released from these
terminals when direct inhibition is produced. The fact that
these effects are blocked by strychnine, an antagonist
selective for the Gly receptor, makes a strong presumptive
case that Gly is responsible for the inhibitory actions. Some
partial support for the localization of Gly has been provided by
autoradiographic localization of Gly-uptake sites as visualized
by electron microscopy. However, demonstration of discretely
evoked release of Gly from spinal cord interneurons has been
more difficult to obtain.

Figure 6-9. Possible metabolic routes for the

formation and degradation of glycine by nervous
tissue. (Modified from Roberts and Hammerschlag,
1972.)

Gly as a Modulator of NMDA Receptors

A new role has been proposed for Gly that is distinct from its
established role as an inhibitory transmitter in lower brain
stem areas and in spinal cord mediated by a strychnine-
sensitive chloride conductance. Several groups have shown
that nanomolar concentrations of Gly increase the frequency
of opening of one of the subsets of Glu receptors, namely, the
NMDA receptor channel. This effect of Gly is strychnine-
insensitive, suggesting a mechanism involving allosteric
regulation of the NMDA receptor complex through a distinct
Gly-binding site (see Fig. 6-10). This action can be mimicked
with Gly agonists and blocked by other Gly antagonists,
although not strychnine. This allosteric concept is supported
by the existence of strychnine-insensitive Gly-binding sites
that have an anatomical distribution identical to that of the
NMDA receptor. When compared with the effects of
benzodiazepines on the GABA receptor, the enhancement of
NMDA responses observed with Gly is much greater. This
suggests that the main effect of Gly is to prevent
desensitization of the NMDA receptor during prolonged
exposure to agonists. Gly appears to accomplish this by
accelerating the recovery of the receptor from its desensitized
state rather than by blocking the onset of desensitization. Gly
facilitation of synaptic responses mediated by NMDA receptors
may be a common regulatory mechanism for excitatory
synapses, which raises the question of the processes that
regulate extracellular Gly concentration in brain regions where
NMDA receptors play a critical role in excitatory transmission.
An important question is whether endogenous Gly antagonists
(e.g., kynurenate) may also play a role in regulating neuronal
function where NMDA receptors are involved. However, recent
clinical trials of a selective Gly antagonist for the NMDA
receptor effect, previously reported to be neuroprotective in
animal models of stroke, have shown no such efficacy when
evaluated in the first 6 hours after a stroke.

Figure 6-10. Schematic illustration of the N-

methyl-D-aspartate (NMDA) receptor and the sites
of action of different agents on the receptor. The
NMDA receptor gates a cation channel that is
permeable to Ca2+ and Na+ and gated by Mg2+
in a voltage-dependent fashion; K+ is the
counterion. The NMDA receptor channel is blocked
by phencyclidine (PCP) and MK801, and the
complex is regulated at two modulatory sites by
glycine and polyamines; 2-amino-5-
 phosphonopentanoic acid (AP5) and 3-(2-
carboxypiperazin-4-yl)- propyl-1-phosphonic acid
(CPP) are competitive antagonists at the NMDA
site.

Glutamic Acid
Long before a role for Glu in neurotransmission was
established, it was recognized that certain amino acids, such
as Glu and Asp, occur in uniquely high concentrations in the
brain and that they can exert very powerful stimulatory
effects on neuronal activity. Thus, if any amino acid is
involved in the regulation of nerve cell activity, as an
excitatory transmitter or otherwise, it seems unnecessary to
look beyond these two candidates. The excitatory potency of
Glu was first demonstrated in crustacean muscle and later by
direct topical application to mammalian brain. However,
except for the invertebrate model, where substantial evidence
has accumulated to support a role for Glu as an excitatory
neuromuscular transmitter, its status as a neurotransmitter in
mammalian brain was uncertain for many years. This is
probably in part explainable by the fact that Glu (and Asp) is
a compound that is also intimately involved in intermediary
metabolism in neural tissue. For example, it has an important
function in the detoxification of ammonia in the brain, is an
important building block in the synthesis of proteins and
peptides including glutathione, and plays a role as a precursor
for the inhibitory neurotransmitter GABA. Thus, it has been
extremely difficult to dissociate the role this amino acid plays
in neuronal metabolism and as a precursor for GABA from its
possible role as a transmitter substance. Transport of
circulating Glu to the brain normally plays only a very minor
role in regulating the levels of brain Glu. In fact, the influx of
Glu from the blood across the blood-brain barrier is much
lower than the efflux of Glu from the brain.
Synthesis and Metabolism
In brain L-Glu is synthesized in the nerve terminals from two
sources: from glucose via the Krebs cycle and transamination
of -oxoglutatrate (Fig. 6-1) and from glutamine that is
synthesized in glial cells, transported into nerve terminals,
and locally converted by glutaminase into Glu (see Fig. 6-11).
In the Glu-containing nerve terminals, Glu is stored in
synaptic vesicles and, upon depolarization of the nerve
terminal, it is released by a calcium-dependent exocytotic
process. The action of synaptic Glu is terminated by a high-
affinity uptake process via the plasma membrane Glu
transporter on the presynaptic nerve terminal and/or on glial
cells. The Glu taken up into glial cells is converted by
glutamine synthetase into glutamine, which is then
transported via a low-affinity process into the neighboring
nerve terminals, where it serves as a precursor for Glu. In
astrocytes, glutamine can also be oxidized (via the Krebs
cycle) into -ketoglutarate, which can be actively transported
into the neuron to replace the -ketoglutarate lost during the
synthesis of neuronal Glu. As noted earlier in this chapter,
glutamine can replenish the transmitter pool of GABA via this
pathway as well.

Even though neuronal systems believed to utilize Glu or Asp

as transmitter substances have been described in the CNS,
because of their role in intermediary metabolism, it seems
quite unlikely that it will be possible to map these systems
accurately by simply tracking the presence of Glu or Asp or
their synthesizing enzymes, as has been done in the past for
the monoamines and GABA. The development of antibodies
against excitatory amino acids (EAAs), especially antisera that
can distinguish between Glu and Asp with a high degree of
selectivity as well as antibodies against their receptors and
transporters, has facilitated the anatomical mapping of EAA
pathways. The development of reliable methods for combining
anterograde labeling of primary afferent terminals with
immunocytochemistry has helped to identify afferent nerve
terminals enriched in Glu or Asp and to dissociate the role
played by these amino acids in neurotransmission from their
general role in metabolism. Nerve terminal enrichment in a
specific EAA provides the most direct anatomical evidence
that a pathway uses a particular amino acid as a
neurotransmitter. The elegant studies of Rustioni and co-
workers employing immunocytochemistry at the electron
microscopic level have identified primary afferent fibers
terminating in spinal laminae of the lumbar spinal cord with
nerve terminals enriched in Glu and/or Asp, providing direct
anatomical evidence that these primary afferents use these
EAAs as neurotransmitters. Use of these techniques in
conjunction with other less direct approaches, including
mapping the EAA receptors by ligand binding
autoradiography, has provided strong support for a
neurotransmitter role of EAAs in the mammalian CNS.

Figure 6-11. Pathways for glutamate utilization

and metabolism. Glutamate (Glu) released into
the synaptic cleft is recaptured by neuronal-type
(GT[n]) and glial-type (GT[g]) Na+-coupled
glutamate transporters. Glial glutamate is
converted to glutamine (Gln) by the enzyme
glutamine synthetase. Gln is present at high
concentrations in the cerebrospinal fluid (about
0.5 mM) and can enter the neuron to help
replenish glutamate after hydrolysis by
mitochondrial glutaminase. (Modified from
Nicholls, 1994.)

Release
Although the status of Glu and Asp as neurotransmitters has
suffered through many cycles of acceptability and
nonacceptability in the past several decades, the rush to clone
the Glu receptors and the extensive research on long-term
potentiation demonstrating a function for Glu have stopped
the doubters from speaking. In spite of considerable evidence
that Glu may be an excitatory transmitter in the CNS, little is
known about the biosynthesis and release of the pool of
releasable transmitter Glu. Utilizing the molecular layer of the
dentate gyrus of the hippocampal formation to provide a
definitive system in which the major input appears to be
glutaminergic, Cotman and co-workers addressed these
questions. Glu was shown to be released by depolarization
from slices of the dentate gyrus in a Ca2+-dependent
manner, and lesions of the major input to the dentate gyrus
originating from the entorhinal cortex diminished this release
and the high-affinity uptake of Glu. Glu biosynthesis in the
releasable pools was rapidly regulated by the activity of
glutaminase and by uptake of glutamine. These properties are
consistent with the properties expected of a neurotransmitter,
and the observations strengthened the premise that Glu may
be an important neurotransmitter in the molecular layer of
the dentate gyrus. Furthermore, these studies demonstrated
that the regulation of Glu synthesis and release share many
properties with other transmitters. For example, similar to
acetylcholine synthesis, Glu synthesis is regulated in part via
the accumulation of its major precursor, glutamine, and newly
synthesized Glu, like acetylcholine, is preferentially released.
In addition, Glu synthesis is regulated by end-product
inhibition. This is similar to the mechanism by which the rate-
limiting enzyme in catecholamine synthesis, tyrosine
hydroxylase, is regulated in catecholaminergic neurons by
dopamine and norepinephrine. It is interesting that these
similarities are demonstrable despite the involvement of Glu
in general brain metabolism.

Storage
The vesicle hypothesis describing the quantal release of
acetylcholine at the neuromuscular junction was introduced in
the mid-1950s. Since then, the concept of vesicular storage
and release of acetylcholine has become firmly established
and has been extended to a number of other synapses and
neurotransmitters. However, there was no direct experimental
evidence for the participation of synaptic vesicles in the
storage and release of EAA neurotransmitters until recently.
The concept has received strong support from the studies of
Jahn and Ueda and co-workers on isolated synaptic vesicles
from mammalian brain. These studies have shown that
vesicles are capable of Glu uptake and storage and that they
have a specific carrier for L-Glu. This concept was finally
validated with the cloning of the vesicular transporter for Glu.
The difficulty in cloning this vesicular transporter is most
likely due to the fact that it is unrelated to other known
transmitter transporters, although it shares significant
sequence homology with EAT-4, a worm protein implicated in
glutamatergic transmission. The vesicular Glu transporter was
identified as a protein that was previously suggested to
mediate the sodium-dependent transport of inorganic
phosphate across the membrane. It is found predominantly in
axon terminals, particularly those that form asymmetric
(excitatory) synapses. In contrast to the GABA transporter, for
which Gly is also a substrate (see Release and Reuptake), the
Glu vesicular transporter has a high specificity for Glu. This
transporter has been extensively characterized biochemically
and shown to play a key role in exocytosis.

Plasma Membrane Glu Transporter

Most of the molecular biological research on EAA transmission
has focused on receptors rather than the process of
transmitter inactivation. Inactivation is an especially
important process in EAA neurotransmission. The rapid
removal of Glu from the synapse by high-affinity uptake not
only serves to terminate the excitatory signal and recycle the
Glu but also plays an important role in the maintenance of
extracellular levels of Glu below those that could induce
excitotoxic damage. To date, five distinct subtypes of EAA
transporter (EAAT) have been identified, which together with
the neutral amino acid transporters appear to be part of a
novel gene family. These high-affinity, sodium-dependent
EAATs exhibit distinct anatomical and cellular distributions and
appear to have marked differences in pharmacological
specificity (Table 6-2).

These Glu transporter genes share 50% sequence homology

and exhibit minimal homology with other eukaryotic proteins,
including the superfamily of neurotransmitter transporters
that mediate the uptake of GABA, Gly, choline, and the
biogenic amines. These Glu transporters have distinct brain
distributions, and even the glial transporters exhibit regional
and intracellular differences in expression, underscoring the
heterogeneity of glia as well as neurons. All of the Glu
transporters demonstrate a strong Na+ dependence and are
enantioselective (i.e., D-Asp, L-Asp, and L-Glu are substrates,
whereas D-Glu is not). These transporters are also inhibited
by well-characterized uptake blockers including -threo-
hydroxy-Asp, dihydrokainate, and L-trans-2,4-pyrrolidine
decarboxylate. Preferential inhibition of several subtypes of
EAAT can be achieved with several of the newer nonsubstrate
antagonists. Neurons and glial cells appear to possess a
similar plasma membrane glutamate uptake carrier that
serves to terminate the postsynaptic action of
neurotransmitter Glu and to maintain the extracellular Glu
concentrations below levels that may damage neurons. The
presence of certain Glu transporters on glial cells is consistent
with the intricate interplay of glial and neuronal elements in
the synthesis and fate of Glu. Because the Glu that is released
from neurons is accumulated by glia and then metabolized to
glutamine, there is an ultimate recycling of the released
amino acid. The fate of Glu released from a neuron containing
the neuronal Glu transporter is unclear. It has not been
established if Glu released from a given neuron is taken up in
part by a Glu transporter on that particular neuron or
primarily by Glu transporters on other neurons or glia.

The normal mechanism for terminating the synaptic action of

Glu may involve all three processes to varying degrees.
Radiotracer studies with labeled Glu applied to the brain
indicate that most of the label is taken up into glial cells.
However, this may be an artifact of the method employed
since synaptically released, compared to exogenous, Glu may
be better positioned with regard to the uptake carrier in the
presynaptic nerve membrane than to the glial transporter.
Regardless, uptake appears to be the major process for
terminating the action of released Glu. There does not appear
to be any significant role for enzymatic inactivation of Glu
similar to that observed with GABA and the other classical
neurotransmitters.

It is troublesome to some analysts that Glu, Asp, and

synthetic derivatives of these dicarboxylic acids result in
almost universal excitation of neuronal discharge, perhaps
reflected in their far more ubiquitous distribution. Glu does
not bring the cell membrane potential to the same level as
the natural excitatory transmitter. In addition, both the D and
the naturally occurring L isomers of EAA are active, although
in the case of Glu the D isomer is often reported to be
somewhat less active. These findings have led some highly
skeptical investigators to suggest that the response to amino
acids represents a nonspecific receptivity of the neuron to
these agents and is therefore not necessarily indicative of a
transmitter function.

Based on the abundant current evidence, however, Glu

appears to have satisfied four of the main criteria for
classification as an excitatory neurotransmitter in the
mammalian CNS: (1) it is localized presynaptically in specific
neurons, where it is stored in and released from synaptic
vesicles; (2) it is released by a Ca-dependent mechanism by
physiologically relevant stimuli in amounts sufficient to elicit
postsynaptic responses; (3) a mechanism (reuptake and
specific transporters) exists that will rapidly terminate its
transmitter action; and (4) it demonstrates pharmacological
identity with the naturally occurring transmitter.

The most clear-cut evidence that EAAs can act physiologically

as excitatory neurotransmitters at a given synapse comes
from experiments in which intracellular recordings of pre- and
postsynaptic events have been made. In studies of this
nature, the criterion of identical and parallel change induced
by antagonists on synaptic events elicited by stimulation and
those elicited by the action of the exogenously administered
putative transmitter substance is of critical importance in
identifying the excitatory amino transmitter involved. In many
situations, however, such studies are not feasible, so more
indirect studies like those summarized above have been
utilized.

Quite often at excitatory synapses the actual molecule acting

at the postsynaptic receptor has not been definitively
identified, even though pharmacological analysis indicates
that the synaptic response is mediated by a particular EAA
receptor. Thus, a critical link in establishing a
neurotransmitter role for an EAA within a specific pathway is
the demonstration and characterization of the EAA receptor
that mediates synaptic transmission at that synapse.

EAA Receptors
Our understanding of EAA transmitters and their function and
regulation has been greatly enhanced by studies directed at
the identification, characterization, localization, and isolation
of receptors for these amino acids. In fact, progress on the
definition of receptor subtypes and the availability of more
selective agonists and antagonists has produced a quantum
leap in knowledge about EAAs at synaptic sites throughout
the vertebrate CNS. Furthermore, many aspects of Glu
receptor dynamics have suggested an extraordinary degree of
functional regulation.

Until the mid-1980s, neuropharmacologists were content with

two major classes of EAA receptors, the NMDA receptors and
the non-NMDA receptors, the latter composed at that time of
kainate and quisqualate receptors. With the development of
more selective agonists and antagonists, however, the classes
of EAA receptors have expanded to at least five different
types (NMDA, kainate, -amino-3-hydroxy-5-methylisoxazole-
4-propionic acid [AMPA], 1,2-amino-4-phosphonobutyrate
[AP4], and 1-aminocyclopentane-1,3-dicarboxylic acid
[ACPD]; see below) in the CNS, each displaying distinct
physiological characteristics. Three of these receptors have
been defined by the depolarizing excitatory actions of select
synthetic agonists (NMDA, kainate, and AMPA) and the
blockade of the effects of these agonists by selective
antagonists. A fourth, the AP-4 receptor, appears to represent
an inhibitory autoreceptor. The fifth receptor, activated by
ACPD, modifies inositol phosphate metabolism and has been
called a metabotropic Glu receptor (mGluR). A summary of
representative agonists and antagonists for each of these EAA
receptor classes is given in Table 6-3.

Synaptic transmission in synapses using EAAs does not

appear to follow the simple model of fast-acting synaptic
transmission mediated by a single receptor class. In fact,
individual synapses that use EAAs may not be restricted to
distinct receptors but rather may have a combination of
receptors and thereby exhibit different input/output
properties and second-messenger responses.

One specific subtype of EAA receptor, the NMDA receptor, has

become a major focus of attention because of evidence that it
may be involved in a wide range of both neurophysiological
and pathological processes as important and diverse as
memory acquisition (see Chapter 13), developmental
plasticity, epilepsy, and the neurotoxic effects of brain
ischemia.

NMDA Receptor-Ionophore Complex

The NMDA receptor is a ligand-gated ion channel composed of
two different protein subunits, NMDAR1 and NMDAR2.
NMDAR1 can exist in seven splice variants, and there are four
different genes encoding variants of NMDAR2 (NMDAR2A, -B,
-C, and -D). At present, it is not clear how many NMDAR1 and
NMDAR2 subunits are present in each functional NMDA
receptor. This receptor complex has been extensively
characterized physiologically and pharmacologically and is
widely distributed in mammalian brain and spinal cord, with
particularly high receptor densities found in hippocampus and
cerebral cortex. NMDA receptors appear to have a pivotal role
in long-term depression, long-term potentiation (LTP), and
developmental plasticity. However, overactivation or
prolonged stimulation of NMDA receptors can damage and
eventually kill target neurons via a process referred to as
excitotoxicity. Like the GABAA receptor, the NMDA receptor is
a complex molecular entity endowed with a number of distinct
recognition sites for endogenous and exogenous ligands, each
with discrete binding domains. At present, there appear to be
at least six pharmacologically distinct sites through which
compounds can alter the activity of this receptor (see Fig. 6-
10): (1) a transmitter binding site that binds L-glutamate and
related agonists, promoting the opening of a high-
conductance channel that permits entry of Na and Ca into
target cells, L-Glu being virtually ineffective unless the site
that binds Gly, the strychnine-insensitive Gly-modulatory site
(2), is also occupied (this Gly site is distinct from the Gly-
binding site on the strychnine-sensitive Gly-inhibitory
receptor, see discussion of Gly receptors above Glycine as an
Inhibitory Transmitter, which is present in high density in the
brain stem and spinal cord); (3) a site within the channel that
binds phencyclidine (PCP site) and related noncompetitive
antagonists (MK-801, ketamine), which act most effectively
when the receptor is activated (i.e., open channel block); (4)
a voltage-dependent Mg2+-binding site; (5) an inhibitory
divalent cation site near the mouth of the channel that binds
Zn2+ to produce a voltage-independent block; and (6) a
polyamine-regulatory site whose activation by spermine and
spermidine facilitates NMDA receptor-mediated transmission.

In addition, two distinct binding sites are apparently

associated with the transmitter recognition site, one that
preferentially binds agonists and one that preferentially binds
antagonists. It is of interest that quinolinic acid, a metabolite
of tryptophan and thus a natural brain constituent, may be a
specific antagonist for a particular subtype of NMDA receptor
since it shows regional variation in potency. A number of Gly
agonists and antagonists have already been identified for the
Gly-regulatory site, even though the Gly-binding component
of the NMDA receptor was discovered only recently.

The Gly-modulatory site has attracted a great deal of interest

as a potential site for the action of new antiepileptic drugs or
agents that might be useful in preventing ischemic brain
damage. D-Serine is a potent agonist at this site, and (+)HA-
966 is a selective antagonist. Gly in submicromolar
concentrations increases the frequency of the NMDA receptor
channel opening in a strychnine-insensitive manner, and,
even though brain and cerebrospinal fluid concentrations of
Gly are in the millimolar range, Gly is effective in vivo,
indicating that the Gly site is not saturated. Thus, conditions
that alter the extracellular concentration of Gly or compete
with its binding site can dramatically alter NMDA receptor-
mediated responses.
D-Serine is an endogenous ligand for the Gly site of the NMDA
receptor. Strikingly, D-serine occurs exclusively in glia and not
in neurons. It appears to be selectively localized in astrocytes
containing its biosynthetic enzyme, serine racemase, which
ensheathe nerve terminals in brain areas enriched in NMDA
receptors. Thus, D-serine has been suggested to play a role in
the modulation of synaptic transmission at NMDA receptors,
providing a mechanism by which astrocytes may play an
active role in supporting synaptic transmission via the
activation of NMDA receptors.

Polyamines such as spermine and spermidine function as

allosteric modulators of NMDA receptors and potentiate NMDA
currents in the presence of saturating concentrations of Glu
and Gly. However, in contrast to Gly or D-serine, their
presence is not a requirement for NMDA receptor activation.
Under pathological conditions, such as brain ischemia or
trauma, in which the production of polyamines is dramatically
increased, polyamines may mediate or potentiate the
excitotoxic mechanisms responsible for the neuronal damage
produced. This idea is supported by findings that
phenylethanolamines, for instance, ifenprodil and eliprodil
(SL820715), which are potent antagonists of the polyamine
modulatory site of the NMDA receptor complex in a number of
biochemical models, exhibit effective neuroprotective action in
ischemia and trauma.

Kynurenate is a tryptophan metabolite that blocks the high-

affinity Gly-binding site. It can be thought of as an
endogenous neuroprotective agent that is released from glial
cells following the transamination of L-kynurenine.
Manipulation of endogenous kynurenate formation may
provide further insight into the role it plays in modulating the
action of Gly on the NMDA receptor.

Quinolinic acid is an endogenous excitatory neurotoxin,

synthesized from L-tryptophan via the kynurenine pathway,
that has the potential of mediating NMDA-induced
neurotoxicity and dysfunction. Its endogenous occurrence in
normal brain is relatively low (approximately 50-100
pmoles/g). Toxicity induced by exogenous quinolinic acid can
be prevented or reversed by noncompetitive or competitive
NMDA antagonists, suggesting that the neurotoxicity
produced by locally administered quinolinic acid is mediated
through the NMDA subtype of EAA receptors. Although
quinolinic acid was first identified in the human brain in 1983,
knowledge of the possible role played by this toxin in
neuropathology was uncertain while speculation abounded.
The observation by Heyes and co-workers in the late 1980s
that quinolinate levels were dramatically elevated in acquired
immunodeficiency syndrome (AIDS) was a major advance
toward understanding the pathobiology of quinolinate. Further
studies have demonstrated that the motor signs of AIDS
dementia correlate strikingly with the levels of quinolinate in
the cerebrospinal fluid. Treatment with the antiretroviral
agent azidothymidine decreased the viremia and associated
dementia and lowered the cerebrospinal fluid levels of
quinolinate. Because cerebrospinal fluid levels of quinolinate
are often higher than those found in blood, it has been argued
that the origin of this neurotoxin may be via intracerebral
synthesis, although this has not yet been documented. The
highest reported levels of cerebrospinal fluid quinolinate are
often obtained in AIDS patients with opportunistic infections.
Elevated levels of cerebrospinal fluid quinolinate are found in
human patients and in nonhuman primates with inflammatory
neurological diseases, but it is unclear whether the
mechanisms underlying the increased levels of quinolinate
induced by bacterial and viral infection are similar. It also
remains to be determined if the bacteria are actually involved
in the synthesis of some of the increased quinolinate. No
specific organism has been identified as responsible for
elevating quinolinic acid since this neurotoxin is elevated in a
variety of bacterial and viral infections. The discovery of an
effective inhibitor of quinolinic acid synthesis will help to
elucidate the importance of this neurotoxin in inflammatory
and infectious disease.

In addition to the multiple ligand-binding and regulatory sites

on the NMDA receptor for these relatively small molecules,
the NMDA receptor exhibits an extremely high degree of
interaction with many other membrane and cytoplasmic
proteins. In fact, the complexity of the proteins with which
this receptor has demonstrable interactions has established a
new form of high-throughput molecular analysis combining
the classical forms of two-dimensional gel chromatography
with very sensitive methods of peptide sequence identification
employing mass fragmentography. Using such "proteomic"
methods, the NMDA receptor's known interacting partners
increased from a single association with the intrinsic
membrane protein PSD95, which was thought to be critical to
anchor the receptor to the postsynaptic specialization site, to
more than six dozen other proteins. Within this NMDA
receptor protein complex, five main classes of proteins have
so far been identified: neurotransmitter receptors (including
other Glu receptors), cell adhesion proteins, adaptors,
signaling enzymes, and cytoskeletal proteins.

Non-NMDA Receptors
Both AMPA and kainic acid (KA) receptors mediate fast
excitatory synaptic transmission and are associated primarily
with voltage-independent channels that gate a depolarizing
current primarily carried by the influx of Na+ ions. While
these receptors are easily distinguished from NMDA receptors,
they are more difficult to distinguish from each other.
Molecular biological studies have confirmed the existence of
AMPA and KA classes of non-NMDA receptors but have
revealed a considerable degree of heterogeneity within these
two families. Some pharmacological discrimination can be
achieved, with AMPA and quisqualate being the preferred
agonists for AMPA receptors and domoate and kainate the
preferred agonists for KA receptors. The most selective and
potent non-NMDA antagonists available are a series of
dihydroxyquinoxaline derivatives including 2,3-dihydro-6-
nitro-7-sulfamoyl-benzo(F)quinoxaline (NBQX), 6-cyano-7-
nitroquinoxaline-2,3-dione (CNQX), and 6,7-
dinitroquinoxaline-2,3-dione (DNQX) (see Table 6-3).
However, these agents competitively block both types of non-
NMDA receptor, although NBQX appears to exhibit the best
selectivity for AMPA receptors. Very few KA receptor-selective
compounds have been identified. A new class of 2,3-
benzodiazepine derivatives (most notably GYKI-52466) has
been shown to block AMPA-induced responses
noncompetitively and to attenuate ischemic neuronal damage
effectively in animal models, highlighting the fact that non-
NMDA receptors also play an important role in CNS pathology.

In studies of immature hippocampal neurons in culture, AMPA

receptors were surprisingly found almost exclusively within
the dendritic cytoplasm rather than in the dendritic
membranes and spines. Using special constructs of the mRNA
for the AMPA receptor coupled with green fluorescent protein,
regular dynamic insertion and removal could be demonstrated
for the receptors, which was accelerated by activity. This
receptor cycling was quickly advanced as a mechanism for
activating silent synapses by recruiting these receptors to the
synaptic surfaces. However, whether this explanation would
hold for mature neurons is dubious since such neurons exhibit
abundant spine and synaptic AMPA receptors. Recent
evidence suggests that in mature neurons, it may well be the
NMDA receptors that are recruited to the cell surface during
depolarizing events.

Metabotropic Glutamate Receptors

The mGluRs constitute a family of EAA receptors that are
linked to G proteins and second-messenger systems and are
distinct from the ionotropic EAA receptors that form ion
channels and are comprised of the NMDA, AMPA, and kainate
subtypes discussed above. This more recently characterized
group of receptors is coupled to a variety of signal-
transduction pathways via G proteins, producing alterations in
intercellular second messengers and generating slow synaptic
responses. This is in clear contrast to the ionotropic Glu
receptors, which are directly coupled to cation-specific ion
channels and mediate fast excitatory synaptic responses.
Metabotropic is a term that was coined to indicate that these
receptors, unlike the inotropic receptors which form ion
channels, affect cellular metabolic processes. Unfortunately,
this nomenclature is very misleading since mGluRs, similar to
other G protein-coupled receptors, exert profound effects on
neuronal function through the regulation of ion channels,
protein phosphorylation, and second-messenger cascades.
The widespread distribution of metabotropic receptors in the
CNS coupled with the prevalence of Glu as a neurotransmitter
indicates that this system is a major modulator of second
messengers in the mammalian CNS. Molecular cloning studies
have revealed the existence of at least eight different
subtypes of mGluR, mGluR1 through mGluR8, which have a
common structure of a large extracellular domain preceded by
the seven-member spanning domains.

Members of the mGluR family can be divided into three

subgroups according to their sequence similarities, signal-
transduction properties, and pharmacological profiles to
agonists (i.e., relative potencies when expressed in cell lines
of Glu, quisqualate, ACPD, and AP-4). The first subgroup,
comprised of mGluR1 and mGluR5, is coupled to the
stimulation of phosphatidylinositol hydrolysis/Ca2+ signal
transduction. The second group, mGluR2 and mGluR3, is
negatively coupled through adenylyl cyclase to cyclic
adenosine monophosphate formation. The third group,
mGluR4, mGluR6, mGluR7, and mGluR8, is also negatively
linked to adenylyl cyclase activity but shows a different
agonist preference from that of mGluR2 and mGluR3. As a
group, the mGluRs are widely expressed throughout the
brain, but the individual subtypes show some differential
distribution. The pharmacology of the individual subtypes
expressed in Chinese hamster ovary or in baby hamster
kidney cells shows some interesting differences (Table 6-3).
LAP4 is a potent agonist of mGluR4, mGluR6, mGluR7, and
mGluR8 but has little effect on the other receptor subtypes. L-
(2S,3S,4S)- -(carboxycyclopropyl) Gly (L-CCG-I), however,
activates mGluR2 at concentrations that have little or no
effect on mGluR1 and mGluR4. No agonist yet identified
appears to be specific for any single metabotropic receptor
subtype.

Considerable experimental evidence indicates that the

mGluRs are involved in the regulation of synaptic
transmission in the CNS. However, the lack until recently of
specific antagonists has limited the precise characterization of
the role of individual mGluRs in glutamatergic transmission
and has severely hampered progress in identifying their
physiological and pathological roles. The discovery that
phenylglycine derivatives are selective antagonists of mGluRs
has permitted more rigorous testing of the physiological role
of this receptor subclass in brain function and dysfunction.
Data are emerging that suggest a role in both synaptic
transmission and synaptic plasticity. The most exciting new
development concerning MGluRs is the finding by Conquet
and collaborators using knockout mice, that mGluR5 is
probably an essential factor in cocaine self-administration and
locomotor effects. Their studies showed that the reinforcing
properties of cocaine are absent in mice lacking mGluR5 and
that a selective mGluR5 antagonist, 2-methyl-6-
(phenylethynyl)-pyridine, dose-dependently decreased
cocaine self-administration (Fig. 6-12). This finding should
provide a new incentive for the rapid development of more
potent mGluR antagonists with greater subtype specificity and
enhanced bioavailability. With the availability of these new
subtype-selective antagonists, the next few years should
witness major advances in our knowledge of the roles played
by mGluRs in physiological and pathological processes and
perhaps even a new treatment of addictive disorders.

The function and distribution of the four classes of EAA

receptors are summarized in Table 6-4. The NMDA receptor is
an essential component in the generation of LTP. LTP results in
an increase in synaptic efficacy that has been proposed as an
underlying mechanism involved in memory and learning (see
Chapter 13).

In addition to the roles excitotoxic mechanisms may play in

various chronic neurodegenerative disorders like Huntington's
disease and viral diseases like AIDS, two chronic neurological
syndromes have been linked to dietary consumption of amino
acid toxins of plant origin. Neurolathyrism, a spastic disorder
occurring in eastern Africa and southern Asia, is associated
with dietary consumption of the chick pea Lathyrus sativus. -
N-Oxalylamino- L-alanine (L-BOAA) has been identified as the
responsible toxin in this plant. This amino acid acts as an
agonist at AMPA receptors. One of the primary effects of L-
BOAA toxicity is the inhibition of mitochondrial complex 1
selectively in the motor cortex and lumbar spinal cord. Guam
disease, also referred to as amyotrophic lateral
sclerosis/parkinsonism/dementia, is thought to be related to
the consumption of flour prepared from the seeds of the
cycad Cycas circinalis, which contains the amino acid -N-
methyl-amino-L-alanine (BMAA). Although BMAA is a neutral
amino acid that is not directly excitatory or toxic in vitro, in
the presence of bicarbonate it becomes excitotoxic and acts
as an agonist at AMPA and NMDA receptors.
Neurotoxicity has also been observed following ingestion of
domoic acid. Domoic acid is an analog of KA that is about
three times as potent. This substance is synthesized by
seaweed and can be consumed in toxic amounts by eating
mussels that have fed on the seaweed. An outbreak of domoic
poisoning occurred in 1987 in Canada. Consumption of this
neurotoxin by humans can damage the hippocampus and
produce dementia.

With the availability of more specific pharmacological agents,

it should be possible to evaluate in more detail the
involvement of EAA pathways in normal brain function and in
neuropathological conditions. The participation of NMDA
receptors in LTP provides a strong link between these systems
and the mechanisms of learning and memory. NMDA and
other EAA receptors also appear to play a role in cell damage
caused by hypoglycemia, hypoxia, seizures, and other
disturbances associated with excess EEAs.

Figure 6-12. Structures of group 1-selective

metabotropic glutamate (mGlu) receptor
antagonists. LY344545 is a selective mGlu5
receptor competitive antagonist. SIB1893 and
MPEP [2-methyl-6-(phenylethynyl)-pyridine] are
two new noncompetitive selective mGlu5 receptor
antagonists.

SELECTED REFERENCES

GABA
Bowery, N. G. (1993). GABAB receptor pharmacology. Annu.
Rev. Pharmacol. Toxicol. 33, 109-147.

Jursky, F., S. Tamura, A. Tamura, S. Mandiyan, H. Nelson, and

N. Nelson (1994). Structure, function and brain localization of
neurotransmitter transporters. J. Exp. Biol. 196, 283-295.

LUddens, H. and W. Wisden (1991). Function and

pharmacology of multiple GABAA receptor subunits. Trends
Pharmacol. Sci. 12, 49-51.

Macdonald, R. L. and Olsen, R. W. (1994). GABAA receptor

channels. Annu. Rev. Neurosci. 17, 569-602.

Mellon, S. H. and L. S. Griffin (2002). Neurosteroids:

biochemistry and clinical significance. Trends Endocrinol.
Metab. 13, 35-43.

Mohler, H., F. Crestani, and U. Rudolph (2001). GABAA

receptor subtypes: a new pharmacology. Curr. Opin.
Pharmacol. 1, 22-25.

Paul, S. M. (1995). GABA and glycine. In

Psychopharmacology: The Fourth Generation of Progress (F.
E. Bloom and D. J. Kupfer, eds.). Raven Press, New York, pp.
87-94.

Purdy, R. H., A. L. Morrow, P. H. Moore, Jr., and S. M. Paul

(1991). Stress-induced elevations of -aminobutyric acid type
A receptor-active steroids in the rat brain. Proc. Natl. Acad.
Sci. U.S.A. 88, 4553-4557.

Rudolph, U., F. Crestani, and H. Mohler (2001). GABAA

receptor subtypes: dissecting their pharmacological functions.
Trends in Pharmacol. Sci. 22, 1188-194.

Scott, R. H., K. G. Sutton, and A. C. Dolphin (1993).

Interactions of polyamines with neuronal ion channels. Trends
Neurosci. 16, 153-160.
Sieghart, W. (2000). Unraveling the function of GABAA
receptor subtypes. Trends Pharmacol. Sci. 21, 411-413.

Glycine
Becker, C.-M. (1990). Disorders of the inhibitory glycine
receptor: the spastic mouse. FASEB J. 4, 2767-2774.

Betz, H. (1992). Structure and function of inhibitory glycine

receptors. Q. Rev. Biophys. 25, 381-394.

Ottersen, O. P. and J. Storm-Mathisen (eds.) (1990). Glycine

Neurotransmission. John Wiley Sons, New York.

Roberts, E. and R. Hammerschlag (1972). Basic

Neurochemistry. Little, Brown, Boston, pp. 131-163.

Vandenberg, R. J., C. A. Handford, and P. R. Schofield (1992).

Distinct agonist- and antagonist-binding sites on the glycine
receptor. Neuron 9, 491-496.

Vannier, C. and A. Triller (1997). Biology of the postsynaptic

glycine receptor. Int. Rev. Cytol. 176, 201-244.

Zafra, F., C. Aragon, and C. Gimenez (1997). Molecular

biology of glycinergic neurotransmission. Mol. Neurobiol. 14,
117-142.

Excitatory amino acids

Bridges, R. J. (2001). The ins and outs of glutamate
transporter pharmacology. Tocris Rev. 17, 1-5.

Conn, P.J., and J.-P. Pin (1997). Pharmacology and functions

of metabotropic glutamate receptors. Annu. Rev. Pharmacol.
Toxicol. 37, 205-237.
Chiamulera, C., M. P. Epping-Jordan, A. Zocchi, C. Marcon, C.
Cottiny, S. Tacconi, M. Corsi, F. Orzi, and F. Conquet (2001).
Reinforcing and locomotor stimulant effects of cocaine are
absent in mGluR5 null mutant mice. Nat. Neurosci. 4, 873-
874.

Cotman, C. W., J. S. Kahle, S. E. Miller, J. Ulas, and R. J.

Bridges (1995). Excitatory amino acid neurotransmission. In
Psychopharmacology: The Fourth Generation of Progress (F.
E. Bloom and D. J. Kupfer, eds.). Raven Press, New York, pp.
75-85.

Grant, S. N. and W. P. Blackstock (2001). Proteomics in

neuroscience: from protein to network. J. Neurosci. 21, 8315-
8318.

Grosshaus, D. R., D. A. Clayton, S. J. Coultrop, and M. D.

Browning (2002). LTP leads to rapid surface expression of
NMDA but not AMPA receptors in adult rat CA1. Nat. Neurosci.
5, 27-33.

Hayashi, Y., N. Sekiyama, S. Nakanishi, D. E. Jane, D. C.

Sunter, E. F. Birse, P. M. Udvarhelyi, and J. C. Watkins (1994).
Analysis of agonist and antagonist activities of phenylglycine
derivatives for different cloned metabotropic glutamate
receptor subtypes. Neuroscience 14, 3370-3377.

Husi, H, M. A. Ward, J. S. Choudhary, W. P. Blackstock, and S.

G. Grant (2000). Proteomic analysis of NMDA receptor-
adhesion protein signaling complexes. Nat. Neurosci. 3, 661-
669.

Ishii, T., K. Moriyoshi, H. Sugihara, K. Sakurada, H. Kadotani,

M. Yokoi, C. Akazawa, R. Shigemoto, N. Mizuno and M. Masu
(1993). Molecular characterization of the family of the N-
methyl-D-aspartate receptor subunits. J. Biol. Chem. 268,
2836- 2843.

Kemp, J. A. and P. D. Leeson (1993). The glycine site of the

NMDA receptor-five years on. Trends Pharmacol. Sci. 14, 20-
25.

Mothet, J.-P., A. T. Parent, H. Wolosker, R. O. Brady, Jr., D. J.

Linden, C. D. Ferris, M. A. Rogawski, and S. H. Snyder
(2000). D-Serine is an endogenous ligand for the glycine site
of the N-methyl-D-aspartate receptor. Proc. Natl. Acad. Sci.
U.S.A. 97, 4926-4931.

Nicholls, D. G. (1994). Proteins, Transmitters and Synapses.

Blackwell Science, Cambridge, MA.

Ravindranath, V. (2002). Neurolathyrism. mitochondrial

dysfunction in excitotoxicity mediated by L-beta-oxalyl
aminoalanine. Neurochem. Int. 40(6), 505-509.

Reinhard, J. F., Jr., J. B. Erickson, and E. M. Flanagan (1994).

Quinolinic acid in neurological disease: opportunities for novel
drug discovery. Adv. Pharmacol. 30, 85-127.

Rogawski, M. A. (1993). Therapeutic potential of excitatory

amino acid antagonists: channel blockers and 2,3-
benzodiazepines. Trends Pharmacol. Sci. 14, 325-331.

Sacco, R. L., J. T. DeRosa, E. C. Haley, Jr., B. Levin, P.

Ordronneau, S. J. Phillips, T. Rundek, R. G. Snipes, J. L.
Thompson and for the GAIN Americas Investigators (2001).
Glycine antagonist in neuroprotection for patients with acute
stroke GAIN Americas: a randomized controlled trial. JAMA
285, 1719-1728.

Scatton, B. (1993). The NMDA receptor complex. Fundam.

Clin. Pharmacol. 7, 389-400.
Schoepp, D. D., D. E. Jane, and J. A. Monn (1999).
Pharmacological agents acting at subtypes of metabotropic
glutamate receptors. Neuropharmacology 38, 1431-1476.

Seeburg, P. H. (1993). The molecular biology of mammalian

glutamate receptor channels. Trends Pharmacol. Sci. 14, 297-
303.

Shi, S.-H. (2001). AMPA receptor dynamics and synaptic

plasticity. Science 294, 1851-1852.

Spooren, W. P. S., F. Gasparomo, T. E. Salt, and R. Kuhn

(2001). Novel allosteric antagonists shed light on mGlu5
receptors and CNS disorders. Trends Pharmacol. Sci. 22, 331-
337.

Takamori, S., J. S. Rhee, C. Rosenmund, and R. Jahn (2000).

Identification of a vesicular glutamate transporter that defines
a glutaminergic phenotype in neurons. Nature 407, 189-194.

Valtschanoff, J. G., K. D. Phend, P. S. Bernardi, R. J.

Weinberg, and A. Rustioni (1994). Amino acid
immunocytochemistry of primary afferent terminals in the rat
dorsal horn. J. Comp. Neurol. 346, 237-252.

Watkins, J. and G. Collingridge (1994). Phenylglycine

derivatives as antagonists of metabotropic glutamate
receptors. Trends Pharmacol. Sci. 15, 333-342.

(+/-) Show / Hide Bibliography

Send Feedback Teton Server (5.0.0) - ©2007 Teton Data Title Updates
Customer Service Systems User
800.901.5494 Send Us Your Comments Responsibilities
Training Center
What's New
7. Acetylcholine
INTRODUCTION
The neurophysiological activity of acetylcholine (ACh) has
been known since the turn of the century and its
neurotransmitter role since the mid-1920s. With this long
history, it is not surprising when students assume that
everything is known about the subject. Unfortunately, the
delay in developing sophisticated methods for determining the
presence of ACh in cholinergic tracts and terminals, which
only recently has been overcome, has left this field far behind
the biogenic amines. The structural formula of ACh is
presented below:

ASSAY PROCEDURES
ACh may be assayed by its effect on biological test systems or
by physiochemical methods. Popular bioassay preparations
include the frog rectus abdominis, the dorsal muscle of the
leech, the guinea pig ileum, the blood pressure of the rat (or
cat), and the heart of Venus mercenaria. In general,
bioassays tend to be laborious, to be subject to interference
by naturally occurring substances, and on occasion to behave
in a mysterious fashion (e.g., the frog rectus abdominis is not
as sensitive to ACh in the summer months as in the winter).
Nevertheless, bioassays currently represent one of the most
sensitive (0.01 pmol in the toad lung) and, under properly
controlled conditions, the most specific procedures for
determining ACh. It is probably fair to say that the
neurochemically oriented investigators' natural fear and
distrust of a bioassay have hampered progress in elucidating
the biochemical and biophysical aspects of ACh. This
statement is supported by a consideration of the plethora of
information on norepinephrine. This neurotransmitter can also
be bioassayed, but it was only after the development of
sensitive fluorometric and radiometric procedures for
determining components of the adrenergic nervous system
that the information explosion occurred.

Until about 1965, physiochemical methods for determining

ACh were so insensitive as to be virtually useless for
measuring endogenous levels. Since then, however, papers
have been published on enzymatic, fluorometric, gas
chromatographic, chemiluminescent, and radioimmunoassay
techniques that approach the sensitivity and specificity of the
bioassays. Currently, the most popular assays are the gas
chromatographic-mass spectrometric procedure of Jenden
and co-workers, the radiometric procedure of McCaman and
Stetzler, the chemiluminescent assay of Israel and Lesbats,
and the high-performance liquid chromatographic-
electrochemical procedure of Potter and colleagues. Ricny and
co-workers have modified and combined several existing
methods to develop a highly sensitive (0.2 pmol)
determination of the transmitter, and the sensitivity has been
further increased in the procedure of Flentge et al.

SYNTHESIS
ACh is synthesized in a reaction catalyzed by choline
acetyltransferase (ChAT):

Before entering into a discussion of ChAT, we should take note

of Figure 7-1, which depicts the possible sources of acetyl
coenzyme A (CoA) and choline. In theory, the acetyl CoA for
ACh synthesis may arise from glucose, through glycolysis and
the pyruvate dehydrogenase system; from citrate, either by
reversal of the condensing enzyme (citrate synthetase) or by
the citrate cleavage enzyme (citrate lyase); or from acetate
through acetate thiokinase. In brain slices, homogenates,
acetone powder extracts, and preparations of nerve-ending
particles, glucose, or citrate are the best sources for ACh
synthesis, with acetate rarely showing any activity. In lobster
axons, the electric organ of the Torpedo, corneal epithelium,
and frog neuromuscular junction, acetate appears to be the
preferred substrate. However, all of these systems are in vitro
and do not necessarily reflect the situation in vivo. Regardless
of its source, acetyl CoA is primarily synthesized in
mitochondria. Since, as detailed below, ChAT appears to be in
the synaptosomal cytoplasm, another still unsolved problem is
how acetyl CoA is transported out of the mitochondria to
participate in ACh synthesis. A probable carrier for acetyl CoA
is citrate, which can diffuse into the cytosol and produce
acetyl CoA via citrate lyase; a possible carrier is acetyl
carnitine, and another possibility is Ca2+-induced leakage of
acetyl CoA from mitochondria.

Although choline can be synthesized de novo in brain by

successive methylations of ethanolamine, the extent is too
minor to be of significance. Rather, choline is transported to
the brain both free and in phospholipid form (possibly as
phosphatidylcholine) by the blood. Following the hydrolysis of
ACh, about 35%-50% of the liberated choline is transported
back into the presynaptic terminal by a sodium-dependent,
high-affinity active transport system, to be reutilized in ACh
synthesis. As outlined in Figure 7-1, the remaining choline
may be catabolized or become incorporated into
phospholipids, which can again serve as a source of choline. A
curious observation is that when brain cortical slices are
incubated for 2 hours in a Krebs-Ringer medium, choline
accumulates to about 10 times its original concentration.
Similarly, a rapid postmortem increase in choline has been
observed. The precise source of this choline is unknown; a
probable candidate is phosphatidylcholine.
 Figure 7-1. Acetylcholine (ACh) metabolism. CoA,
coenzyme A; CTP, cytidine triphosphate.

CHOLINE TRANSPORT
Recent studies of choline transport have produced a number
of significant findings.

1. Choline crosses cell membranes by two processes, referred

to as high-affinity and low-affinity transport. High-affinity
transport, with a Km for choline of 1-5 uM, is saturable,
carrier-mediated, dependent on sodium, and stimulated by
chloride. It is also dependent on the membrane potential of
the cell or organelle so that any agent (e.g., K+) that
depolarizes the cell will concurrently inhibit high-affinity
transport. Low-affinity choline transport, with a Km of 40-80
mM, appears to operate by a passive diffusion process, to be
linearly dependent on the concentration of choline, and to be
virtually nonsaturable.

2. In contrast to the other neurotransmitters, ACh is taken up

in terminals only via low-affinity transport; it is only choline
that exhibits high-affinity kinetics.

3. Current evidence suggests that the high-affinity transport

of choline is specific for cholinergic terminals and is not
present in aminergic nerve terminals. Furthermore, it is
kinetically (but not physically) coupled to ACh synthesis.
About 50%-85% of the choline that is transported by the
high-affinity process is utilized for ACh synthesis. Low-affinity
transport, however, is found in cell bodies and in tissues such
as the corneal epithelium, and it is thought to function in the
synthesis of choline-containing phospholipids. Tissues that do
not synthesize ACh (e.g., fibroblasts, erythrocytes,
photoreceptor cells) exhibit high-affinity choline transport that
is coupled to phospholipid synthesis.

4. Hemicholinium-3 is an extremely potent inhibitor of high-

affinity transport (Km of 0.05-1 uM) but a relatively weak
inhibitor of low-affinity transport (Km of 10-120 uM).

5. There are three obvious mechanisms for regulating the

level of ACh in cells: feedback inhibition by ACh on ChAT,
mass action, and the availability of acetyl CoA and/or choline.
Of these three possibilities, the major regulatory factor seems
to be high-affinity choline transport. This view derives from
early observations that choline is rate-limiting in the synthesis
of ACh coupled with findings in a number of laboratories.
Using the septal-hippocampal pathway, a known cholinergic
tract, Kuhar and associates showed that changes in impulse
flow induced via electrical stimulation or pentylenetetrazol
administration (both of which increase impulse flow) or via
lesioning or the administration of pentobarbital (both of which
decrease neuronal traffic) will alter high-affinity transport of
choline into hippocampal synaptosomes. In their studies,
procedures that activated impulse flow increased the maximal
velocity (Vmax) of choline transport, while agents that
stopped neuronal activity decreased Vmax. In neither
situation was the Km changed, a result to be expected since
the concentration of choline outside the neuron (5-10 mM)
normally exceeds the Km for transport (1-5 mM). Recent
evidence, however, suggests that this relationship between
impulse traffic and choline transport does not occur in all
brain areas (e.g., in the striatum, where cholinergic
interneurons abound). In addition, the endogenous
concentration of ACh is implicated in regulating the level of
the transmitter in the brain. Thus, in several studies, an
increase in choline uptake following depolarization of a
preparation has been attributed to the release of endogenous
ACh upon depolarization. Other studies, however, suggest
that this increased choline uptake is not related to ACh
release but rather to an increase in Na-K adenosine
triphosphatase (ATPase) activity.

6. Recently, the high-affinity choline transporter has been

cloned by the Okuda group, who developed an antibody to
map cholinergic neurons.

CHOLINE ACETYLTRANSFERASE
With respect to the cellular localization of ChAT, the highest
activity is found in the interpeduncular nucleus, caudate
nucleus, retina, corneal epithelium, and central spinal roots
(3000-4000 mg ACh synthesized g-1 hour-1). In contrast,
dorsal spinal roots contain only trace amounts of the enzyme,
as does the cerebellum.

Intracellularly, after differential centrifugation in a sucrose

medium, ChAT is found in mammalian brain, predominantly in
the crude mitochondrial fraction. This fraction contains
mitochondria, nerve-ending particles (synaptosomes) with
enclosed synaptic vesicles, and membrane fragments. When
this fraction is subjected to sucrose density gradient
centrifugation, the bulk of the ChAT is associated with nerve-
ending particles. When these synaptosomes are ruptured by
hypo-osmotic shock, synaptosomal cytoplasm can be
separated from synaptic vesicles. In a solution of low ionic
strength, ChAT is adsorbed to membranes and to vesicles; but
in the presence of salts at physiological concentration, the
enzyme is solubilized and remains in the cytoplasm. In vivo,
the enzyme is most likely present in the cytoplasm of the
nerve-ending particle. However, a particulate form of the
enzyme has been found at nerve terminals and, more
recently, at synaptic vesicle membranes. This ChAT, which
when solubilized has the same kinetic characteristics as the
soluble form, occurs in a much lower concentration but with a
higher specific activity. Since kinetic coupling has been
observed between uptake of choline and acetylation, it is
conceivable that this membrane-bound ChAT is the
physiologically relevant form. Support for this contention
derives from three experimental findings: (1) homocholine
cannot be acetylated with purified soluble ChAT but can form
acetylhomocholine when a lysed synaptosomal preparation is
used; (2) choline mustard aziridinium ion, an inhibitor of
choline transport as well as of ChAT, is a much more potent
inhibitor of membrane-bound ChAT than the soluble form of
the enzyme; (3) a monoclonal antibody raised against
Torpedo terminal membranes inhibits ChAT and ACh release.
Regardless of these observations, most cholinergic experts
favor the soluble form of ChAT as the major source of ACh.

A cell-free system of ChAT was first described by

Nachmansohn and Machado in 1943. Since that time, the
enzyme from squid head ganglia, human placenta, Drosophila
melanogaster, Torpedo californica, and brain has been
purified and some of its characteristics have been defined.

When highly purified from rat brain, ChAT has a molecular

weight of 67-75 kDa: it has an apparent Michaelis constant
(Km) for choline of 7.5 10-4 M and for acetyl CoA of 1.0
10-5 M. Recent estimates suggest an equilibrium constant of
13. The enzyme is activated by chloride and inhibited by
sulfhydryl reagents. A variety of studies on the substrate
specificity of the enzyme indicate that various acyl derivatives
of both CoA and ethanolamine can be utilized. The major gap
in our knowledge of ChAT is that as yet we do not know of
any useful (i.e., potent and specific) direct inhibitor.
Styrylpyridine derivatives inhibit it but suffer from the fact
that they are light-sensitive, somewhat insoluble, and possess
varying degrees of anticholinesterase activity. Hemicholinium
inhibits the synthesis of ACh indirectly by preventing the
transport of choline across cell membranes. The genomic
aspects of ChAT have been reviewed by Wu and Hersh
(1994). Dobransky et al. (2001) have described the results of
phosphorylation of the enzyme.

ACETYLCHOLINESTERASE
Everybody agrees that ACh is hydrolyzed by cholinesterases,
but nobody is sure just how many cholinesterases exist in the
body. All cholinesterases will hydrolyze not only ACh but other
esters as well. Conversely, hydrolytic enzymes such as
arylesterases, trypsin, and chymotrypsin will not hydrolyze
choline esters. The problem in determining the number of
cholinesterases that exist is that different species and organs
sometimes exhibit maximal activity with different substrates.
For our purposes, we will divide the enzymes into two rigidly
defined classes: acetylcholinesterase (also called "true" or
specific cholinesterase) and butyrylcholinesterase (also called
"pseudo" or nonspecific cholinesterase; the term
propionylcholinesterase is sometimes used since in some
tissues propionylcholine is hydrolyzed more rapidly than
butyrylcholine). Although their molecular forms are similar,
the two enzymes are distinct entities, encoded by specific
genes. Current evidence suggests that in lower forms
butyrylcholinesterase predominates, gradually giving way to
acetylcholinesterase with evolution. When distinguishing
between the two types of cholinesterase, at least two criteria
should be used because of the aforementioned species or
organ variation.

The first criterion is the optimum substrate.

Acetylcholinesterase hydrolyzes ACh faster than
butyrylcholine, propionylcholine, or tributyrin; the reverse is
true with butyrylcholinesterase. In addition, acetyl-N-methyl
choline (methacholine) is split only by acetylcholinesterase.
That this criterion is not inviolate and must be used along
with other indices is illustrated by the fact that chicken brain
acetylcholinesterase will hydrolyze acetyl- -methyl choline but
will also hydrolyze propionylcholine faster than ACh. Also, the
beehead enzyme will not hydrolyze either ACh or
butyrylcholine but will split acetyl- -methyl choline.

The second criterion is the substrate concentration versus

activity relationship. Acetylcholinesterase is inhibited by high
concentrations of ACh so that a bell-shaped substrate
concentration curve results. This is observed also when
butyrylcholine or propionylcholine is used. In contrast,
butyrylcholinesterase is not inhibited by high substrate
concentrations so that the usual Michaelis-Menten type of
substrate concentration curve is obtained. The reason for this
difference is that in acetylcholinesterase there is at least a
two-point attachment of substrate to enzyme, whereas with
butyrylcholinesterase the substrate is attached at only one
site.

The type of cholinesterase found in a tissue is often a

reflection of the tissue. This fact is used as a discriminating
index between cholinesterases. In general, neural tissue
contains acetylcholinesterase, while glial cells and nonneural
tissue usually contain butyrylcholinesterase. However, this is a
generalization, and some neural tissue (e.g., autonomic
ganglia) contains both esterases, as do some extraneural
organs (e.g., liver, lung). In the blood, erythrocytes contain
only acetylcholinesterase, while plasma contains
butyrylcholinesterase. However, plasma has primary
substrates varying from species to species. Because of its
ubiquity, cholinesterase activity cannot be used as the sole
indicator of a cholinergic system in the absence of additional
supporting evidence. To generalize on this point, until neuron-
specific, transmitter-degrading enzymes are discovered, it is a
neurochemical commandment that, to delineate a neuronal
tract, one must always assay an enzyme involved in the
synthesis of a neurotransmitter and not one concerned with
catabolism.
A final criterion that may be applied to differentiate between
the esterases is their susceptibility to inhibitors. Thus, the
organophosphorous anticholinesterases, such as diisopropyl
phosphorofluoridate and iso-octamethyl pyrophosphoramide,
are more potent inhibitors of butyrylcholinesterase whereas
WIN8077 (Ambinonium) is about 2000 times better an
inhibitor of acetylcholinesterase. The compound BW284C51
[1,5-bis-(4-allyldimethylammoniumphenyl)penta-3-one
dibromide] is presumed to be a specific reversible inhibitor of
acetylcholinesterase.

In discussing the various techniques used to classify the

cholinesterases, we touched on some aspects of the molecular
properties of the enzymes. Because very little work has been
done on butyrylcholinesterase and no physiological role for
this enzyme (or enzymes) has been demonstrated, we will
focus our attention on acetylcholinesterase. In sucrose
homogenates of mammalian brain subjected to differential
centrifugation, acetylcholinesterase is found in both the
mitochondrial and the microsomal fractions. The latter,
consisting of endoplasmic reticulum and plasma cell
membranes, exhibit a higher specific activity. This localization
of the enzyme is supported by electron microscopic and
histochemical studies that fix the activity at membranes of all
kinds in both the CNS and the peripheral nervous system.

Both cholinesterases occur in several molecular forms, which

are classified as either globular or asymmetric. The globular
forms, G1, G2, and G4 (Fig. 7-2), exist as monomers, dimers,
and tetramers. Elongated forms, which contain as many as 12
subunits and are attached to a collagen tail, are classified as
asymmetric. Regardless of the form, both cholinesterases
occur in a water-soluble and a membrane-bound state.

With its turnover time of 150 milliseconds, equivalent to

hydrolyzing 5000 molecules of ACh per molecule of enzyme
per second, acetylcholinesterase ranks as one of the most
efficient enzymes extant.

With respect to the topography of the enzyme, the twin-

hatted diagram of the anionic and esteratic sites has been
reproduced countless times and need not be presented again
here. However, some discussion is in order since this was the
first enzyme to be dissected at the molecular level. For this
initiation into molecular biology, we owe a debt of gratitude to
Nachmansohn and colleagues, particularly Wilson.

The active center of acetylcholinesterase has two main

subsites. The first is an anionic site, which attracts the
positive charge in ACh; the second, about 5 A distant, is an
esteratic site, which binds the carbonyl carbon atom of ACh.
Current information suggests that the anionic site contains at
least one carboxyl group, possibly from glutamate, and the
esteratic site involves a histidine residue adjacent to serine.
The overall reaction is written as follows:

Information on the architecture of the active center has been

derived not only from kinetic studies using model compounds
but also from a group of inhibitors known as the
anticholinesterases. (The pharmacology of these agents will
be discussed below, see ACh in Disease States.) The
anticholinesterases are classified as reversible and irreversible
inhibitors of the enzyme. Like ACh, both types of inhibitor
acylate the enzyme at the esteratic site. However, in contrast
to ACh or to a reversible inhibitor such as physostigmine, the
irreversible inhibitors, which are organophosphorous
compounds, irreversibly phosphorylate the esteratic site. This
phosphorylation occurs on the hydroxyl group of serine when
diisopropylfluorophosphonate (DFP) is incubated with purified
acetylcholinesterase. Although the organophosphorous agents
(referred to as nerve gases, although they are actually oils)
are classed as irreversible anticholinesterases, there is a slow
detachment of the compounds from the enzyme. Wilson
observed that hydroxylamine speeded up this dissociation and
regenerated active enzyme. He then designed a nucleophilic
agent with a spatial structure that would fit the active center
of acetylcholinesterase and produced a compound that is very
active in displacing the inhibitor. This is 2-pyridine aldoxime
methiodide (PAM), which has been used with moderate
success in treating poisoning from the organophosphorous
compounds in insecticides. PAM, with its quaternary
ammonium group, does not penetrate the blood-brain barrier
well enough to overcome central actions of the
anticholinesterase. For this reason, atropine is usually used as
an antidote along with PAM. Atropine blocks the effect of ACh
at neuroeffector sites and has nothing to do with
acetylcholinesterase. Currently, the oxime of choice in
dephosphorylating organophosphorous compound-inactivated
cholinesterase is HI-6 (1 [2-
{hydroxyimino}methylpyridinium]-2-[4-
carboxyamidopyridinium] dimethyl ether dichloride).

A curious observation (useful for cocktail party conversation)

is that heroin is deacetylated to 6-monoacetylmorphine by
both serum butyrylcholinesterase and erythrocyte
acetylcholinesterase, but only acetylcholinesterase will
hydrolyze the monoacetylmorphine to morphine.
Interestingly, brain acetylcholinesterase will not deacetylate
heroin. Another interesting and potentially useful observation
is that serum butyrylcholinesterase will hydrolyze and thus
inactivate cocaine. It has been suggested that mutants of the
enzyme can be prepared that will inactivate the drug at a
faster rate, to be useful in cocaine overdosage. Reviews of the
cholinesterases by Taylor et al. and by Soreq and Seidman are
recommended.
 Figure 7-2. Schematic model of the molecular
polymorphism of acetylcholinesterase (AChE) and
butyrylcholinesterase (BChE). Open circles
designate catalytic subunits. Disulfide bonds are
indicated by S-S. Hydrophilic forms are G1, G2,
and G4. The asymmetric A12 forms have three
hydrophilic G4 heads linked to a collagen tail via
disulfide bonds. The G4 amphiphilic forms of brain
are anchored into a phospholipid membrane
through a 20 kDa anchor. The G2 amphiphilic
forms of erythrocytes have a glycolipid anchor. In
Torpedo, AChE hydrophilic forms and amphiphilic
G2 forms are produced by alternative splicing so
that the proteins are identical at 535 amino acids
but non-identical at their C termini (From
Chatonnet and Lockridge, 1989.)

UPTAKE, SYNTHESIS, AND RELEASE OF ACH

Superior Cervical Ganglion, Brain, and

Skeletal Muscle
To date, the only major, thorough studies of ACh turnover in
nervous tissue were done originally by MacIntosh and
colleagues and subsequently by Collier using the superior
cervical ganglion of the cat. Using one ganglion to assay the
resting level of ACh and perfusing the contralateral organ,
these investigators determined the amount of transmitter
synthesized and released under a variety of experimental
conditions, including electrical stimulation, addition of an
anticholinesterase to the perfusion fluid, and perfusion media
of varying ionic composition. Their results may be
summarized as follows:
1. During stimulation, ACh turns over at a rate of 8%-10% of
its resting content every minute (i.e., about 24-30
ng/minute). At rest, the turnover rate is about 0.5 ng/minute.
Since there is no change in the ACh content of the ganglion
during stimulation at physiological frequencies, it is evident
that electrical stimulation not only releases the transmitter
but also stimulates its synthesis.

2. Choline is the rate-limiting factor in the synthesis of ACh.

3. In the perfused ganglion, Na+ is necessary for optimum

synthesis and storage and Ca2+ is necessary for release of
the neurotransmitter.

4. Newly synthesized ACh appears to be more readily

released upon nerve stimulation than depot or stored ACh.

5. About half of the choline produced by cholinesterase

activity is reutilized to make new ACh.

6. At least three separate stores of ACh in the ganglion are

inferred from these studies: surplus ACh, considered to be
intracellular, which accumulates only in an eserine-treated
ganglion and is not released by nerve stimulation but is
released by K depolarization; depot ACh, which is released by
nerve impulses and accounts for about 85% of the original
store; and stationary ACh, which constitutes the remaining
15% that is nonreleasable.

7. Choline analogs, such as triethylcholine, homocholine, and

pyrrolcholine, are released by nerve stimulation only after
they are acetylated in the ganglia.

8. Increasing the choline supply in the plasma during

perfusion of the ganglion only transiently increases the
amount of ACh that is releasable with electrical stimulation,
despite accumulation of the transmitter in the ganglion.

9. The compound AH5183 (vesamicol), shown by the Parsons

lab to inhibit ACh transport into synaptic vesicles, ultimately
blocks release of ACh from the stimulated ganglia. This
finding supports the contention that there is vesicular release
of the transmitter in the periphery.

As stated above, this work on the superior cervical ganglion

represents the most complete information on the turnover of
ACh in the nervous system. With respect to the regulation of
ACh turnover in cholinergic terminals of skeletal muscle, Vaca
and Pilar, using chick iris, have elaborated on the work of
Potter with the rat phrenic nerve-diaphragm preparation. With
the iris and the diaphragm preparation, the same relationship
of high-affinity choline uptake, ACh synthesis, and regulation
by endogenous ACh has been demonstrated as previously
described with CNS and autonomic nervous system
preparations.

As noted earlier, ACh is not taken up into cholinergic terminals

by a high-affinity transport system. However, as first
described by Parsons' laboratory, ACh is transported into
synaptic vesicles via a proton-pumping ATPase activity. A
glycosylated ATPase pumps protons out of vesicles and drives
ACh via a separate transporter into vesicles in exchange for
the protons. This uptake is blocked by vesamicol.
Interestingly, this vesicular transporter and ChAT arise from
the same gene locus on chromosome 10.

Brain Slices, Nerve-Ending Particles

(Synaptosomes), and Synaptic Vesicles
In 1939 Mann, Tennenbaum, and Quastel demonstrated the
synthesis and release of ACh in cerebral cortical slices. Since
then, these observations have been repeatedly confirmed but
only moderately extended. The major finding of interest in all
of these studies is that in the usual incubation medium the
level of ACh in the slices reaches a limit and cannot be raised.
In a high K+ medium, the total ACh is increased substantially
because much of it leaks into the medium from the slices. The
experiments again suggest that the intracellular concentration
of the neurotransmitter plays a role in regulating its rate of
synthesis, in addition to the high-affinity uptake system for
choline. This concept of a feedback mechanism is supported
by the findings that the administration of drugs such as
morphine, oxotremorine, and the anticholinesterases at the
most succeed only in doubling the original level of ACh in the
brain. Regardless of the dose of the drug, no higher level can
be obtained.

Much of the current neurochemical work on the release of

ACh from the brain involves the use of nerve-ending particles
(synaptosomes). This preparation, independently developed
by DeRobertis and by Whittaker, is derived from sucrose
density gradient centrifugation of a crude mitochondrial
fraction of brain. Although synaptosomes represent
presynaptic terminals with enclosed vesicles and
mitochondria, some postsynaptic fragments are often
attached to them. A disadvantage of the preparation is its
heterogeneity; the usual synaptosome fraction is a mixture of
cholinergic, noradrenergic, serotonergic, and other terminals.
In addition, as judged by electron microscopy and enzyme
markers, the purity is around 60%; the contaminants may
include glial cells, ribosomes, and membrane fragments that
may be axonal, mitochondrial, or perikaryal. However, the
advantage of these preparations is that they can be isolated
easily and that synaptic vesicles can be collected by hypo-
osmotically shocking the synaptosomes, followed by
centrifugation. With respect to the disposition of ACh in
synaptosomes, roughly half of the transmitter is found in
vesicles and the other half in synaptosomal cytoplasm. The
cytosolic localization could be artifactual, resulting from the
preparation methods.

Following the discovery of these presynaptically localized

vesicles that contained ACh, the conclusion was almost
unavoidable that these organelles are the source of the
quantal release of transmitter as described in the
neurophysiological experiments of Katz and collaborators.
Thus, the obvious interpretation has been that as the nerve is
depolarized Ca2+ enters the terminal, vesicles in apposition
to the terminal fuse with the presynaptic membrane, and ACh
is released into the synaptic cleft to interact with receptors on
the postsynaptic cell to change ion permeability. Synapsin I, a
phosphoprotein that is localized in vesicles, may mediate the
translocation of vesicles to the plasma membrane. Other
vesicular proteins that have been implicated in the exocytotic
process are the synaptotagmins, synaptophysins, and
synaptobrevins (see Chapter 3). Synaptobrevin is of particular
interest in that both tetanus toxin and botulinum toxin type B,
which are zinc endopeptidases, inhibit ACh release by cleaving
it. Synaptotagmin has been implicated as a Ca2+-sensor in
the release process. The subsequent sequence of events is
not clear, but in some fashion the presynaptic membrane is
pinocytotically recaptured and vesicles are resynthesized and
simultaneously or subsequently repleted with ACh. This
endocytotic event is apparently triggered by calcineurin, a
Ca2+-dependent protein phosphatase. Another, though minor,
possibility to explain ACh release comes from the laboratories
of Israel and Dunant. The Israel group isolated a lipoprotein
from presynaptic terminals of the Torpedo electric organ,
which they refer to as a mediatophore. A 15 kDa subunit of
this protein exhibits vacuolar ATPase activity. When the
mediatophore is incorporated into ACh-loaded
proteoliposomes, addition of calcium and the calcium
ionophore A23187 triggers the release of ACh. The Dunant
group transfected mediatophore cDNA into neuroblastoma
cells and restored quantal ACh release. Dunant and Israel
suggest that the vesicles release ACh to the cytosol and that
the mediatophore, acting as a gate around a calcium channel,
will release the transmitter. The action of calcium is
terminated by uptake into synaptic vesicles, where it is
ultimately excreted.

CHOLINERGIC PATHWAYS
The identification of cholinergic synapses in the peripheral
nervous system has been relatively easy, and we have known
for a long time now that ACh is the transmitter at autonomic
ganglia, at parasympathetic postganglionic synapses, and at
the neuromuscular junction. In the CNS, however, until
relatively recently, technical difficulties have limited our
knowledge of cholinergic tracts to the motoneuron collaterals
to Renshaw cells in the spinal cord. With respect to the
aforementioned technical difficulties, the traditional approach
has been to lesion a suspected tract and then assay for ACh,
ChAT, or high-affinity choline uptake at the presumed terminal
area. Problems with lesioning include making discrete, well-
defined lesions and interrupting fibers of passage. This latter
problem is illustrated by the discovery that a habenula-
interpeduncular nucleus projection that, based on lesioning of
the habenula, was always described as a cholinergic pathway
is not: it turned out that what was lesioned were cholinergic
fibers that passed through the habenula. Thus, although the
interpeduncular nucleus has the highest choline uptake and
ChAT activity of any area in the brain, the origin of this
innervation remains largely unknown. A quantum leap in
technology for tracing tracts in the CNS has occurred in the
past several years. Through the use of histochemical
techniques (originally developed by Koelle and co-workers)
that stain for regenerated acetylcholinesterase after DFP
treatment (Butcher, Fibiger), autoradiography with muscarinic
receptor antagonists (Rotter, Kuhar), and
immunohistochemical procedures with antibodies to ChAT
(McGeer, Salvaterra, Cuello, Wainer), a clear picture of
cholinergic tracts in the CNS is now emerging. The well-
documented tracts are depicted in Figure 7-3. There is
additional information that in the striatum and the nucleus
accumbens septi, only cholinergic interneurons are found.
Also, intrinsic cholinergic neurons have been reported to exist
in the cerebral cortex, colocalized with vasoactive intestinal
polypeptide and often in close proximity to blood vessels.

With respect to other neurotransmitter functions, ACh may

participate in circuits involved with pain reception. Thus, the
findings that nettles (Urtica dioica) contain ACh and
histamine, that high concentrations of ACh injected into the
brachial artery of humans result in intense pain, and that ACh
applied to a blister produces a brief but severe pain indicate a
relationship between ACh and pain. That ACh may act as a
sensory transmitter in thermal receptors, taste fiber endings,
and chemoreceptors has also been suggested, based on the
excitatory activity of the compound on these sensory nerve
endings.

Figure 7-3. Schematic representation of the

major cholinergic systems in the mammalian
brain. Central cholinergic neurons exhibit two
basic organizational schemata: local circuit cells
(i.e., those that morphologically are arrayed
wholly within the neural structure in which they
are found), exemplified by the interneurons of the
caudate-putamen nucleus, nucleus accumbens,
olfactory tubercle, and Islands of Calleja complex
(ICj), and projection neurons (i.e., those that
connect two or more different regions). Of the
cholinergic projection neurons that interconnect
central structures, two major subconstellations
have been identified: the basal forebrain
cholinergic complex composed of choline
acetyltransferase (ChAT)-positive neurons in the
medial septal nucleus (ms), diagonal band nuclei
 medial septal nucleus (ms), diagonal band nuclei
(td), substantia innominata (si), magnocellular
preoptic field (poma), and nucleus basalis (bas)
and projecting to the entire nonstriatal
telencephalon and the
pontomesencephalotegmental cholinergic complex
composed of ChAT-immunoreactive cells in the
pendunuclopontine (tpp) and laterodorsal (dltn)
tegmental nuclei and ascending to the thalamus
and other diencephalic loci and descending to the
pontine and medullary reticular formations (Rt),
deep cerebellar (DeC) and vestibular (Ve) nuclei,
and cranial nerve nuclei. Not shown are the
somatic and parasympathetic cholinergic neurons
of cranial nerves III-VII and IX-XII and the
cholinergic and motor and autonomic neurons
of the spinal cord. amyg, amygdada; ant cg,
anterior cingulate cortex; CrN, dorsal cranial nerve
nuclei; diencep, diencephalon; DR, dorsal raphe
nucleus; ento, entorhinal cortex; frontal, frontal
cortex; IP, interpeduncular nucleus; ins, insular
cortex; LC, locus ceruleus; LR, lateral reticular
nucleus; olfact, olfactory; pir, piriform cortex; PN,
pontine nuclei; pr, perirhinal cortex; parietal,
parietal cortex; post cg, posterior cingulate
cortex; SN, substantia nigra; Sp5, spinal nucleus
of cranial nerve V. (From Butcher and Woolf, 1986,
and Woolf and Butcher, 1989.)

CELLULAR EFFECTS
A variety of actions of ACh that may be viewed as cellular
effects rather than neurotransmitter activity have been
described. These include ciliary movement in the gill plates of
Mytilus edulis, ciliary motility of mammalian respiratory and
esophageal tracts, water resorption and photosynthesis in
plants, a hyperpolarizing effect on atrial muscle, limb
regeneration in salamanders, protein production in the silk
gland of spiders, induction of sporulation in the fungus
Trichoderma, protoplasmic streaming in slime molds, and
photic control of circadian rhythms and seasonal reproductive
cycles.

There are also a variety of tissues and organisms, such as

human placenta, immune cells, Lactobacillus plantarum,
Trypanosoma rhodesiense, and the fungus Claviceps purpurea
in which ACh is found but nothing is known of its action. One
of the most interesting situations is the corneal epithelium,
which contains the highest concentrations of ACh of any
tissue in the body. ACh may be involved in sodium transport
in this tissue.

All of the activities of ACh and its localization in nonnervous

tissue that we have noted above suggest that this agent may
be a hormone as well as a neurotransmitter. All known
neurotransmitters may possess this dual function. These two
activities have already been shown to occur with the biogenic
amines. Even when a certified neurotransmitter is found in
nervous tissue, its action may satisfy the criteria for defining
a hormone or a modulator rather than the currently strict
criteria for a "classic" neurotransmitter.

CHOLINERGIC RECEPTORS
As noted in Chapter 4, cholinergic receptors fall into two
classes, muscarinic (Table 7-1) and nicotinic (Table 7-2). At
last count, five muscarinic receptors (M1-M5) had been
cloned. All of them exhibit a slow response time (100-250
milliseconds), are coupled to G proteins, and either act
directly on ion channels or are linked to a variety of second-
messenger systems. M1, M3, and M5 via Gq are coupled to
phosphatidylinositol hydrolysis; M2 and M4 via Gi are coupled
to cyclic adenosine monophosphate (cAMP). When activated,
the final effect can be to open or close K channels, Ca
channels, or Cl channels, depending on the cell type. With this
array of channel activity, therefore, stimulation of muscarinic
receptors will lead to either depolarization or
hyperpolarization. As noted in Table 7-1, second-messenger
systems have been described following activation of the
muscarinic receptors. Knowing the messengers is fine, but it
does not tell us anything about the message; that is, what
the ultimate physiological effect is.

The pharmacological antagonists that have been used to

define three of the muscarinic subtypes are pirenzepine,
which has a high affinity for M1; AFDX-116 and
methoctramine with a high affinity for M2; and 4-
diphenylacetoxy-N-methylpiperidine methiodide, which
exhibits the highest affinity for M3. Most antagonists do not
show more than a fivefold selectivity for one subtype over all
other subtypes. The two classic muscarinic antagonists
atropine and quinuclidinylbenzylate do not distinguish the
subtypes but block all equally well. What is needed are
selective muscarinic agonists that will define the various
subtypes. More importantly, it would be a major advance if we
knew whether each subtype subserved a specific function,
such as bradycardia or smooth muscle contractibility. If this
were the case (and it probably is), it could lead to the
development of specific drugs devoid of side effects.

Until relatively recently, the identification of nicotinic

cholinergic receptors in the CNS was an enigma. Using labeled
-bungarotoxin, nicotine, mecamylamine, or dihydro- -
erythroidine, each investigation yielded mystifying results in
which the antagonist could not be easily displaced by ACh or
by unlabeled ganglionic or neuromuscular antagonists but on
occasion was displaced by muscarinic agonists and
antagonists. A major problem has been the low density of
nicotinic compared to muscarinic receptors in the brain.

Conversely, much is known about the properties of the

nicotinic cholinergic receptor of Torpedo and Electrophorus
electricus organs. This reflects the abundance of the receptor
in this tissue and the availability of two snake toxins, -
bungarotoxin and Naja naja siamensis, that specifically bind
to the receptor and have facilitated its isolation and
purification. In the past 10 years, however, research with
monoclonal antibodies and cDNA has yielded considerable
information about the mammalian nicotinic receptors (Table
7-2). At least nine different functional receptors have been
identified and can be tentatively differentiated by CNS,
ganglionic, and muscle types as well as pre- and postsynaptic
localizations in the CNS. Cholinergic nicotinic receptors from
muscle or electric organ contain four different subunits ( , ,
and ), with a stoichiometry of two subunits and each of the
other three. In contrast, neuronal nicotinic receptors contain
only two kinds of subunit ( and ), with the occurring in at
least eight different forms and the in three. When one
considers the variety of combinations of these and
subunits that are theoretically possible, there appears to be a
surfeit of nicotinic receptors in the brain. Current and future
electrophysiological and behavioral studies are attempting to
assign a specific function to each of them. A review of this
problem by Klink et al. and a review of neuronal nicotinic
receptors by Clementi et al. are recommended.

Meanwhile, considerable attention is being devoted to nicotine

since tobacco smoking has been shown to reduce the
incidence of Parkinson's disease and Alzheimer's disease.
Thus, nicotine patches have been shown to increase cognition
in Alzheimer patients and 7 nicotine receptor agonists
specifically appear to be involved in regulating neuronal
growth during development (Reviewed by Picciotto et al.,
2000). The 4 2 receptor has been implicated in learning
deficiencies. Finally, nicotine antagonists such as dihydro- -
erythroidine impair working memory. Obviously,
pharmaceutical chemists have more than a casual interest in
pursuing this topic. A currently promising compound is
galantamine, an anticholinesterase that sensitizes nicotinic
receptors.

As noted above, the nicotinic ACh receptor from electric

organs is a pentameric integral membrane protein composed
of four glycosylated polypeptide chains designated , , , and
, with a stoichiometry of two subunits to one each of the
other three (Fig. 7-4). All subunits traverse the membrane
and, when viewed face on, resemble a five-petal rosette with
a central pit. ACh binds to the subunits and produces a
conformational change in the channel that selectively allows
cations rather than anions with a diameter of about 0.65 nm
to pass through. The molecular weight of the receptor
complex is about 255,000. Desensitization of the receptor
increases when it is phosphorylated by cAMP protein kinase
(protein kinase A), protein kinase C, or tyrosine kinase.

An unexpected windfall resulting from the isolation of the

neuromuscular nicotinic cholinergic receptor has been the
demonstration that when this lipoprotein, isolated from
Electrophorus, is injected into rabbits, all signs of myasthenia
gravis appear. This finding, coupled with autoradiographic
studies of muscle biopsies of myasthenics using labeled -
bungarotoxin as a tag where a marked deficiency of receptors
is observed, has led to a better understanding of the disease
at the molecular level. It is now clear that myasthenia gravis
is an autoimmune disease in which a circulating antibody
appears to be involved in an increased rate of degradation
and damage, as well as antagonism of the ACh receptor.
Since antibodies produced from Electrophorus or Torpedo ACh
receptor result in myasthenic signs in rabbits, guinea pigs,
and monkeys, this work also carries the implication that the
ACh receptor is a phylogenetically conserved protein that can
exhibit immunological cross-reactivity.

Figure 7-4. Model of a acetylcholine synapse

illustrating the presynaptic and postsynaptic
molecular entities involved in the synthesis,
storage, release, reuptake, and signaling of
acetylcholine. Choline is transported into the
presynaptic terminal by an active uptake
mechanism and converted to acetylcholine by a
single enzymatic step. The acetyl coenzyme A
required for acetylcholine synthesis is provided by
presynaptic mitochondria. Muscarinic M2
autoreceptors in the presynaptic terminal
modulate the release of acetylcholine. In contrast
to monoamine synapses, the plasma membrane
transporter of an acetylcholine synapse does not
return the neurotransmitter to the presynaptic
terminal but rather recycles precursor choline.
Acetylcholine is metabolized by acetylcholine
esterase (AchE) present on both pre- and
postsynaptic membranes, which serves to
terminate its action. Both G protein-coupled
muscarinic receptors and ligand-gated ion
channels (nicotinic receptors) may be present as
depicted on the postsynaptic cell. Sites of drug
action:

Site 1: ACh synthesis can be blocked by styryl

pyridine derivatives such as NVP.

Site 2: ACh transport into vesicles is blocked by

vesamicol (AH5183).

Site 3: Release is promoted by -bungarotoxin,

black widow spider venom, and La3+. Release is
blocked by botulinum toxin, cytochalasin B,
collagenase pretreatment, and Mg2+.

Site 4: Acetylcholinesterase is inhibited reversibly

by physostigmine (eserine) or irreversibly by DFP,
or soman.

Site 5: Choline uptake competitive blockers

include hemicholinium-3, troxypyrrolium tosylate,
or AF64A (noncompetitive).

Site 6: Presynaptic muscarinic receptors may be

blocked by AFDX-116 (an M2 antagonist),
atropine, or quinuclidinyl benzilate. Muscarinic
agonists (e.g., oxotremorine) will inhibit the
evoked release of ACh by acting on these
receptors.

Site 7: Postsynaptic receptors are activated by

cholinomimetic drugs and anticholinesterases.
Nicotinic receptors, at least in the peripheral
nervous system, are blocked by rabies virus,
curare, hexamethonium, or dihydro- -
erythroidine; n-methylcarbamylcholine and
dimethylphenyl piperazinium are nicotinic
agonists. Muscarinic receptors are blocked by
atropine, pirenzepine, and quinuclidinyl benzilate.

AC, adenylyl cyclase; AChE, acetylcholine

esterase; ChAT, choline acetyltransferase; CT,
plasma membrane choline transporter; DAG,
diacylglycerol; IP3, inositol triphosphate; M,
muscarinic receptors; nAChR, nicotinic
acetylcholine receptors; PLC, phospholipase C;
VAT, vesicular acetylcholine transporter (not yet
isolated). (Modified from Nestler et al., Molecular
 Neuropharmacology)

ACH IN DISEASE STATES

Aside from myasthenia gravis and other autoimmune
diseases, such as the Lambert-Eaton myasthenic syndrome (a
presynaptic problem involving diminished release of ACh), the
role of ACh in nervous system dysfunction is unclear.
Certainly, a strong case can be made for familial
dysautonomia, an autosomal recessive condition affecting
Ashkenazi Jews that is diagnosed by a supersensitivity of the
iris to methacholine. Huntington's disease, involving a
degeneration of Golgi type 2 cholinergic interneurons in the
striatum, is partially ameliorated by physostigmine.
Administration of physostigmine to patients with tardive
dyskinesia has produced mixed results.

In Alzheimer's disease, characterized behaviorally by a severe

impairment in cognitive function and neuropathologically by
the appearance of neuritic plaques and neurofibrillary tangles,
a cholinergic dysfunction has been implicated. As discussed
more fully in Chapter 13, this is based on the fact that a
variety of cholinergic abnormalities have been found.
However, efforts to treat patients with choline, lecithin, or
anticholinesterases such as tetrahydroaminoacridine (tacrine)
have met with little success. In animal models with imposed
cognitive deficits, memory has been restored with synthetic
cholinergic agonists such as YM796, DuP996, L-689,660, and
WEB-1881-FU. Since there are no animal models of
Alzheimer's disease, it is still unknown if any of these agents
will produce a clinical benefit in humans. At any rate, the
major focus currently is on the -amyloid peptide, which is
found in the plaques of Alzheimer's disease brain.
Considerable evidence suggests that aberrant proteolytic
processing of this -amyloid protein is associated with
degenerating nerve terminals. Thus, current research is
directed to developing specific protease inhibitors as well as
exploring techniques for delivering neurotrophic factors such
as nerve growth factor (incidentally stimulated by nicotine) to
the brain. Sites of drugs that affect the synthesis, release,
and neurotransmitter activity of ACh are shown in Figure 7-5.
Structures of drugs that affect the cholinergic nervous system
are depicted in Figure 7-6.

Figure 7-5. Sites of drug action at cholinergic

synapses.

Site 1: Acetylcholine (ACh) synthesis can be

blocked by styryl pyridine derivatives such as
naphylvinylpyridine (NVP).

Site 2: ACh transport into vesicles is blocked by

vesamicol (AH5183).

Site 3: Release is promoted by -bungarotoxin,

black widow spider venom, and La3+. Release is
blocked by botulinum toxin, cytochalasin B,
collagenase pretreatment, and Mg2+.

Site 4: Postsynaptic receptors are activated by

cholinometic drugs and anticholinesterases.
Nicotinic receptors, at least in the peripheral
nervous system, are blocked by rabies virus,
curare, hexamethonium, or dihydro- -
erythroidine; n-methylcarbamylcholine and
dimethylphenyl piperazinium are nicotinic
agonists. Muscarinic receptors are blocked by
atropine, pirenzepine, and methoctramine.

Site 5: Presynaptic muscarinic receptors may be

blocked by AFDX-116 (an M2 antagonist),
atropine, or quinuclidinyl benzilate. Muscarinic
 agonists (e.g., oxotremorine) will inhibit the
evoked release of ACh by acting on these
receptors.

Site 6: Acetylcholinesterase is inhibited reversibly

by physostigmine (eserine) or irreversibly by
diisopropylfluorophosphate or sarin.

Site 7: Choline uptake competitive blockers

include hemicholinium-3, troxypyrrolium tosylate,
or AF64A (noncompetitive).

Figure 7-6. Structure of some drugs that affect

the cholinergic nervous system.

SELECTED REFERENCES
Allen, D. D. and Q. R. Smith (2001). Characterization of the
blood-brain barrier choline transporter using the in situ rat
brain perfusion technique. J. Neurochem. 76, 1032.

Benishin, C. G. and P. T. Carroll (1983). Multiple forms of

choline-O-acetyltransferase in mouse and rat brain:
solubilization and characterization. J. Neurochem. 41, 1016.

Bloc, A., E. Bugnard, Y. Dunant, J. Falk-Vairant, M. Israel, F.

Loctin, and E. Roulet (1999). Acetylcholine synthesis and
quantal reconstituted by transfection of mediatophore and
choline acetyltransferase cDNAs. Eur. J. Neurosci. 11, 1523.

Butcher, L. L. and N. J. Woolf (1986). Central cholinergic

systems: synopsis of anatomy and overview of physiology and
pathology. In The Biological Substrates of Alzheimer's Disease
(A. B. Scheibel and A. F. Wechsler, eds.). Academic Press,
New York, pp. 73-86.
Chatonnet, A. and O. Lockridge (1989). Comparison of
butyrylcholinesterase and acetylcholinesterase. Biochem. J.
260, 625.

Clementi, F., D. Fornasari, and C. Gotti (2000). Neuronal

nicotine acetylcholine receptors: from structure to
therapeutics. Trends Pharmacol. Sci. 21, 33.

Collier, B., A. Tandon, M. A. M. Prado, and M. Bachoo (1993).

Storage and release of acetylcholine in a sympathetic
ganglion. Prog. Brain Res. 98, 183.

Cooper, J. R. (1994). Unsolved problems in the cholinergic

nervous system. J. Neurochem. 63, 395.

Cordell, B. (1994). -Amyloid formation as a potential

therapeutic target for Alzheimer's disease. Annu. Rev.
Pharmacol. Toxicol. 34, 69.

Dobransky, T., W. L. Davis, and F. J. Rylett (2001). Functional

characterization of phosphorylation of 69-kDa human 440 by
protein kinase C. J. Biol. Chem. 276, 222.

Eder-Colli, L., P. A. Briand, M. Pellegrini, and Y. Dunant

(1989). A monoclonal antibody raised against plasma
membrane of cholinergic nerve terminals of the Torpedo
inhibits choline acetyltransferase activity and acetylcholine
release. J. Neurochem. 53, 1419.

Eiden, L.E. (1998). The cholinergic gene locus. J. Neurochem.

70, 2227.

Ericson, J. P., H. Varoqui, M. K. H. Schafer, W. Modi, M. F.

Diebler, E. Weihe, J. Rand, L. Eiden, T. I. Bonner, and T. B.
Usdin (1994). Functional identification of a vesicular
acetylcholine transporter and its expression from a cholinergic
gene locus. J. Biol. Chem. 269, 21929.
Flentge, F., K. Venema, T. Koch, and J. Korf (1992). An
enzyme-reactor for electrochemical monitoring of choline and
acetylcholine: applications in high-performance liquid
chromatography, brain tissue, microdialysis, and
cerebrospinal fluid. Anal. Biochem. 204, 305.

Israel, M. and Y. Dunant (1998). Acetylcholine release and the

cholinergic genomic locus. Mol. Neurobiol. 16, 1.

Israel, M. and B. Lesbats (1981). Chemiluminescent

determination of acetylcholine and continuous detection of its
release from Torpedo electric organ synapses and
synaptosomes. Neurochem. Int. 3, 81.

Jenden, D. J., M. Roch, and R. A. Booth (1973). Simultaneous

measurement of endogenous and deuterium-labeled tracer
variants of choline and acetylcholine in subpicomole quantities
by gas chromatography/mass spectrometry. Anal. Biochem.
55, 438.

Kawashima, K., T. Hayakawa, Y. Icashima, T. Suzuki, K.

Jufimoto, and H. Oohata (1991). Determination of
acetylcholine release in the striatum of anesthetized rats
using in vivo microdialysis and a radioimmunoassay. J.
Neurochem. 57, 882.

Klink, R., A. de Kerchove d'Exaerde, M. Zoli, and J.P.

Changeux (2001). Molecular and physiological diversity of
nicotinic acetylcholine receptors in the midbrain dopaminergic
nuclei. J. Neurosci. 21, 1452.

Massouli, J., L. Pezzementi, S. Bon, E. Krejci, and F. M.

Vallette (1993). Molecular and cellular biology of
cholinesterases. Prog. Neurobiol. 41, 39.

McCaman, R. E. and J. Stetzler (1977). Radiochemical assay

for ACh: modifications for sub-picomole measurements. J.
Neurochem. 28, 669.

Misawa, H., K. Nakata, J. Matsura, M. Nagao, T. Okuda, and T.

Hagg (2001). Distrubution of the high-affinity choline
transporter in the central nervous system of the rat.
Neuroscience 105, 87.

Okuda, T., T. Haga, T. Kanai, H. Endou, T. Ishihara, and I.

Katsura (2000). Identification and characterization of the
high-affinity choline transporter. Nat. Neurosci. 3, 120.

Parsons, S. M., C. Prior, and I. G. Marshall (1993).

Acetylcholine transport, storage, and release. Int. Rev.
Neurobiol. 35, 279.

Picciotto, M.R., B.J. Calderone, S.L. King, and V. Zachariou

(2000). Nicotinic receptors in the brain: links between
molecular biology and behavior. Neuropsychopharmacology
22, 451.

Potter, L. T. (1970). Synthesis, storage and release of

[14C]acetylcholine in isolated rat diaphragm muscles. J.
Physiol. 206, 145.

Potter, P. E., M. Hadjiconstantinou, J. L. Meek, and N. H. Neff

(1984). Measurement of acetylcholine turnover rate in brain:
an adjunct to a simple HPLC method for choline and
acetylcholine. J. Neurochem. 43, 288.

Ricny, J., J. Coupek, and S. Tucek (1989). Determination of

acetylcholine and choline by flow-injection with immobilized
enzymes and fluorometric or luminometric detection. Anal.
Biochem. 176, 221.

Roghani, A., J. Feldman, S. A. Kohan, A. Shirzadi, C. B.

Gundersen, N. Brecha, and R. H. Edwards (1994). Molecular
cloning of a putative vesicular transporter for acetylcholine.
Proc. Natl. Acad. Sci. U.S.A. 91, 10620.

Rotter, A., N. J. Birdsall, A.S.V. Burgen, P. M. Field, E. C.

Hulme, G. Sastry, and C. Sadarangivivad (1979). Cholinergic
systems in nonnervous tissue. Pharmacol. Rev. 30, 65.

Simon, R., S. Atweh, and M. J. Kuhar (1976). Sodium-

dependent high affinity choline uptake: a regulator step in the
synthesis of acetylcholine. J. Neurochem. 26, 909.

Soreq, H. and S. Seidman (2001). Acetylcholinesterase-new

roles for an old actor. Nat. Rev. Neurosci. 2, 294.

Taylor, P., Z.D. Luo, and S. Camp (2000). Cholinesterase and

Cholinesterase Inhibitors. (E. Giacobini, ed.) Martin Dynitz,
London.

Tucek, S. (1993) Short-term control of the synthesis of

acetylcholine. Prog. Biophys. Mol. Biol. 60, 59.

Vaca, K. and G. Pilar (1979). Mechanisms controlling choline

transport and acetylcholine synthesis in motor nerve
terminals during electrical stimulation. J. Gen. Physiol. 73,
605.

Wainer, B. H., A. I. Levey, E. J. Mutson, and M.-M. Mesulam

(1984). Cholinergic systems in mammalian brain identified
with antibodies against choline acetyltransferase. Neurochem.
Int. 6, 163.

Watling, R. S. (2001). The Sigma-RBI Handbook of Receptor

Classification and Signal Transduction. Sigma-RBI, Natick,
Mass.

Wessler, I., H. Kilbinger, F. Bittinger, and C. J. Kirkpatrick

(2001). The biological role of non-neuronal acetylcholine in
plants and humans. Jpn. J. Pharmacol., 85, 2.

Woolf, N. J. and L. L. Butcher (1989). Cholinergic systems:

synopsis of anatomy and overview of physiology and
pathology. In The Biological Substrates of Alzheimer's Disease
(A. B. Scheibel and A. F. Wechsler, eds.). Academic Press,
New York, pp. 73-86.

Wu, D. and L. B. Hersh (1994). Choline acetyltransferase:

celebrating its fiftieth year. J. Neurochem. 62, 1653.

(+/-) Show / Hide Bibliography

Send Feedback Teton Server (5.0.0) - ©2007 Teton Data Title Updates
Customer Service Systems User
800.901.5494 Send Us Your Comments Responsibilities
Training Center
What's New
8. Norepinephrine and Epinephrine
INTRODUCTION
Norepinephrine (NE) and epinephrine (E) are chemically
catecholamines. The term catecholamine refers generically to
all organic compounds that contain a catechol nucleus (a
benzene ring with two adjacent hydroxyl substituents) and an
amine group (Fig. 8-1). In practice, the term usually means
dihydroxyphenylethylamine (dopamine, DA) and its metabolic
products NE and E. Major advances in the understanding of
the biochemistry, physiology, and pharmacology of these
compounds have come mainly through the development of
sensitive assay techniques and methods for visualizing
catecholamine neurons, their connections and their metabolic
enzymes and receptors in vivo and in vitro.

In 1946, it was demonstrated by von Euler in Sweden, and

shortly thereafter by Holtz in Germany, that mammalian
peripheral sympathetic nerves use NE as a transmitter instead
of E (which is the sympathetic transmitter in the frog). With
only minimal evidence based on tissue content, von Euler
predicted that NE was highly concentrated in the nerve
terminal region from which it was released to act as a
neurotransmitter. This prediction was conclusively
documented some 10 years later with the development of
techniques for visualizing catecholamines in freeze-dried
tissue sections, which helped to define the anatomy of the
peripheral noradrenergic neuron.

Shortly after NE was established as the neurotransmitter

substance of adrenergic nerves in the peripheral nervous
system, Holtz identified it as a normal constituent of
mammalian brain. For some years, however, it was thought
that the presence of NE in the mammalian brain only reflected
vasomotor innervation to the cerebral blood vessels. In 1954,
Vogt demonstrated that NE was not uniformly distributed in
the central nervous system (CNS) and that this nonuniform
distribution did not in any way coincide with the density of
blood vessels found in a given brain region. This regional
localization within the mammalian brain suggested that NE
might subserve some specialized function, perhaps as a
central neurotransmitter. The relative regional distribution of
NE in the brain is quite similar in most mammalian species.

Figure 8-1. Catechol and catecholamine

structure.

MORPHOLOGY OF THE SYMPATHETIC

NEURON
The morphology of the peripheral noradrenergic neuron
became clear following the development by Falck and Hillarp
of the fluorescent histochemical method for the visualization
of catecholamines by condensation with dry formaldehyde
vapor at 60 C-80 C. Under the fluorescent microscope, all
parts of the adrenergic neuron can be visualized (cell bodies,
dendrites, axons, and nerve terminals), but the highest
concentration of NE and the strongest fluorescence are found
in the nerve terminal varicosities. A diffuse, widespread
innervation pattern is characteristic of the sympathetic
nervous system and its noradrenergic nerves. A single neuron
can give rise to nerve terminal branches with lengths on the
order of 10-20 cm, possessing several thousand nerve
terminal varicosities (Fig. 8-2). The localization of
catecholamines and their precursors within morphologically
recognizable microscopic structures by fluorescence and
electron microscopy has been a great advantage for
investigators studying noradrenergic mechanisms where
fluorescence intensity, vesicle contents, and amine content
correlate. The lack of a suitable histochemical technique for
the visualization of acetylcholine has been a serious handicap
by comparison for those interested in cholinergic mechanisms
(see Chapter 7).

DA is also present in the mammalian CNS, and its distribution

differs markedly from that of NE, an early indication that DA
functions as more than a precursor of NE in the CNS (see
Chapter 9). DA is also present in the carotid body and
superior cervical ganglion, where it also likely plays a role
independent of NE. The superior cervical ganglion appears to
have at least three distinct populations of neurons: cholinergic
neurons, noradrenergic neurons, and, small, intensely
fluorescent cells that contain DA but whose functional
significance is unclear.

The endogenous occurrence of E in the mammalian CNS at

relatively low levels (approximately 5%-10% by bioassay of
the NE content) was reported in the early 1960s. Many
investigators suggested that these original estimates are
subject to error and in the past have discounted the
importance of E in the mammalian brain. However, its
presence has now been documented by more sophisticated
analytical techniques, such as gas chromatography-mass
spectrometry and liquid chromatography coupled with
electrochemical detection and confirmed by
immunohistochemical techniques.

The detailed topographical survey of brain catecholamines at

different levels of organization within the CNS has provided a
framework for organizing and conducting logical experiments
concerning the possible functions of these amines. The
anatomy, biochemistry, and pharmacology of CNS NE and E
systems are discussed in detail in the latter half of this
chapter.

Figure 8-2. Schematic drawing of a noradrenergic

 neuron in the peripheral autonomic nervous
system. Examples of innervated structures are
illustrated. (Modified from Dahlstrom and
Carlsson, 1986.)

LIFE CYCLE OF THE CATECHOLAMINES

Biosynthesis
Catecholamines are formed in the brain, chromaffin cells,
sympathetic nerves, and sympathetic ganglia from their
amino acid precursor tyrosine by a sequence of enzymatic
steps first postulated by Blaschko in 1939. This pathway was
confirmed in 1964 by Nagatsu and co-workers, who
demonstrated that tyrosine hydroxylase (TH) converts L-
tyrosine to 3,4-dihydroxyphenylalanine (DOPA). Tyrosine is
normally present in the circulation at a concentration of about
5 to 8 10-5 M. It is taken up from the bloodstream and
concentrated within the brain and presumably in other
sympathetically innervated tissue by an active transport
mechanism. Once inside the peripheral neuron, tyrosine
undergoes a series of chemical transformations, resulting in
the formation of NE or, in the brain or chromaffin cell, NE, DA,
or E, depending on the availability of the required
downstream synthetic enzymes (see Fig. 8-3). Both
phenylalanine and tyrosine are normal constituents of the
mammalian brain, present in a free form at a concentration of
about 5 10-5 M. However, NE biosynthesis is usually
considered to begin with tyrosine, which represents a branch
point for many important biosynthetic processes in animal
tissues. The percentage of tyrosine used for catecholamine
biosynthesis as opposed to other biochemical pathways is
minimal ( 2%).
 Figure 8-3. Primary and alternative pathways in
the formation of catecholamine: (1) tyrosine
hydroxylase; (2) aromatic amino acid
decarboxylase; (3) dopamine- -hydroxylase; (4)
phenylethanolamine-N-methyltransferase; (5)
nonspecific N-methyltransferase in lung and
folate-dependent N-methyltransferase in brain;
(6) catechol-forming enzyme.

Tyrosine Hydroxylase
The first enzyme in the biosynthetic pathway, TH, was the last
enzyme in this series of reactions to be identified. It was
demonstrated by Udenfriend and colleagues in 1964, and its
properties have been reviewed repeatedly. It is present in the
adrenal medulla, brain, and all sympathetically innervated
tissues as a unique constituent of catecholamine-containing
neurons and chromaffin cells. The enzyme is stereospecific;
requires molecular O2, Fe2+, and a tetrahydropteridine
cofactor; and shows a fairly high degree of substrate
specificity for L-tyrosine and, to a smaller extent, L-
phenylalanine. The single human gene for TH has been cloned
and found to encode multiple mRNAs that are heterogeneous
at the 5 end of the coding region. The functional significance
of the varient messages remains to be determined.

Tyrosine hydroxylation is the rate-limiting step in the

biosynthesis of NE in the peripheral nervous system and of NE
and DA in the brain. In most sympathetically innervated
tissues, including the brain, the activity of DOPA
decarboxylase and that of dopamine- -hydroxylase have a
magnitude 100-1000 times that of TH. Further proof that this
enzyme is the rate-limiting step in catecholamine biosynthesis
is that pharmacological intervention at this step reduces NE
biosynthesis while blockade of the last two steps in the
synthesis of NE does not. Inhibitors of TH markedly reduce
endogenous NE and DA in the brain and NE in the heart,
spleen, and other sympathetically innervated tissues.
Effective inhibitors of this enzymatic step can be categorized
into four main groups: (1) amino acid analogues, (2) catechol
derivatives, (3) tropolones, and (4) selective iron chelators.
Some effective amino acid analogues include -methyl-p-
tyrosine and its ester, -methyl-3-iodotyrosine, 3-
iodotyrosine, and -methyl-5-hydroxytryptophan. In general,
-methyl-amino acids are more potent than the unmethylated
analogues and a marked increase in activity in the case of the
tyrosine analogues can also be produced by substituting a
halogen at the 3 position of the benzene ring. Most of the
agents in this category act as competitive inhibitors of the
substrate tyrosine. -methyl-p-tyrosine and its methyl ester
have been the inhibitors most widely used to demonstrate the
effects of exercise, stress, and various drugs on the turnover
of catecholamines and to lower NE formation in patients with
pheochromocytoma and malignant hypertension.

Dihydropteridine Reductase
Although not directly involved in catecholamine biosynthesis,
dihydropteridine reductase is intimately linked to the TH step.
This enzyme catalyzes the reduction of the quinonoid
dihydropterin, which has been oxidized during the
hydroxylation of tyrosine to DOPA. Since reduced pteridines
are essential for tyrosine hydroxylation, alterations in the
activity of dihydropteridine reductase affect the activity of TH.
Dihydropteridines with amine substitution at positions 2 and 4
are effective inhibitors of this enzyme, while folic acid
antagonists such as aminopterin and methotrexate are
relatively ineffective. The distribution of dihydropteridine
reductase is quite widespread, the highest activity being
found in the liver, brain, and adrenal gland. The distribution of
this enzyme activity in the brain extends well beyond
catecholamine or serotonin innervation, suggesting that
reduced pterins most likely participate in other reactions
besides the hydroxylation of tyrosine and tryptophan. In fact,
reduced pteridines are critical for nitric oxide (NO)
synthetase.

Dihydroxyphenylalanine Decarboxylase
The second enzyme involved in catecholamine biosynthesis is
DOPA-decarboxylase, which was actually the first
catecholamine-synthesis enzyme to be discovered. Although
originally believed to remove carboxyl groups only from L-
DOPA, a study of purified enzyme preparations and specific
inhibitors demonstrated that this DOPA-decarboxylase acts on
all naturally occurring aromatic L-amino acids, including
histidine, tyrosine, tryptophan, and phenylalanine as well as
both DOPA and 5-hydroxytryptophan. Therefore, this enzyme
is more appropriately referred to as "L-aromatic amino acid
decarboxylase." There is no appreciable binding of this
enzyme to particles within the cell since, when tissues are
disrupted and the resultant homogenates centrifuged at high
speeds, the decarboxylase activity remains associated largely
with the supernatant fraction. The exception to this is in the
brain, where some of the decarboxylase activity is associated
with synaptosomes. Since synaptosomes are in essence
pinched-off nerve endings, however, they would be expected
to retain entrapped cytoplasm as well as other intracellular
organelles. The DOPA-decarboxylase found in synaptosomal
preparations is thought to be present in the entrapped
cytoplasm. DOPA-decarboxylase is, relative to other enzymes
in the biosynthetic pathway for NE formation, very active and
requires pyridoxal phosphate (vitamin B6) as a cofactor. The
apparent Km value for this enzyme is 4 10-4 M. The high
activity of this enzyme may explain why it has been difficult
to detect endogenous DOPA in sympathetically innervated
tissue and brain. It is rather ubiquitous in nature, occurring in
the cytoplasm of most tissues, including the liver, stomach,
brain, and kidney in high levels, suggesting that its function in
metabolism is not limited solely to catecholamine
biosynthesis. Although decarboxylase activity can be reduced
by vitamin B6 deficiency in animals, this does not usually
result in significant reduction of tissue catecholamines,
although it appears to interfere with the rate of repletion of
adrenal catecholamines after insulin depletion. In addition,
potent decarboxylase inhibitors have very little effect on
endogenous levels of NE in tissue. However, these inhibitors
have been useful as pharmacological tools (e.g., DOPA
accumulation following administration of a decarboxylase
inhibitor as an in vivo index of tyrosine hydroxylation).

Dopamine- -Hydroxylase
Although it has been known for many years that the brain,
sympathetically innervated tissue, sympathetic ganglia, and
adrenal medulla can transform DA into NE, it was not until
1960 that the enzyme responsible for this conversion was
isolated from the adrenal medulla. This enzyme, called
dopamine- -hydroxylase, is, like TH, a mixed-function
oxidase. It requires molecular oxygen and utilizes ascorbic
acid as a cofactor. Its Km for its substrate DA is about 5 10-
3 M. Dicarboxylic acids such as fumaric acid are not absolute
requirements, but they stimulate the reaction. Dopamine- -
hydroxylase is a Cu2+-containing protein, with about 2 moles
of cupric ion per 1 mole of enzyme, associated with the
particulate fraction from the heart, brain, sympathetic nerve,
and adrenal medulla, localized primarily in the membrane of
the amine storage granules. The mRNA is expressed only in
noradrenergic neurons or adrenal chromaffin tissue.
Dopamine- -hydroxylase does not show a high degree of
substrate specificity and acts in vitro on a variety of
substrates besides DA, oxidizing almost any phenylethylamine
to its corresponding phenylethanolamine (i.e., tyramine to
octopamine, -methyldopamine to -methylnorepinephrine).
A number of the resultant structurally analogous metabolites
can replace NE at the noradrenergic nerve endings and
function as "false" neurotransmitters.

Dopamine- -hydroxylase can be inhibited by a variety of

compounds, including the copper chelators: D-cysteine and L-
cysteine, glutathione, mercaptoethanol, and coenzyme A.
Inhibition can be reversed by addition of N-ethylmaleimide,
which reacts with sulfydryl groups amd interferes with the
chelating properties of these substances. Copper-chelating
agents such as diethyldithiocarbamate(disulfuram) and bis(1-
methyl-4-homopiperazinyl-thiocarbonyl)-disulfide (FLA-63)
have proved to be effective inhibitors both in vivo and in vitro.
Thus, it has been possible to treat animals with these agents
and produce a reduction in brain NE and an elevation of brain
DA.

Dopamine- -hydroxylase purified from the bovine adrenal

medulla will produce a specific antibody to the enzyme that
will inactivate bovine dopamine- -hydroxylase but will not
cross-react with either DOPA-decarboxylase or TH. However,
anti-bovine dopamine- -hydroxylase will cross-react with
human, guinea pig, and dog dopamine- -hydroxylase,
indicating that the enzymes from these various sources are
probably structurally related. By coupling immunochemical
techniques with fluorescence and electron microscopy, this
antibody has already proved useful in the localization of the
enzyme in intact peripheral and central tissue. Dopamine- -
hydroxylase has been cloned, and further studies on its
molecular structure and expression should yield interesting
information.

Phenylethanolamine-N-Methyltransferase
In the adrenal medulla, NE is N-methylated by the enzyme
phenylethanolamine-N-methyltransferase (PNMT) to form E.
This enzyme is largely restricted to the adrenal medulla,
although low levels of activity exist in heart and mammalian
brain. Like the decarboxylase, this enzyme appears in the
supernatant of homogenates. Demonstration of activity
requires the presence of the methyl donor S-
adenosylmethionine. Interest in the biosynthetic pathway for
catecholamines has also led to the cloning of a single PNMT
gene with three exons. The transcript is present in the
adrenal medulla, heart, and brain stem; and PNMT is found in
these tissues. Regulation of this enzyme in the brain has not
been extensively studied, but glucocorticoids are known to
regulate the activity of PNMT in the adrenal gland.

Synthesis Regulation
It has been known for a long time that the degree of
sympathetic activity does not influence the endogenous levels
of tissue NE, and it has been speculated that there must be
some homeostatic mechanism whereby the level of
transmitter is maintained relatively constant in the
sympathetic nerve endings despite the additional losses
assumed to occur during enhanced sympathetic activity.

More than 40 years ago, von Euler hypothesized, on the basis

of experiments carried out in the adrenal medulla, that during
periods of increased functional activity, the sympathetic
neuron must also increase the synthesis of its transmitter
substance NE to meet the increased demands placed on the
neuron. If the sympathetic neuron had the ability to increase
transmitter synthesis, this would enable the neuron to
maintain a constant steady-state level of transmitter despite
substantial changes in transmitter utilization. Some years
later, experiments carried out by several laboratories on
peripheral sympathetically innervated tissues as well as brain
directly demonstrated that this was, in fact, the case.
Electrical stimulation of sympathetic nerves, median forebrain
bundle, or locus ceruleus both in vivo and in vitro resulted in
increased formation of NE in the tissues innervated by these
NE neurons. Further studies demonstrated that the observed
acceleration of NE biosynthesis produced by enhanced
noradreneregic activity was due to an increase in the activity
of the rate-limiting enzyme involved in catecholamine
biosynthesis, TH. Emphasis then shifted toward defining the
mechanisms by which impulse flow changed enzyme activity.
It is now well appreciated that the function of TH is
determined by two major factors: changes in enzyme activity
(the rate at which the enzyme converts the precursor into its
product) and changes in the amount of enzyme protein. One
major determinant of TH activity is it state of
phosphorylation, which occurs at four different serine sites at
the N terminus of the TH protein. These four serine residues
are differentially phosphorylated by various kinases. Protein
kinase C, (cAMP)-dependent protein kinase (protein kinase
A), and calcium-calmodulin-dependent kinase phosphorylate
TH, producing kinetic activation of the enzyme; the question
of which of these mechanisms is operative physiologically
remains uncertain. The most recent studies favor the
involvement of Ca2+-calmodulin-dependent phosphorylation
in the impulse-dependent activation process (Fig. 8-4). 2-
Adrenergic or D2 dopaminergic autoreceptor-mediated
reduction of TH may also be achieved via regulation of the
cAMP or Ca2+-calmodulin-dependent protein phosphatases.

A second means of regulating the activity of the enzyme is

through end-product inhibition: catecholamines can inhibit the
activity of TH through competition for a required pterin
cofactor for the enzyme. Finally, the amount of reduced
pteridine cofactor (tetrahydobiopterin, BH4) can also influence
TH activity because the levels of the cofactor are not
saturating under basal conditions.
In general, very rapid, short-term upregulation of TH occurs
primarily by these posttranslational changes, whereas longer-
term changes in TH occur through transcriptional upregulation
of the TH gene. However the degree to which increases in
catecholamine synthesis depend on de novo synthesis of new
enzyme protein or changes in enzymatic activity appears to
differ between CNS noradrenergic and dopaminergic neurons.
For example, increased synthetic demand in noradrenergic
neurons of the brain stem nucleus locus ceruleus appears to
be accomplished primarily by increasing TH gene expression.
In contrast, the same conditions and treatments that increase
TH gene expression in brain stem noradrenergic neurons fail
to increase TH mRNA levels in DA-containing neurons of the
midbrain. In these dopaminergic neurons, it appears that
synthesis is regulated primarily by altering the activity of TH,
i.e., by phosphorylation (see Chapter 9).

Figure 8-4. Model illustrating mechanisms for

regulation of transmitter synthesis in
noradrenergic neurons. TH, tyrosine hydroxylase;
NE, norepinephrine; DA, dopamine; DOPA, 3,4-
dihydroxyphenylalanine.

Storage
A great conceptual advance made in the study of
catecholamines more than 40 years ago was the recognition
that in almost all tissues a large percentage of the NE present
is located within highly specialized subcellular particles (later
shown to be synaptic vesicles but colloquially referred to as
"granules") in sympathetic nerve endings and chromaffin
cells. Much of the NE in the CNS is also located within similar
vesicles. These granules contain adenosine triphosphate (ATP)
in a molar ratio of catecholamine to ATP of about 4:1. Some
complex of the amines with ATP and protein is probable since
the intravesicular concentration of amines, at least in the
adrenal chromaffin granules and probably in the splenic nerve
granules (0.3-1.1 M), would be hypertonic if present in free
solution and might be expected to lead to osmotic lysis of the
vesicles.

Two vesicular monoamine transporters (VMATs) localized to

the membranes of synaptic vesicles and chromaffin granules
have been cloned (VMAT-1 and VMAT-2). VMAT-1 is found in
the adrenal medulla, in the adrenal chromaffin cells that
synthesize and release monoamines. VMAT-2 is present in
catecholamine and serotonin neurons in the CNS. VMAT-2, the
isoform found in the brain, shows modest substrate specificity
and can transport catecholamines and indoleamines, as well
as histamine, into vesicles. These VMATs define another new
family of transporter proteins that display the common motif
of 12 hydrophobic, putative transmembrane domains and
move transmitters into acidic intracellular compartments such
as neurotransmitter vesicles. These transporters display no
significant sequence homology with the plasma membrane
Na+/Cl--coupled family (see below), but they do resemble
bacterial drug-resistance transporters. VMATs can also be
distinguished from plasma membrane transporters by their
use of transmembrane H+ gradients instead of Na+ (see
below). All of the amine storage vesicles studied in brain and
adrenal chromaffin cells contain a vascular-type H+-pumping
ATPase similar to that found in lysosomes and Golgi
membranes. The significant homology of these vesicular
transporters to a group of bacterial drug-resistance
transporters suggests that VMATs may play a role in
detoxification. VMAT enables vesicles to sequester toxins and
reduce their cellular toxicity. For example, mice heterozygous
for one knockout copy of VMAT-2 show increased toxicity of
DA neurons to the DA neurotoxin 1-methyl-4-phenyl-1,2,3,6-
tetrahydropyridine. Loss of VMAT means less sequestration of
the toxin, which can then exert its toxic action by targeting
mitochondrial respiration (see Chapter 9).
Release
The release of NE, similar to the release of acetylcholine
(ACh), is Ca2+-dependent and appears to occur by the same
Ca2+-dependent process (exocytosis) that has been
described for other transmitters (see Chapter 2). Exocytosis
requires that the entire content of the granular vesicle be
released (i.e., catecholamine, ATP, and soluble protein). Since
the nerve terminal region, as far as we know, cannot
synthesize any protein, high rates of axonal flow from the
nerve cell body would be required to replenish the protein lost
during exocytosis. Alternatively, one might propose a "protein
reuptake" mechanism, to recapture that protein released
during synaptic transmission. In peripheral nerves, release of
NE is frequency-dependent within a physiological range of
frequencies. Some evidence has also been presented that
newly synthesized NE may be released preferentially. This
preferential release is additional evidence to support the
contention that NE exists in more than one pool within the
sympathetic neuron.

It has been a great deal more difficult to demonstrate NE

release from its nerve endings in the CNS. However, with the
independent development of the push-pull cannula,
microdialysis techniques, and electrochemical detectors,
catecholamines can be detected extracellularly from certain
deep nuclear masses of the CNS or measured directly by in
vivo voltammetry. NE release detemined by the above
techniques is Ca2+-dependent and influenced by the
functional activity of the noradrenergic neuron.

Regulation of Release
The major homeostatic mechanism for regulation of
catecholamine release in both central and peripheral
catecholamine neurons involves interaction of the released
catecholamine with specific presynaptic receptors
(autoreceptors), which are located on the nerve terminals.

In most catecholamine-containing systems, administration of

catecholamine agonists attenuates stimulus-induced release
while administration of catecholamine receptor blockers
augments release. These pharmacological studies have
established the concept that presynaptic receptors modulate
release by responding to the concentration of catecholamine
in the synapse (high concentrations inhibiting release and low
concentrations augmenting release). Presynaptic
autoreceptors have also been implicated in the regulation of
DA synthesis (see Chapter 9).

Several types of presynaptic receptor are involved in the

inhibition of transmitter release from adrenergic nerves.
These include 2-adrenergic autoreceptors as well as
muscarinic, opiate, and DA receptors. Different presynaptic
receptors are linked to facilitation of NE release, including 2-
adrenergic adrenoceptors, nicotinic cholinergic receptors, and
angiotensin II receptors. The precise mechanism by which
autoreceptors influence neurotransmitter release from
adrenergic neurons varies depending on their location on the
neuron. Activation of 2-adrenoceptors inhibits NE release by
several possible mechanisms, including (1) attenuation of the
rate of Ca2+ entry through inhibition of voltage-gated Ca2+
channels; (2) opening of K channels, leading to
hyperpolarization of the neuron terminals; and (3) inhibition
of adenylate cyclase, resulting in a decrease of intracellular
cAMP and Ca2+ concentration.

Presynaptic -adrenergic receptors are also present in some

sympathetic tissues. These autoreceptors are usually believed
to be of the 2 subtype. The 2-adrenoceptor-mediated
facilitation of NE release may be due to stimulation of
adenylate cyclase, leading to an increase in cAMP and a
subsequent increase in intracellular Ca2+ concentrations.
Alternatively, it has been suggested that the facilitation may
be due to a transsynaptic signal involving a local renin-
angiotensin response. This hypothesis posits that -
adrenoceptor agonists activate postsynaptic 2-adrenoceptors
in the vascular wall. This activation results in the synthesis of
angiotensin II, which diffuses across the synaptic cleft to
activate presynaptic angiotensin II receptors that facilitate NE
release.

The presence of presynaptic facilitatory and inhibitory

adrenergic receptors on the same nerve terminals may
provide for a fine-tuning control of stimulus-evoked NE
release. Since the affinity of E for 2-adrenoceptors is much
greater than that of NE, this subset of adrenoceptors is more
likely to be activated by endogenous E than NE. Thus, low
synaptic concentrations of E could activate presynaptic 2-
adrenoceptors preferentially, leading to an increase in NE
release. At higher concentrations of E or NE, activation of 2-
adrenoceptors would predominate and NE release would be
diminished rather than enhanced.

Prostaglandins of the E series are also potent inhibitors of

neurally induced release of NE in a great number of tissues,
and their action appears to be dissociated from any
interaction with presynaptic receptors. These substances are
released from sympathetically innervated tissues, and most
evidence indicates that inhibition of local prostaglandin
production is associated with an increase in the release of NE
and subsequent effector responses induced by neuronal
activity. The control of NE release by this prostaglandin-
mediated feedback mechanism appears to operate through
restriction of calcium availability for the NE release process
and to be most efficient within the physiological frequency
range of nerve impulses.

Metabolism
The metabolism of exogenously administered or endogenous
catecholamines is markedly slower than that of ACh by the
ACh-ACh-esterase system. The major mammalian enzymes of
importance in the metabolic degradation of catecholamines
are monoamine oxidase (MAO) and catechol-O-
methyltransferase (COMT) (Fig. 8-5). MAO is an enzyme that
converts catecholamines to their corresponding aldehydes.
This aldehyde intermediate is rapidly metabolized, usually by
oxidation by the enzyme aldehyde dehydrogenase to the
corresponding acid. In some circumstances, the aldehyde is
reduced to the alcohol or glycol by aldehyde reductase. In the
case of brain NE, reduction of the aldehyde metabolite
appears to be the favored route of metabolism. MAO is a
particle-bound protein, localized largely in the outer
membrane of mitochondria, although partial microsomal
localization cannot be excluded. There is also some evidence
for a riboflavin-like material in MAO isolated from liver
mitochondria. MAO is usually considered to be an
intraneuronal enzyme, but it occurs in abundance
extraneuronally. In fact, most experiments indicate that
chronic denervation of a sympathetic end organ leads only to
a relatively small reduction in MAO, suggesting that the
greater proportion of this enzyme is extraneuronal. However,
it is the intraneuronal enzyme that seems to be important in
catecholamine metabolism. MAO present in human and rat
brain exists in at least two different forms, designated MAO-A
and MAO-B based on substrate specificity and sensitivity to
inhibition by selected inhibitors.

Clorgyline is a specific inhibitor of MAOA, which has a

substrate preference for NE and serotonin. Deprenyl is a
selective inhibitor of MAOB, which has a substrate preference
for -phenylethylamine and benzylamine as substrates. DA,
tyramine, and tryptamine appear to be equally good
substrates for both forms of the enzyme. MAO-A and -B arise
from distinct genes on chromosome X and are expressed in
different regions of the nervous system. The differential
expresssion of the two forms of MAO in biogenic amine
neurons is noteworthy. MAO-A mRNA appears to be the main
form of the enzyme expressed in peripheral and central
noradrenergic neurons, whereas MAO-B mRNA is found
primarily in serotonergic neurons in the midbrain raphe and
histaminergic neurons in the hypothalamus. Neither mRNA
appears be expressed in midbrain dopaminergic neurons,
although this remains a matter of controversy.

The role and relative importance of these two types of MAO in

physiological and pathological states is currently unknown,
but this is an important area for further research. It has been
speculated that under certain circumstances MAO could
regulate NE biosynthesis by controlling the amount of
substrate DA available to the enzyme dopamine- -
hydroxylase. MAO is not an exclusive catabolic enzyme for the
catecholamines, since it also oxidatively deaminates other
biogenic amines, such as 5-hydroxytryptamine, tryptamine,
and tyramine. The intraneuronal localization of MAO in
mitochondria or other structures suggests that this would
limit its action to amines that are present in a free (unbound)
form in the axoplasm. Here, MAO can act on amines that have
been taken up by the axon before they are sequestered by
VMAT or even on amines that are released from the VMAT
before they pass out through the axonal membrane.
Interestingly, the latter possibility seems of minor
physiological importance since MAO inhibition does not
potentiate the effects of peripheral sympathetic nerve
stimulation.

The second enzyme of importance in the catabolism of

catecholamines is COMT, discovered by Axelrod in 1957. This
enzyme is a relatively nonspecific enzyme that catalyzes the
transfer of methyl groups from S-adenosylmethionine to the
m-hydroxyl group of catecholamines and various other
catechol compounds. COMT is found in the cytoplasm of most
animal tissue, being particularly abundant in kidney and liver.
A substantial amount is also found in the CNS and in various
sympathetically innervated organs. The precise cellular
localization of COMT has not been determined, although it has
been suggested that it functions extraneuronally. The purified
enzyme requires S-adenosylmethionine and Mg2+ ions for
activity. As with MAO, inhibition of COMT activity does not
markedly potentiate the effects of sympathetic nerve
stimulation, although in some tissues it tends to prolong the
duration of the response to stimulation. Therefore, neither
MAO nor COMT seems to be the primary mechanism for
terminating the action of NE liberated at sympathetic nerve
terminals. It may be, however, that these enzymes play a
more important role in terminating transmitter action and
regulating catecholamine function in the CNS. In fact, recent
data have suggested that COMT may play an important role in
the regulation of synaptic DA in the prefrontal cortex (see
Chapter 9) in contrast to its minimal action in the
mesoaccumbens and mesostriatal DA systems.

Figure 8-5. Dopamine and norepinephrine

metabolism. DOPA, 3,4-dihydroxyphenylalanine;
DA, dopamine; NE, norepinephrine; DOMA, 3,4-
dihydroxymandelic acid; DOPAC, 3,4-
dihydroxyphenylacetic acid; DOPEG, 3,4-
dihydroxyphenylglycol; DOPET, 3,4-
dihydroxyphenylethanol; MHPG, 3-methoxy-4-
hydroxyphenylglycol; HVA, homovanillic acid;
VMA, 3-methoxy-4-hydroxymandelic acid; NM,
normetanephrine; MTA, 3-methoxytyramine; MAO,
monoamine oxidase; COMT, catechol-O-
 methyltransferase. Dashed arrows indicate steps
that have not been firmly established.

Reuptake
When sympathetic postganglionic nerves are stimulated at
frequencies low enough to be comparable to those
encountered physiologically, very little intact NE overflows
into the circulation, suggesting that local inactivation is very
efficient. This local inactivation is not significantly blocked
when COMT, MAO, or both are inhibited; and it is believed to
involve mainly reuptake of the transmitter by sympathetic
neurons.

Much attention has been focused on the role of tissue-uptake

mechanisms in the physiological inactivation of
catecholamines, but only in recent years has this concept
received direct experimental support via isolation of the NE
transporter and the cloning of its gene.

This uptake process is a saturable membrane transport

process dependent on temperature and requiring energy. The
stereochemically preferred substrate is L-NE; furthermore, NE
is taken up more efficiently than its N-substituted derivatives.
It is now appreciated that neuronal reuptake of
catecholamines is a major means of inactivation of the
released transmitter in the brain and peripheral nervous
system. The reuptake process conserves transmitter and
allows intracellular enzymes that degrade catecholamines to
act, thus bolstering the actions of extracellular enzymes.
Neuronal reuptake of catecholamines, and indeed of all
transmitters for which a reuptake process has been identified,
has several characteristics. The reuptake process is energy-
dependent, is saturable, and depends on Na+ cotransport. In
addition, extracellular Cl- is necessary for transport. Because
reuptake depends on coupling to the Na+ gradient across the
neuronal membrane, toxins that inhibit Na+,K+-ATPase also
inhibit reuptake. These plasma membrane transporters, in
contrast to the monoamine transporters associated with
neuronal vesicles, are not Mg2+-dependent and not inhibited
by reserpine. A DNA clone encoding a human NE transporter
has been isolated. The isolated cDNA sequence predicts a
protein of 617 amino acids with the typical transporter motif
of 12 highly hydrophobic regions, probable membrane-
spanning domains. Expression of the cDNA clone in
transfected HeLa cells indicates that the NE transport activity
is sodium-dependent and sensitive to NE-transport inhibitors.
Striking sequence homology is notable between the NE
transporter and the rat and human -aminobutyric acid
(GABA) transporters, suggesting a new transporter gene
family.

The plasma membrane transporters have been intensively

studied at the biochemical, pharmacological, and molecular
levels. It has become clear that these Na+- and Cl--coupled
transporters represent a group of integral membrane proteins
encoded by a closely related family of genes that includes the
transporters of monoamines, GABA, glycine, and choline. A
different class of plasma membrane transporters is
represented by the glutamate transporter (see Chapter 6).
Plasma membrane transporters can be classified into families
and subfamilies based on ion dependence, topology, and
sequence relatedness. The NE transporters (NETs) are
members of the Na+/Cl--dependent neurotransmitter
transporter family, which includes the DA transporter (DAT)
and the serotonin transporter (SERT). Expression of NET,
SERT, and DAT in nonneuronal cells has established model
systems for analysis of the structural basis of transporter
specificity for transmitters and transporter-specific
antagonists. The catecholamine transporters, NET and DAT
are not very specific, with each accumulating both DA and NE.
In fact, the NET has a higher affinity for DA than for NE and,
under certain conditions, can play a role in the modulation of
DA transmission (see Chapter 9). No specific transporter for
E-containing neurons has been found in mammalian species,
but one has been identified in the frog. The availability of
pure transporter proteins has facilitated the development of
transporter-specific antibodies and nucleic acid probes and
stimulated renewed interest in the endogenous control
mechanisms acutely regulating monoamine transport in vivo.
Also, the availability of human cDNA encoding the NET offers
the opportunity to determine whether alterations in
transporter genes could have important etiological
implications for major depression or affective disorder.

Since many aspects of NE neurotransmission, including

synthesis and release, are tightly regulated, it would not be
unexpected to find that the NET is also subject to regulatory
control. In fact, almost 40 years ago, Schneider and Gillis
noted that, following stimulation of the sympathetic input to
the heart, a rapid increase in the retention of NE occurred in
the cat atrium. A number of similar observations have been
made over the past four decades, suggesting that NE
reuptake increases in parallel with an increased rate of
sympathetic discharge and NE release. Although the
mechanism responsible for the increase in NET activity
remains to be determined, the presence of serine/threonine
phosphorylation sites on human NET raises the possibility that
this effect might be mediated by protein phosphorylation. The
role of protein phosphorylation or other posttranslational
modifications in regulating transporter function is only
beginning to be evaluated. However, the fact that NET
proteins can now be visualized suggests that this should be a
fertile area for future investigations. Little is currently known
about promotor regions and other regulatory elements
involved in the transcriptional regulation of the NET and SERT
genes. However, rats treated chronically with antidepressant
drugs exhibit a decrease in SERT mRNA levels and a reduction
in NET using [3H]nisoxetine autoradiography.

The availability of cDNAs encoding transporter proteins may

facilitate the development of sensitive screening techniques to
help develop new and selective agents that target specific
neurotransmitter transporters. Although this screening
technology has not yet led to the development of uniquely
selective inhibitors of the monoamine transporters, a number
of selective inhibitors of NE uptake are available, several of
which are used clinically in the treatment of affective
disorders. Several of these NE-uptake blockers are illustrated
in Figure 8-6.

Figure 8-6. Representative compounds that

selectively inhibit norepinephrine reuptake.

Adrenergic Receptors
When catecholamines are released from either noradrenergic
nerve terminals or the adrenal medulla, they are recognized
by and bind to specific receptor molecules on the plasma
membrane of the neuroeffector cells that transduce
catecholamine interactions with the cell into a physiological
response. Depending on the nature of the receptor, this
interaction sets off a cascade of membrane and intracellular
events, which cause the cell to carry out its specialized
function. Classically, peripheral adrenergic receptors have
been divided into two distinct classes, called and
receptors. The development of synthetic compounds active at
adrenergic receptors has allowed further differentiation of
adrenergic receptors into 1, 2, 1, and 2 subtypes (Fig. 8-
7). Genes for these subtypes of adrenergic receptor have
been cloned; this has shown that these receptors are
members of a larger family of hormone receptors that
mediate their activity through interaction with one of a series
of guanine nucleotide-binding regulatory proteins (G
proteins). In addition to traditional classification based on
their pharmacological profile, adrenergic receptors can be
divided into three major classes by their differential coupling
to G proteins (Table 8-1). -Adrenergic receptors activate Gs
to stimulate adenylate cyclase. 2-Adrenergic receptors
decrease adenylyl cyclase activity through coupling to Gi. 1-
Adrenergic receptors stimulate phospholipase C action
through a still ill-defined Gq. As schematically illustrated in
Figure 8-7, -adrenoceptor signals are transmitted through
the effector cell membrane via the adenylyl cyclase system.
Occupation of the adrenoceptor by the catecholamines
stimulates adenylyl cyclase to generate cAMP by a series of
intramembrane events (see Chapter 5). Initially, the receptor
interacts with a guanosine triphosphate (GTP)-dependent
protein, Gs. Gs is linked to a catalytic unit, which generates
cAMP upon activation by Gs; the catalytic unit then converts
adenosine monophosphate (AMP) to cAMP. The latter
compound triggers a series of intracellular events involving
protein kinases. The protein kinases activate further unknown
biochemical changes to generate the final biological response
to the transmitter.

The order of potency for stimulation of 1-adrenergic

receptors by catecholamines is isoproterenol NE E. For the
2 adrenoceptor, the order of potency is isoproterenol E
NE. Adrenergic blocking agents such as propranolol and
alprenolol prevent the activation of receptors. The
interaction of adrenergic agents with their receptors is
saturable, stereospecific, and reversible. Prolonged exposure
of adrenoceptors to endogenous or exogenous agonists
often reduces the responsiveness of these receptors
(desensitization). Desensitization can also be caused by
uncoupling of receptors from the adenylyl cyclase and by a
decrease in the number of receptors (downregulation) (Fig. 8-
8). Depriving the adrenoceptors of catecholamine (by
chemical or surgical denervation) increases their
responsiveness.

The recently identified 3 adrenoceptor appears to be

responsible for lipolysis in white adipose tissue and
thermogenesis in brown adipose tissue found in rodents. The
3 adrenoceptor shows lower affinity for agonists, with a rank
order of potency of NE isoproterenol E and a lower affinity
for known -adrenoceptor antagonists. All three -
adrenoceptor subtypes appear to be linked to adenylyl cyclase
activation through a stimulatory G protein, and at present
there is no evidence for subtype-related differences in
receptor-mediated cyclase interactions. However, the 3
adrenoceptor is insensitive to blockade by most -
adrenoceptor antagonists, making pharmacological evaluation
of its function difficult. Several selective -adrenoceptor
antagonists have been developed that exhibit high specificity
for the 3 adrenoceptor. These new agents have facilitated
further studies on the functions subserved by this receptor
subtype. In contrast to the 1 and 2 adrenoceptors, the 3
adrenoceptor is resistant to both agonist-mediated short-term
desensitization and, on a prolonged time scale, agonist-
induced receptor downregulation. The biological importance of
this differential response to the regulation by agonists is
uncertain but suggests that during prolonged activation of the
sympathetic nervous system 3 receptor-mediated effects
might be preserved while 1- and 2-mediated responses are
diminished. The 1 adrenoceptor appears to be associated
with calcium mobilization (Fig. 8-7), possibly via stimulation
of phospholipase C and phospholipase. The 1 receptor is
distinguished from the 2 adrenoceptor by its inhibition by
the selective 1-blocking agent prazosin. The 2-adrenergic
receptor is negatively linked to the adenylyl cyclase complex
via an inhibitory G protein (Gi). Like Gs, Gi is activated by
GTP, but in this case it inhibits the generation of AMP by the
catalytic unit. In the peripheral sympathetic nervous system,
the 2 receptor is localized mainly on presynaptic nerve
terminals, where it modulates NE release. Stimulation of this
receptor by catecholamines inhibits impulse-dependent
release of transmitter; blockade of this receptor facilitates
release.

Although it is clear that activation of 2 adrenoceptors inhibits

adenylyl cyclase activity (Fig. 8-7) mediated through an
inhibitory G protein, this may not in all cases represent the
signal-transduction mechanism responsible for the effects
associated with receptor activation. For example, the 2-
mediated inhibition of neurotransmitter release is generally
insensitive to inactivation of inhibitory G proteins by pertussis
toxin, suggesting the involvement of another as yet
undetected signal-transduction mechanism.

Figure 8-7. Model of a norepinephrine synapse

illustrating the presynaptic and postsynaptic
molecular entities involved in the synthesis,
storage, release, reuptake, and signaling of
norepinephrine. Tyrosine is transported into the
presynaptic terminal by an active uptake
mechanism and converted to NE by a series of
enzymatic steps. NE is taken up from the
axoplasm and stored by the VMAT. Once released,
NE can interact with two categories of adrenergic
receptors, alpha and beta. Alpha 2 autoreceptors
localized in the nerve terminals modulate the
synthesis and release of NE, while cell body and
dendritic 2 autoreceptors modulate impulse flow.

The majority of G protein-coupled adrenergic

The majority of G protein-coupled adrenergic
receptors are localized postsynaptically where they
mediate the cellular responses of the postsynaptic
neuron. NE is metabolized by MAO and COMT
giving rise to MHPG, a major brain NE metabolite.
The synaptic action of NE is terminated by
reuptake into the presynaptic terminal by NET. AC,
adenylyl cyclase; AR, adrenergic receptor; DAG,
diacylglycerol; IP3, inositol triphosphate; MHPG,
3-methoxy-4-hydroxy phenethylene glychol; NM,
normetanephrine; PLC, phospholipase C; NET,
plasma membrane norepinephrine transporter,
VMAT, vesicular monoamine transporter. (Modified
from Nesther et al., Molecular
Neuropharmacology)

Figure 8-8. Schematic illustration of multiple

mechanisms underlying regulation of the -
adrenergic receptor ( AR). (1) AR stimulation of
the cyclic adenosine monophosphate (cAMP)
system results in phosphorylation of the receptor
by cAMP-dependent protein kinase, which leads to
uncoupling of the receptor from Gs. The activated
protein kinase would also phosphorylate many
other proteins not shown in the figure, which
would then mediate many of the actions of AR
activation. (2) Prolonged activation of the receptor
leads to phosphorylation by another kinase, AR
kinase ( ARK), which phosphorylates only the
agonist-activated form of the receptor. This results
in binding of the receptor to -arrestin, which
competes with Gs and thereby inhibits AR
stimulation of adenylyl cyclase (AC). (3) Loss of
ARs from the membrane occurs when receptors
are internalized and sequestered into intracellular
vesicles. This pool of receptors is then available
for either recycling back to the membrane or
 for either recycling back to the membrane or
degradation. Such sequestration, internalization,
degradation, and membrane reinsertion may be
mediated via receptor phosphorylation and
dephosphorylation involving the cAMP and/or
ARK pathways. Another mechanism by which
receptor activation leads to downregulation of the
AR is via regulation of receptor mRNA levels,
which may occur by two primary mechanisms. (4)
The level of receptor mRNA is regulated by the
stability or half-life of the mRNA. Although the
mechanisms responsible for regulation of mRNA
stability have not been identified, they may also
involve cAMP-dependent protein kinase. (5) The
level of receptor mRNA is also regulated via
changes in AR gene transcription. This effect is
mediated by the cAMP pathway and appears to
involve translocation of the cAMP-dependent
protein kinase catalytic subunit into the nucleus
and the phosphorylation of constitutively
expressed transcription factors (e.g., cAMP
response element-binding protein [CREB]). It
might also depend on the subsequent induction of
other transcription factors (e.g., immediate early
gene [IEG] products such as c-Fos). In addition to
regulation of receptor gene transcription, such
regulation of transcription factors would mediate
the effects of AR activation on the expression of
many other genes, for example, those for G
proteins, cAMP-dependent protein kinase,
neurotrophins, and neuropeptides. This, in turn,
would mediate many of the more long-term
effects of AR activation on brain function. (From
Duman and Nestler, 1995.)

Dynamics of Adrenergic Receptors

Adrenergic receptors do not seem to be static entities but
change in number and affinity in response to altered synaptic
activity. In both brain and the peripheral sympathetic nervous
system, destruction of adrenergic neurons is associated with
functional supersensitivity of postsynaptic sites. Conversely,
increasing synaptic NE by administering uptake blockers
(tricyclic antidepressants) or MAO inhibitors leads to
functional subsensitivity. These changes appear to be a
compensatory response to altered levels of synaptic
transmitter. The number of 1 and 2 receptors also increases
after noradrenergic neurons in the brain, and sympathetically
innervated tissues have been destroyed by administration of
6-hydroxydopamine. After lesions in NE-containing neurons
are made in the cerebral cortex, the number of 1 receptors
increases markedly but no change occurs in the number of 2
receptors. This may be because 2 receptors have a low
affinity for NE and the concentration of E in the cerebral
cortex is relatively low. Likewise, chronic administration of
tricyclic antidepressants leads to a selective decrease in the
density of 1-adrenergic receptors in the cerebral cortex. This
finding implies that the 1-adrenergic receptors in the cortex
are functionally innervated by adrenergic neurons.

Desensitization
The waning of a stimulated response in the face of continuous
agonist exposure is termed desensitization. Desensitization
has been demonstrated in many hormone- and
neurotransmitter-receptor systems, including those that
activate different G proteins and effector systems.
Desensitization of the cAMP response elicited by -adrenergic
agonists is a useful model system for studying this
phenomenon. Desensitization has been viewed historically as
two separate processes: heterologous and homologous.
Heterologous desensitization occurs when exposure of cells to
a desensitizing agent leads to diminished responsiveness to a
number of different stimuli. In contrast, homologous
desensitization is much more specific and involves only the
loss of responsiveness to the specific desensitizing agent. A
number of molecular mechanisms operating at the receptor
level have been implicated in the process of desensitization.
The heterologous pattern of desensitization is thought to be
mediated by phosphorylation of receptors coupled to Gs by
protein kinase A, whereas the homologous pattern of
desensitization is thought to involve phosphorylation of
receptors by a novel, cAMP-independent kinase, such as -
adrenergic receptor kinase. Recent studies on the -
adrenergic receptor have suggested that the principal
mechanisms underlying rapid, agonist-induced desensitization
of -adrenergic receptor function in intact mammalian cells
are a combination of these two processes and involve
phosphorylation of the receptor by both types of kinase.
Phosphorylation of the receptor is thought to lead to
uncoupling of the -adrenergic receptor from the stimulatory
G protein (Gs). It is not yet clear exactly how phosphorylation
leads to uncoupling of -adrenergic receptor from Gs; for
example, phosphorylation of one specific residue may directly
impair interaction of the receptor with Gs, whereas
phosphorylation of another residue may allow cytosolic factors
such as -arrestin to bind to the phorphorylated -adrenergic
receptor and decrease or prevent the coupling of the -
adrenergic receptor with Gs, leading to desensitization.

CNS CATECHOLAMINE NEURONS

The cellular organization of the brain and spinal cord has been
studied for many decades by classic histological and silver
impregnation techniques. With the development in the 1960s
of completely different histological techniques based on the
presence of a given type of transmitter substance or on
specific synthetic enzymes involved in the formation of a
given transmitter, it became possible to map chemically
defined neuronal systems in the CNS of many species. By the
time such techniques had been employed for several years, it
became clear that the distribution of these chemically defined
monoamine neuronal systems did not necessarily correspond
to that of systems described earlier with the classic
techniques.

However, it was really not until the mid-1960s that the

histochemical fluorescent technique of Falck and Hillarp was
applied to brain tissue and the anatomy of the monoamine-
containing neuronal systems was described. By a variety of
pharmacological and chemical methods, it has subsequently
been possible to discriminate between NE-, E-, and DA-
containing neurons and to describe in detail the distribution of
these catecholamine-containing neurons in the mammalian
CNS. Several recent and thorough surveys of these systems
are available using multiple histochemical measures including
mapping of mRNAs, immunodetection of synthetic proteins,
and ligand binding to receptors and transporters.

Figure 8-9 provides a schematic model of a central

monoamine-containing neuron as observed by fluorescence
and electron microscopy. The cell bodies contain relatively low
concentrations of amine (about 10-100 mg/g), while the
terminal varicosities contain a very high concentration (about
1000-3000 mg/g). The axons, however, consist of highly
branched, largely unmyelinated fibers that have such a low
concentration of amine that they are barely visible in
untreated adult animals. With the electron microscope,
depending on the type of fixation, it is possible to observe
small granular vesicles that are thought to represent the
subcellular storage sites containing the catecholamines. These
granular vesicles are concentrated in the terminal varicosities
of the central noradrenergic neuron, just as they are observed
in the peripheral sympathetic neuron.

Figure 8-9. Schematic illustration of a central

monoamine-containing neuron. Right side depicts
the general appearance and intraneuronal
distribution of monoamines based on fluorescence
microscopy. Cell bodies contain a relatively low
concentration of catecholamine (about 10-100
ug/g), while the terminal varicosities contain a
very high concentration (1000-3000 ug/g).
Preterminal axons contain very low concentrations
of amine. At the electron microscopic level (left
side), dense core granules can be observed in the
cell bodies and axons but appear to be highly
concentrated in the terminal varicosities. (Modified
from Fuxe and Hokfelt, 1971.)

SYSTEMS OF CATECHOLAMINE PATHWAYS

IN THE CNS
Detailed analysis of catecholamine pathways in the CNS has
been greatly accelerated by improvements in the application
of fluorescence histochemistry, such as the use of glyoxylic
acid as the fluorogen and by the application of numerous
auxiliary mapping methods mentioned above. Extensive
progress in the functional analysis of these systems has also
been made possible by evaluation with microelectrodes of the
effects produced by selective electrical stimulation of the
catecholamine (especially NE) cell-body groups. From such
studies, it is clear that the systems can be characterized in
simple terms only by ignoring large amounts of detailed
cytological and functional data and that the earlier
catecholamine wiring diagrams are no longer tenable.
Furthermore, anatomic details for monoamine systems in
rodents seem to bear only rudimentary homology to their
detailed selective anatomic configurations in human and
nonhuman primates. There are two major clusterings of NE
cell bodies from which axonal systems arise to innervate
targets throughout the entire neuraxis.

LOCUS CERULEUS

Introduction
This compact cell group within the caudal pontine central gray
is named for the pigment the cells bear in humans and some
higher primates; in the rat, the nucleus contains about 1500
neurons on each side of the brain (Fig. 8-10). In humans, the
locus ceruleus is composed of about 12,000 large neurons on
each side of the brain. The axons of these neurons form
extensive collateral branches, which project widely along well-
defined tracts. At the electron microscope level, terminals of
these axons exhibit, under appropriate fixation methods, the
same type of small granular vesicles seen in the peripheral
sympathetic nerves (Fig. 8-2).

Fibers from the locus ceruleus form five major noradrenergic

tracts (Fig. 8-11): the central tegmental tract (or dorsal
bundle described by Ungerstedt), a central gray dorsal
longitudinal facsiculus tract, and a ventral tegmental-medial
forebrain bundle tract. These tracts remain largely ipsilateral,
although there is a crossing over in some species of up to
25% of the fibers. These three ascending tracts then follow
other major vascular and fascicular routes to innervate all
cortices, specific thalamic and hypothalamic nuclei, and the
olfactory bulb. Another major fascicle ascends in the superior
cerebellar peduncle to innervate the cerebellar cortex. The
fifth major tract descends into the mesencephalon and spinal
cord, where the fibers course in the ventral-lateral column. At
their terminals, the locus ceruleus fibers form a plexiform
network in which the incoming fibers gain access to a cortical
region by passing through the major myelinated tracts,
turning vertically toward the outer cortical surface and then
forming characteristic T-shaped branches, which run parallel
to the surface in the molecular layer; this pattern is seen in
the cerebellar, hippocampal, and cerebral cortices.

Virtually all of the noradrenergic pathways that have been

studied physiologically so far are efferent pathways of the
locus ceruleus neurons; in cerebellum, hippocampus, and
cerebral cortex, the major effect of activating this pathway is
to inhibit spontaneous discharge. This effect has been
associated with the slow type of synaptic transmission, in
which the hyperpolarizing response of the target cell is
accompanied by increased membrane resistance. The
mechanism of this action has been experimentally related to
the second-messenger scheme in which the noradrenergic
receptor elicits its characteristic action on the target cells by
activating the synthesis of cAMP in or on the postsynaptic
membrane. Pharmacologically and cytochemically, target cells
responding to NE or the locus ceruleus projection in these
cortical areas exhibit -adrenergic receptors.

Figure 8-10. Formaldehyde-induced fluorescence

of the rat nucleus locus ceruleus. In this frontal
section through the principal portion of the
nucleus, intensely fluorescent neurons can be
seen clustered closely together. Within the
neurons, the cell nucleus, which is not fluorescent
after this treatment, appears dark except for the
nucleolus. (Bloom, unpublished) ( 650).

Figure 8-11. Diagram of the projections of the

locus ceruleus viewed in the sagittal plane. See
text for description. AON, anterior olfactory
nucleus; AP-VAB, ansa peduncularis-ventral
amygdaloid bundle system; BS, brain stem nuclei;
C, cingulum; CC, corpus callosum; CER,
 cerebellum; CTT, central tegmental tract; CTX,
cerebral neocortex; DPS, dorsal periventricular
system; DTB, dorsal catecholamine bundle; EC,
external capsule; F, fornix; H, hypothalamus; HF,
hippocampal formation; LC, locus ceruleus; ML,
medial lemniscus; MT, mammillothalamic tract;
OB, olfactory bulb; PC, posterior commissure; PT,
pretectal area; RF, reticular formation; S, septal
area; SC, spinal cord; SM, stria medullaris; SOD,
supraoptic decussations; ST, stria terminalis; T,
tectum; TH, thalamus. (Diagram compiled by R. Y.
Moore, from the observations of Lindvall and
Bjorklund, 1974; Jones and Moore, 1977.)

Lateral Tegmental Noradrenergic Neurons

A large number of noradrenergic neurons lie outside of the
locus ceruleus, where they are more loosely scattered
throughout the lateral ventral tegmental fields. In general, the
fibers from these neurons intermingle with those arising from
the locus ceruleus, those from the more posterior tegmental
levels contributing mainly descending fibers within the
mesencephalon and spinal cord and those from the more
anterior tegmental levels contributing to the innervation of
the forebrain and diencephalon. Because of the complex
intermingling of the fibers from the various noradrenergic cell
body sources, the physiological and pharmacological analysis
of the effects of brain lesions becomes extremely difficult. The
neurons of the lateral tegmental system may be the primary
source of the noradrenergic fibers observed in the basal
forebrain, such as amygdala and septum. No specific analysis
of the physiology of these projections has yet been reported;
therefore, it remains to be established whether the -
receptive cAMP mechanism associated with the cortical
projections of the locus ceruleus group also apply to the
synapses of the lateral tegmental neurons.
EPINEPHRINE NEURONS

Introduction
In the sympathetic nervous system and the adrenal medulla,
E shares with NE the role of final neurotransmitter, the
proportion of this sharing being a species-dependent,
hormonally modified arrangement. Until method refinement,
however, little evidence could be gathered to document the
existence of E in the CNS because the chemical methods for
analyzing E levels or for detecting activity attributable to the
synthesizing enzyme PNMT were unable to provide
unequivocal data. With the development of sensitive
immunoassays for PNMT and their application in
immunohistochemistry and with the application of gas
chromatography-mass fragmentography and high-
performance liquid chromatography with electrochemical
detection to brain neurotransmitter measurements, the
necessary data were rapidly acquired and the existence of E-
containing neurons in the CNS confirmed. By
immunohistochemistry, E-containing neurons are defined as
those that are positively stained (in serial sections) with
antibodies to TH, dopamine- -hydroxylase, and PNMT. These
cells are found largely in two groups: one, called C-1 (see
Hokfelt et al., 1984), is intermingled with noradrenergic cells
of the lateral tegmental system; the other, called C-2, is
found in the regions in which the noradrenergic cells of the
dorsal medulla are also found. The axons of these two E
systems ascend to the hypothalamus with the central
tegmental tract, then via the ventral periventricular system
into the hypothalamus. A third group of cells (C-3) in the
midline (medial longitudinal fascicle) has also been described
(Fig. 8-12). Within the mesencephalon, the E-containing
fibers innervate the nuclei of visceral efferent and afferent
systems, especially the dorsal motor nucleus of the vagus
nerve. In addition, E fibers innervate the locus ceruleus, the
intermediolateral cell columns of the spinal cord, and the
periventricular regions of the fourth ventricle.

Although there are considerably fewer E than DA and NE

neurons in the brain, they appear to have a discrete anatomic
distribution and are believed to subserve physiological
functions discrete from other catecholamines. Selective
plasma membrane transporters have been described for DA
and NE, and these transporters have been extensively studied
both in vivo and in vitro. However, little attention has been
given to the hypothetical E transporter, and at the present
time a specific transporter for E has not been isolated from
mammalian brain nor have drugs that selectively block E
reuptake been described. All of the pharmacological agents
that have been shown to block the NET (see Fig. 8-7)
selectively are also effective inhibitors of E uptake. Thus, it
remains uncertain whether a selective transporter for E
actually exists in the mammalian CNS or peripheral nervous
system. The cloning and expression of the E transporter
might permit a search for selective inhibitors, which, if
discovered, could provide valuable pharmacological tools for
elucidating the functional role of central E-containing neurons.

Except for tests of E in the locus ceruleus (where it inhibits

firing), no other tests of the cellular physiology of this system
have thus far been reported. Our understanding concerning
the function of E-containing neurons in the brain is still very
limited, but based on the distribution of E in specific brain
regions, attention has been directed to their possible role in
neuroendocrine mechanisms and blood pressure control.

Figure 8-12. Schematic illustration of the

distribution of the main central neuronal pathways
containing epinephrine. The C1 group represents a
rostral continuation of the noradrenergic cell group
in the ventrolateral medulla oblongata. The dorsal
 C2 group of cells is located mainly in the dorsal
vagal complex, with the vast majority of cell
bodies in the solitary tract nucleus. A C3 group of
cells in the midline (medial longitudinal fascicle)
has also been described. The ventral group of
epinephrine-containing neurons gives rise to both
ascending and descending pathways, innervating
largely periventricular regions such as the
periaqueductal central gray and various
hypothalamic nuclei and the lateral sympathetic
column in the spinal cord. OB, olfactory bulb; CC,
corpus callosum; CAUD, caudate; SEP, septum;
AP, ansa peduncularis; HI, hippocampus; MO,
Medulla Oblongata. (Data from Hokfelt et al.,
1984.)

Coexistence of Classic Transmitters and

Peptides
The finding of numerous peptides in the CNS has raised the
question of the relationship of the neurons containing these
substances to neurons containing classic neurotransmitters. It
was originally believed that the peptide-containing nerves
represented separate identifiable systems. However, when
studies in the peripheral nervous system clearly indicated the
coexistence of peptides with the sympathetic and
parasympathetic transmitters NE and ACh, respectively, this
initiated a series of systematic studies in the CNS to
determine if neurotransmitters in the brain coexisted with one
or more peptides. These studies revealed that the coexistence
of classic transmitters and peptides may represent a rule
rather than an exception. The functional significance of the
occurrence of a catecholamine and a peptide in the same
neuron and their possible release from the same nerve ending
is still unclear in the CNS. However, some insight may be
obtained from studies in the periphery where two distinct
types of interaction have been noted.

In the salivary gland, the parasympathetic cholinergic

neurons contain a vasoactive intestinal polypeptide (VIP)-like
peptide while the sympathetic NE neurons contain a
neuropeptide Y-like peptide. In this system, the peptides
seem to augment the action of the classic transmitters. Thus,
VIP induces vasodilation and enhances the secretory effects of
ACh while neuropeptide Y causes vasoconstriction like NE. In
contrast, in the vas deferens, where the NE neurons
innervating this tissue also contain neuropeptide Y, the
peptide seems to inhibit the release of NE via a presynaptic
action. There is also some indication of preferential storage
and release of NE and neuropeptide Y, with release of the
peptide occurring preferentially at higher frequencies or
during burst firing.

The physiological significance of multiple messengers at the

synapse in the CNS is still unclear, but an appreciation of
coexistence phenomena may influence our view on
interneuronal communication and in a larger perspective may
be of importance in our efforts to treat or prevent various
disease states or abnormalities of the nervous system.

CNS CATECHOLAMINE METABOLISM

Catecholamine metabolism was discussed earlier, and only a
few aspects pertinent to central catecholamine metabolism
will be covered here.

In the peripheral sympathetic nervous system, the aldehyde

intermediate produced by the action of MAO on NE and
normetanephrine can be oxidized to the corresponding acid or
reduced to the corresponding glycol. In contrast to the CNS,
peripheral findings are that oxidation usually exceeds
reduction and, quantitatively, vanilylmandelic acid (VMA) is
the major metabolite of NE and is readily detectable in the
urine. In fact, urinary levels of VMA are routinely measured in
clinical laboratories to provide an index of peripheral
sympathetic nerve function as well as to diagnose the
presence of catecholamine-secreting tumors such as
pheochromocytomas and neuroblastomas. In the CNS,
however, reduction of the intermediate aldehyde formed by
the action of MAO on catecholamines or catecholamine
metabolites predominates, and a major metabolite of NE
found in the brain is the glycol derivative 3-methoxy-4-
hydroxy-phenethyleneglycol (MHPG). Very little, if any, VMA is
found in the brain.

In many species, a large fraction of the MHPG formed in the

brain is sulfate-conjugated. However, in primates, MHPG
exists primarily in an unconjugated "free" form in the brain.
Some normetanephrine is also found in the brain and spinal
cord. Destruction of noradrenergic neurons in the brain or
spinal cord causes a reduction in the endogenous levels of
these metabolites although not as marked as the
corresponding reduction in endogenous NE. Direct electrical
stimulation of the locus ceruleus or severe stress produces an
increase in the turnover of NE as well as an increase in the
accumulation of the sulfate conjugate of MHPG in the rat
cerebral cortex. These effects are completely abolished by
ablation of the locus ceruleus or by transection of the dorsal
pathway, suggesting that the accumulation of MHPG-sulfate in
the brain may reflect the functional activity of central
noradrenergic neurons. Since MHPG-sulfate readily diffuses
from the brain into the cerebrospinal fluid (CSF) or general
circulation, estimates of its concentration in the CSF or in
urine are thought to reflect the activity of noradrenergic
neurons in the brain. Even though MHPG is proportionately a
minor metabolite of NE in the peripheral sympathetic nervous
system, a fairly large portion of the MHPG in the urine still
derives from the periphery. In rodents, it has been estimated
that the brain provides a minor contribution (25%-30%) to
urinary MHPG, while in primates, the brain contribution is
quite large (60%). Thus, at least in rodents, it is quite
probable that relatively large changes in the formation of
MHPG by the brain are necessary to produce detectable
changes in urinary MHPG. For example, in rats, destruction of
the majority of NE-containing neurons in the brain by
treatment with 6-hydroxydopamine leads to only about a
25% decrease in urinary MHPG levels. Nevertheless,
measurement of urinary changes in MHPG is still a reasonable
strategy for obtaining some insight into possible alterations of
brain NE metabolism in patients with psychiatric illnesses.
Measurement of CSF levels of MHPG also provides another
possible approach to assessing central adrenergic function in
human subjects. More recent studies have suggested that
plasma levels of MHPG might provide a reflection of central
noradrenergic activity. In these studies, stimulation of the
locus ceruleus in the rat was shown to result in a significant
increase in the levels of plasma MHPG; and administration of
drugs that are known to alter noradrenergic activity in
rodents has predictably changed plasma levels and venous-
arterial differences in MHPG in nonhuman primates. Changes
in urinary, plasma, and CSF levels of MHPG and their
relationship to central noradrenergic function have to be
interpreted with considerable caution.

PHARMACOLOGY OF CENTRAL
CATECHOLAMINE-CONTAINING NEURONS

Introduction
The psychotropic drugs (drugs that alter behavior, mood, and
perception in humans and behavior in animals) can be divided
into two main categories: the psychotherapeutic drugs and
the psychotomimetic agents. Psychotherapeutic drugs can be
further divided, on the basis of their activity in humans, into
at least four general classes: antipsychotics, antianxiety
drugs, antidepressants, and stimulants.
A great deal of information has become available concerning
the anatomy, biochemistry, and functional organization of
catecholamine systems. This has fostered knowledge
concerning the mechanisms and sites of action of many
psychotropic drugs. It is now appreciated that drugs can
influence the output of monoamine systems by interacting at
several distinct sites. For example, drugs can influence the
output of catecholamine systems by (1) acting presynaptically
to alter the life cycle of the transmitter (i.e., synthesis,
storage, and release, etc. [see Figs. 8-13 and 9-8]); (2)
acting postsynaptically to mimic or block the action of the
transmitter at the level of postsynaptic receptors; and (3)
acting at the level of cell body autoreceptors to influence the
physiological activity of catecholamine neurons. The activity
of catecholamine neurons can also be influenced by
postsynaptic receptors via negative neuronal feedback loops.
In the latter two actions, drugs appear to exert their influence
by interacting with catecholamine receptors. Figure 8-14
illustrates the various neuronal pathways and local
mechanisms regulating synaptic homeostasis in central
noradrenergic neurons. In general, catecholamine receptors
can be subdivided into two broad categories: those that are
localized directly on catecholamine neurons (often referred to
as presynaptic receptors) and those on other cell types, often
termed simply postsynaptic receptors since they are
postsynaptic to the catecholamine neurons, the source of the
endogenous ligand. When it became appreciated that
catecholamine neurons, in addition to possessing receptors on
the nerve terminals, appear to have receptors distributed over
all parts of the neurons (i.e., soma, dendrites, and
preterminal axons), the term presynaptic receptors really
became inappropriate as a description for all of these
receptors. Carlsson, in 1975, suggested that autoreceptor was
a more appropriate term to describe these receptors since the
sensitivity of these catecholamine receptors to the neuron's
own transmitter seemed more significant than their location
relative to the synapse. The term autoreceptor achieved rapid
acceptance, and pharmacological research in the succeeding
years resulted in the detection and description of
autoreceptors on neurons in almost all chemically defined
neuronal systems.

The presence of autoreceptors on some neurons may turn out

to be only of pharmacological interest since they may never
encounter effective concentrations of the appropriate
endogenous agonist in vivo. Others, however, in addition to
their pharmacological responsiveness, may play a very
important physiological role in the regulation of presynaptic
events. This certainly appears to be the case for DA
autoreceptors (see Chapter 9).

The pharmacology of central and peripheral NE neurons is

quite similar (Fig. 8-13). The main differences seem to be
related primarily to drug delivery and the greater complexity
of the neuronal pathways regulating synaptic homeostasis of
the central NE systems (Fig. 8-14).

Figure 8-13. Schematic model of central

noradrenergic neuron indicating possible sites of
drug action.
Site 1: Enzymatic synthesis:

a. Tyrosine hydroxylase reaction blocked by the

competitive inhibitor, -methyltyrosine.
b. Dopamine- -hydroxylase reaction blocked by a
dithiocarbamate de-rivative, Fla-63-bis-(1-methyl-
4-homopip erazinyl-thiocarbonyl)-disulfide (FLA
63).

Site 2: Storage: Reserpine and tetrabenazine

interfere with the uptake-storage mechanism of
the amine granules. Depletion of norepinephrine

(NE) produced by reserpine is long-lasting, and

(NE) produced by reserpine is long-lasting, and
the storage granules are irreversibly damaged.
Tetrabenazine also interferes with the uptake-
storage mechanism of the granules, except the
effects of this drug are of a shorter duration and
do not appear to be irreversible.

Site 3: Release: Am chanism by which

amphetamine causes release is by its ability to
block effectively the reuptake mechanism.

Site 4: Receptor interaction:

a. Presynaptic 2 autoreceptors
b. Postsynaptic receptors

Clonidine appears to be a very potent

autoreceptor-stimulating drug. At higher doses,
clonidine will also stimulate postsynaptic
receptors. Phenoxybenzamine and phenotolamine
are effective receptor-blocking agents. These
drugs may also have some presynaptic 2
receptor-blocking action. However, yohimbine and
piperoxane are more selective as 2 receptor-
blocking agents.

Site 5: Reuptake: NE has its action terminated by

being taken up into the presynaptic terminal. The
tricyclic drug desipramine is an example of a
potent inhibitor of this uptake mechanism.

Site 6: Monoamine Oxidase (MAO): NE or

dopamine (DA) present in a free state within the
presynaptic terminal can be degraded by the
enzyme MAO, which appears to be located in the
outer membrane of mitochondria. Pargyline is an
effective inhibitor of MAO.
effective inhibitor of MAO.

Site 7: Catechol-O-methyltransferase (COMT): NE

can be inactivated by the enzyme COMT, which is
believed to be localized outside the presynaptic
neuron. Tropolone is an inhibitor of COMT. The
normetanephrine (NM) formed by the action of
COMT on NE can be further metabolized by MAO
and aldehyde reductase to 3-methoxy-4-
hydroxyphenylglycol (MHPG). The MHPG formed
can be further metabolized to MHPG-sulfate
(MHPG-S) by the action of a sulfotransferase
found in the brain. Although MHPG-S is the
predominant form of this metabolite found in
rodent brain, the free unconjugated MHPG is the
major form found in primate brain.

Figure 8-14. Neural pathways regulating synaptic

homeostasis of locus ceruleus neurons. Influence
of locus ceruleus neuron (shaded) on its target
cells can be regulated both by modulation of
transmitter release (1-7) and by amplification of
signal provided by that release (8). This involves a
large variety of cell surface receptors, including
receptors that respond to norepinephrine itself ( )
as well as receptors responding to other chemical
signals ( , , , ). Principal pathways for
regulating norepinephrine release: 1, direct action
of recurrent collaterals onto soma; 2, indirect
action of recurrent collaterals, mediated via
influence on presynaptic afferents; 3, direct action
of norepinephrine on presynaptic terminal; 4,
alterations in rate of norepinephrine reuptake; 5,
humoral signals generated by target; 6, neural
signals providing short-loop negative feedback
from target; 7, neural signals providing long-loop
negative feedback from target; 8, extent to which
signal is amplified can be modulated by short-term
modification of the sensitivity of the target, by
long-term changes in number of receptors, and by
other means such as release of cotransmitter.
(Modified from Stricker and Zigmond, 1986.)

Figure 9-8. Postulated pathways by which

dopamine and glutamate may regulate dopamine
and cAMP-regulated phosphoprotein of 32 kDa
(DARPP-32) phosphorylation. Dopamine, by acting
on the D1 class of receptors, stimulates adenylyl
cyclase via a G protein to increase cyclic
adenosine monophosphate (cAMP) formation and
the activity of cAMP-dependent protein kinase
(protein kinase A, PKA), leading to
phosphorylation of DARPP-32. Phosphorylation
converts this phosphoprotein into a potent
inhibitor of protein phosphatase-1. Inhibition of
protein phosphatase-1 increases the
phosphorylation state of numerous
phosphoproteins involved in the regulation of
important physiological processes. Dopamine, via
D2 receptors, decreases DARPP-32
phosphorylation by two intracellular signaling
pathways. One mechanism involves inhibition of
adenylyl cyclase, a decrease in cAMP, a decrease
in the activity of PKA, and decreased
phosphorylation of DARPP-32. The other D2-
mediated effect involves an increase in
intracellular calcium, activation of calcineurin, and
increased dephosphorylation of phospho-DARPP-
32. Glutamate acting on N-methyl-D-aspartate
(NMDA) receptors also gives rise to a large influx
of calcium. Thus, stimulation of NMDA receptors
can also lead to activation of calcineurin and
enhanced dephosphorylation of DARPP-32,
producing an effect very similar to stimulation of
 D2 receptors.

Effect of Drugs on the Electrophysiological

Activity of Noradrenergic Neurons
A number of drugs influence the activity of the noradrenergic
neurons in the locus ceruleus. Table 8-2 lists the drugs that
have been studied and their influence on locus cell firing.
Amphetamine inhibits locus cell firing, apparently by
activating a neuronal feedback loop. The effects of
amphetamine on locus firing are partially blocked by
chlorpromazine. L-Amphetamine is equipotent with D-
amphetamine in its inhibitory effects. -Receptor agonists and
antagonists also influence locus cell firing. This is due in part
to the interaction of these adrenergic agents with
autoreceptors on the NE cell bodies or dendrites in the locus.
2-Agonists, such as clonidine and guanfacine, suppress locus
cell firing, and these inhibitory effects are reversed by 2
antagonists. Piperoxane, yohimbine, and idazoxane ( 2
antagonists) cause a marked increase in single-cell activity.
Morphine and morphine-like peptides, such as enkephalins
and endorphins, exert an inhibitory influence on locus cell
firing. The inhibitory effects of morphine or enkephalin are
reversed by the opiate antagonist naloxone but are
uninfluenced by 2 antagonists. Methylxanthines, such as
isobutyl-methylxanthines, which induce a quasi-opiate
withdrawal syndrome, cause an increase in locus cell firing;
this increase is antagonized by 2 agonists, such as clonidine.
Locus ceruleus cells in chronic morphine-treated rats
dramatically increase locus cell firing during naloxone-induced
withdrawal, and this increase in firing is suppressed by
clonidine. These results, viewed in conjunction with the
clinical data demonstrating the effectiveness of treating opiate
withdrawal with clonidine, have suggested that NE
hyperactivity may be an important component of the opiate
withdrawal syndrome in humans. These studies have thus
provided new insight into the rational design of a new class of
drugs ( 2 agonists) to treat opiate withdrawal, although the
role of the locus ceruleus in this treatment effect has not been
established.

Numerous pharmacological studies, conducted mostly in

rodents, have demonstrated a good correlation between drug-
induced changes in the firing rate of locus ceruleus neurons
monitored by extracellular single-unit recording and
alterations in brain levels of MHPG. For example, drug-
induced suppression of central noradrenergic activity
produced by administration of clonidine or tricyclic
antidepressants is accompanied by a reduction in MHPG. The
2 antagonists (e.g., idazoxane, piperoxane, or yohimbine) or
experimental conditions (e.g., stress or naloxone-precipitated
withdrawal) that cause an increase in noradrenergic activity
produce an increase in brain levels of MHPG. The magnitude
of the increase in MHPG produced by naloxone-precipitated
morphine withdrawal exceeds that produced by administration
of supramaximal doses of 2 antagonists such as piperoxane
or yohimbine. The biochemical observation is consistent with
electrophysiological studies indicating that the increase in
locus ceruleus cell activity produced during naloxone-
precipitated withdrawal exceeds that achieved by
administration of 2 antagonists such as piperoxane and
yohimbine. Acute severe stress increases the turnover of NE
in the CNS, apparently as a result of increased impulse flow in
noradrenergic neurons. If impulse flow in noradrenergic
neurons projecting to the cortex is acutely interrupted by
destruction of the locus ceruleus, the increase in NE turnover
and accumulation of MHPG-sulfate induced by stress are
completely blocked. Also, microdialysis experiments have
demonstrated stress-induced increases in NE release in
hipocampus and frontal cortex. The stress induced activation
of NE release and turnover can be blocked or attenuated with
2 agonists such as clonidine or guanfacine.

Since MHPG levels in brain measured under controlled

experimental conditions provide an index of physiological
activity in the locus ceruleus, measures of this metabolite in
accessible body fluids may be useful for the assessment of
changes in central NE function in intact animals or patients.
Clinical researchers are currently monitoring this metabolite in
CSF, plasma, and urine in the hope that this will provide
insight concerning alterations in central noradrenergic
function.

Physiological Functions of Central

Noradrenergic Neurons
Many functions have been proposed for the central NE
neurons and their several sets of synaptic connections.
Among the hypotheses that have the most supportive data
are those concerning their role in affective psychoses
(described below), learning and memory, reinforcement,
sleep-wake cycle regulation, and the anxiety-nociception
hypothesis. It has also been suggested that a major function
of central noradrenergic neurons is not on neuronal activity
and related behavioral phenomena at all but, rather, a more
general role in cerebral blood flow and metabolism. However,
available data fit better into a more general proposal: the
main function of the locus ceruleus and its projections is to
determine the brain's global orientation concerning events in
the external world and within the viscera. Such a hypothesis
of central noradrenergic neuron function has been generated
by observations of locus ceruleus unit discharge in untreated,
awake, behaviorally responsive rats and monkeys. These
studies reveal the locus ceruleus units to be highly responsive
to a variety of nonnoxious sensory stimuli and that the
responsiveness of these units varies as a function of the
animal's level of behavioral vigilance. Increased neuronal
activity in the locus ceruleus is associated with unexpected
sensory events in the subject's external environment, while
decreased noradrenergic activity is associated with behaviors
that mediate tonic vegetative behaviors. Such a global-
orienting function can also incorporate other aspects of
presumptive function expressed by earlier data, but none of
those more discrete functions can be documented as
necessary or sufficient explanations of locus ceruleus
function.

Pharmacology of Adrenergic Neurons

Limited experiments have suggested that the pharmacology
of central adrenergic neurons is similar to that of central
noradrenergic neurons. Agents that block TH, DOPA-
decarboxylase, and dopamine- -hydroxylase lead to a
reduction of both NE and E in the brain. Depleting agents
such as reserpine, which cause release of NE and DA, also are
effective at releasing E. MAO inhibitors cause an elevation of
NE, DA, and E; but inhibitors of MAO-A are much more
effective at elevating E. In fact, most of the pharmacological
data are consistent with the hypothesis that, at least in the
rat hypothalamus, oxidative deamination is an important
metabolic process by which E is degraded and that MAO-A is
predominantly involved in this degradation. Similar to
observations made in central noradrenergic neurons, 2
agonists such as clonidine decrease E formation and 2
antagonists increase E turnover. These data are also
consistent with the possibility that 2 receptors are involved
in the regulation of synthesis and release of E and perhaps in
the control of the functional activity of adrenergic neurons.

SELECTED REFERENCES
Abell, C. W. and S. W. Kwan (2000). Molecular
characterization of monoamine oxidases A and B. Prog.
Nucleic Acid Res. Mol. Biol. 65, 129-156.

Amara, S. G. (1995). Monoamine transporters: basic biology

with clinical implications. Neuroscience 1, 259-268.

Aston-Jones, G., J. Rajkowski, and J. Cohen (1999). Role of

locus coeruleus in attention and behavioral flexibility. Biol.
Psychiatry 46, 1309-1320.

Blakely, R. D. and A. L. Bauman (2000). Biogenic amine

transporters: regulation in flux. Curr. Opin. Neurobiol. 10,
328-336.

Bonisch, H. and M. Bruss (1993). The noradrenaline

transporter of the neuronal plasma membrane. Ann. N.Y.
Acad. Sci. 733, 193-202.

Bylund, D. B., D. C. Eikenberg, J. P. Hieble, S. Z. Langer, R. J.

Lefkowitz, K. P. Minneman, P. B. Molinoff, R. R. Ruffolo, Jr.,
and U. Trendelenburg (1994). International Union of
Pharmacology Nomenclature of Adrenoceptors. Pharmacol.
Rev. 46, 121-136.

Charney, D. S., J. D. Bremner, and D. E. Redmond, Jr. (1995).

Noradrenergic neuronal substrates for anxiety and fear:
clinical associations based on preclinical research. In
Psychopharmacology: The Fourth Generation of Progress (F.
E. Bloom and D. J. Kupfer, eds.). Raven Press, New York, pp.
387-396.

Dahlstrom, A. and A. Carlsson (1986). Making visible the

invisible. In Discoveries in Pharmacology. Pharmacology
Methods, Receptors and Chemotherapy, vol. 3 (M. J. Parnham
and J. Bruinvels, eds.). Elsevier, Amsterdam, pp. 97-125.
Duman, R. S. and E. J. Nestler (1995). Signal transduction
pathways for catecholamine receptors. In
Psychopharmacology: The Fourth Generation of Progress (F.
E. Bloom and D. J. Kupfer, eds.). Raven Press, New York, pp.
303-320.

Foote, S. L. and G. S. Aston-Jones (1995). Pharmacology and

physiology of central noradrenergic systems. In
Psychopharmacology: The Fourth Generation of Progress (F.
E. Bloom and D. J. Kupfer, eds.). Raven Press, New York, pp.
335-346.

Fuller, R. W. (1982) Pharmacology of brain epinephrine

neurons. Annu. Rev. Pharmacol. Toxicol. 22, 31-55.

Goldstein, M. (1995). Long- and short-term regulation of

tyrosine hydroxylase. In Psychopharmacology: The Fourth
Generation of Progress (F. E. Bloom and D. J. Kupfer, eds.).
Raven Press, New York, pp. 189-196.

Hokfelt, T., O. Johansson, and M. Goldstein (1984). Central

catecholamine neurons as revealed by immunohistochemistry
with special reference to adrenaline neurons. In Handbook of
Chemical Neuroanatomy. Classical Transmitters in the CNS,
vol. 2, part I (A. Bjorklund, and T. Hokfelt, eds.), Elsevier,
Amsterdam, pp. 157-259.

Holmes, P. V. and J. N. Crawley (1995). Coexisting

neurotransmitters in central noradrenergic neurons. In
Psychopharmacology: The Fourth Generation of Progress (F.
E. Bloom and D. J. Kupfer, eds.). Raven Press, New York, pp.
347-354.

Kumer, S. C. and K. E. Brana (1996). Intricate regulation of

tyrosine hydroxylase activity and gene expression. J.
Neurochem. 67, 443-462.
Lindvall, O. and A. Bjorklund (1974). The organization of the
ascending catecholamine neuron systems in the rat brain as
revealed by the glyoxylic acid fluorescence method. Acta
Physiol Scand. 412, 1-47.

Lindvall, O. and A. Bjorklund (1983). Dopamine and

norepinephrine containing neuron systems: their anatomy in
the rat brain. In Chemical Neuroanatomy (P. C. Emson, ed.).
Raven Press, New York.

Maas, J. W., ed. (1983). MHPG: Basic Mechanisms and

Psychopathology. Academic Press, New York.

Mefford, I. N. (1988). Epinephrine in mammalian brain. Prog.

Neuropsychopharmacol. Biol. Psychiatry 12, 365-388.

Melikian, H. E., J. K. McDonald, H. Gu, G. Rudnick, K. R.

Moore, and R. D. Blakely (1994). Human norepinephrine
transporter: biosynthetic studies using a site-directed
polyclonal antibody. J. Biol. Chem. 269, 12290-12297.

Milligan, G., P. Svoboda, and C. M. Brown (1994). Why are

there so many adrenoceptor subtypes? Biochem. Pharmacol.
48, 1059-1071.

Moore, R. Y. and F. E. Bloom (1979). Central catecholamine

neuron systems: anatomy and physiology of the
norepinephrine and epinephrine systems. Annu. Rev.
Neurosci. 2, 113.

Nagatomo, T. and K. Koike, (2000). Recent advances in

structure, binding sites with ligands and pharmacological
function of beta-adrenoceptors obtained by molecular biology
and molecular modeling. Life Sci. 66, 2419-2426.

Nagatsu, T. and H. Ichinose (1999). Regulation of pteridine-

requiring enzymes by the cofactor tetrahydrobiopterin. Mol.
Neurobiol. 19, 79-96.

Redmond, D. E., Jr. (1987). Studies of the nucleus locus

coeruleus in monkeys and hypothesis for
neuropsychopharmacology. In Psychopharmacology: The
Third Generation of Progress (H. Y. Meltzer, ed.). Raven Press,
New York, pp. 967-975.

Reimer, R. J., E. A. Fon, and R. H. Edward (1998). Vesicular

neurotransmitter transport and the presynaptic regulation of
quantal size. Curr. Opin. Neurobiol. 8, 405-412.

Robbins, T. W. and B. J. Everitt (1995). Central

norepinephrine neurons and behavior. In
Psychopharmacology: The Fourth Generation of Progress (F.
E. Bloom and D. J. Kupfer, eds.). Raven Press, New York, pp.
363-372.

Schuldiner, S., A. Shirvan, and M. Linial (1995). Vesicular

neurotransmitter transporters: from bacteria to humans.
Physiol. Rev. 75, 369-392.

Tam, S.-Y. and R. H. Roth (1997). Commentary: Mesprefrontal

dopaminergic neurons: Can tyrosine availability influence the
function of the mesoprefrontal dopaminegic neurons?
Biochem. Pharmacol. 53, 441-453.

Thony, B., G. Auerbach, and N. Blau (2000).

Tetrahydrobiopterin biosynthesis, regeneration and functions.
Biochem. J. 347, 1-16.

Valentino, R. J. and G. S. Aston-Jones (1995). Physiological

and anatomical determinants of locus coeruleus discharge:
behavioral and clinical implications. In Psychopharmacology:
The Fourth Generation of Progress (F. E. Bloom and D. J.
Kupfer, eds.). Raven Press, New York, pp. 373-386.
von Euler, U. S. (1956). Noradrenaline. Charles C. Thomas,
Springfield, IL.

Yamamoto, B. K. and S. Novotney (1998). Regulation of

extracellular dopamine by the norepinephrine transporter. J.
Neurochem. 71, 274-280.

(+/-) Show / Hide Bibliography

Send Feedback Teton Server (5.0.0) - ©2007 Teton Data Title Updates
Customer Service Systems User
800.901.5494 Send Us Your Comments Responsibilities
Training Center
What's New
9. Dopamine
INTRODUCTION
Dopamine is the most recently discovered catecholamine
transmitter in the mammalian brain. Until the mid-1950s it
was exclusively considered to be an intermediate in the
biosynthesis of the catecholamines norepinephrine and
epinephrine. Significant tissue levels of dopamine were first
demonstrated in peripheral organs of ruminant species. A
short time later, Montagu, Carlsson, and co-workers found
that dopamine was also present in the brain in about equal
concentrations to those of norepinephrine but with a quite
different distribution. The very marked differences in regional
distribution of the two catecholamines dopamine and
norepinephrine, both within the central nervous system (CNS)
and in bovine peripheral tissue, led Swedish investigators to
propose a biological role for dopamine independent of its
function as a precursor for norepinephrine biosynthesis.
Studies demonstrating that most brain dopamine is confined
to the basal ganglia led to the hypothesis that it might be
involved in motor control and that decreased striatal
dopamine could be the cause of extrapyramidal symptoms in
Parkinson's disease. The discovery of profound depletions of
dopamine in the striatum of parkinsonian patients and the
demonstration that L-dihydroxyphenylalanine (L-DOPA) has
beneficial effects in these patients substantiated the clinical
relevance of this theory. These and other largely
pharmacological studies were the impetus for the almost
explosive developments in dopamine research during the past
four decades. With the development of histochemical methods
for the visualization of dopamine or its main synthetic enzyme
in brain tissue, the anatomy of brain dopamine systems could
also be described, paving the way for more direct studies on
this neurotransmitter. This research ultimately culminated in
the award of the Nobel Prize in 2000 to three scientists,
Carlsson, Greengard, and Kandel for their seminal
contributions to this field.

DOPAMINERGIC SYSTEMS
The central dopamine-containing systems are considerably
more complex in their organization than the noradrenergic
and adrenergic systems. Not only are there many more
dopamine cells (the number of mesencephalic dopamine cells
has been estimated at about 15,000-20,000 on each side,
while the number of noradrenergic neurons in the entire brain
stem is reported to be about 5000 on each side) but there are
also several major dopamine-containing nuclei. From
anatomical studies of the dopamine systems in the 1970s
(Fig. 9-1), these systems have been divided into three major
categories based on the length of the efferent dopamine
fibers.

1. Ultrashort systems. Among the ultrashort systems are the

interplexiform amacrine-like neurons, which link the inner and
outer plexiform layers of the retina, and the periglomerular
dopamine cells of the olfactory bulb, which link together
mitral cell dendrites in separated adjacent glomeruli. These
neurons make extremely localized connections.

2. Intermediate-length systems. The intermediate-length

systems include (1) the tuberohypophysial dopamine cells,
which project from arcuate and periventricular nuclei into the
intermediate lobe of the pituitary and into the median
eminence (often referred to as the tuberoinfundibular
system); (2) the incertohypothalamic neurons, which link the
dorsal and posterior hypothalamus with the dorsal anterior
hypothalamus and lateral septal nuclei; and (3) the medullary
periventricular group, which includes those dopamine cells in
the perimeter of the dorsal motor nucleus of the vagus nerve,
the nucleus tractus solitarius, and the cells scattered in the
tegmental radiation of the periaqueductal gray matter.
3. Long systems. The long systems are the long projections
linking the ventral tegmental (A8, A10) and substantia nigra
(A9) dopamine cells with three principal sets of targets: the
neostriatum (principally the caudate and putamen); the limbic
cortex (medial prefrontal, cingulate, and entorhinal areas);
and other limbic structures (the regions of the septum,
olfactory tubercle, nucleus accumbens septi, amygdaloid
complex, and piriform cortex). These latter two groups have
frequently been termed the mesocortical and mesolimbic
dopamine projections, respectively. Under certain conditions,
these limbic target systems, when compared to the
nigrostriatal system, exhibit some unique pharmacological
properties, which are discussed in detail in the latter portion
of this chapter. When dopamine systems were first visualized
in the CNS, it was thought that all dopamine cells within the
zona compacta of the substantia nigra projected to the
caudate putamen nucleus. Dopamine cells in the ventral
tegmental area surrounding the interpeduncular nucleus were
believed to project exclusively to parts of the limbic system.
This beautiful simplicity lasted a relatively short time. Soon,
dopamine inputs to the cortex were discovered; and within a
few years, primarily through the use of retrograde tracing
techniques, it became apparent that dopamine cells within the
A8, A9, and A10 areas form an anatomically heterogeneous
population in terms of their projection areas (Bjorklund and
Lindvall, 1984). Retrograde tracing techniques also led to
another important discovery, that dopamine cells project
topographically to the areas that they innervate. Thus,
although there is some overlap, dopamine cells that are near
each other in a given area are more likely to innervate a
common region than are dopamine cells distant from each
other.

With regard to cellular analysis of function, the majority of

reported data are from studies of the nigrostriatal dopamine
projection. Here, most studies using the iontophoretic
administration of dopamine indicate that the predominant
qualitative response is inhibition. This effect, like that of
apomorphine and cyclic adenosine monophosphate (cAMP), is
potentiated by phosphodiesterase inhibitors, providing the
physiological counterpart to the second-messenger hypothesis
suggested by biochemical studies. However, the effects on the
properties of caudate neurons when electrical stimulation is
applied to the ventral tegmentum and substantia nigra are
considerably less homogeneous, ranging from excitation to
inhibition with considerable variations in latency and
sensitivity to dopamine antagonism. In one study, neither the
excitations nor the inhibitions elicited by nigral stimulation
were prevented by 6-hydroxydopamine-induced destruction of
the dopamine cell bodies. Unambiguous pharmacological and
electrophysiological analyses of the dopamine pathway thus
remain to be accomplished. Although at present the bulk of
the evidence favors an inhibitory role for dopamine,
inquisitive students will want to examine the arguments
raised by both sides and make their own evaluations of the
effects reported (Grenhoff and Johnson, 1997). Important
basic information is needed to rule out the spread of current
to nearby nondopamine tracts and to determine accurately
the latency of conduction reported for these extremely fine
unmyelinated fibers. This may be a mute point since the
actions of dopamine can be better described not in terms of
inhibition or excitation, but rather as related to the gating of
inputs and modulation of the states of neuronal activity of
postsynaptic follower cells. The modulation of the integration
of information can then be further influenced at the network
level via the actions of dopamine on interneurons or cell
coupling. This arrangement is consistent with the behavioral
actions of dopamine. Dopamine does not directly produce
reward or motor activity but instead modulates inputs and
adjusts the state of the organism in order to redirect the
stimulus response output to achieve the most effective
behavioral outcome.
It has become apparent that midbrain dopamine neurons are
quite heterogeneous in terms of their biochemistry,
physiology, pharmacology, and regulatory properties when
compared to the prototypic nigrostriatal system, in which
most of the earlier studies were performed. While midbrain
dopamine neurons differ in a number of important ways, their
functional organization generally reflects features of
transmitter dynamics that are shared by all dopamine
neurons. These features have been most thoroughly studied
in the nigrostriatal pathway and are summarized below (see
Fig. 9-2).

Figure 9-1. Schematic diagram illustrating the

distribution of the main central neuronal pathways
containing dopamine. Stippled regions indicate
major nerve terminal areas and their cell groups
of origin. Cell groups are named according to the
nomenclature of Dahlstrom and Fuxe (1965).

Figure 9-2. Schematic model of a prototypic

dopaminergic nerve terminal illustrating the life
cycle of dopamine (DA) and the mechanisms that
modulate its synthesis, release, and storage.
Invasion of the terminal by a nerve impulse
results in the Ca2+-dependent release of
dopamine. This release process is attenuated by
release-modulating autoreceptors. Increased
impulse flow also stimulates tyrosine
hydroxylation. This appears to involve the
phosphorylation of tyrosine hydroxylase (TH),
resulting in conversion to an activated form with
greater affinity for tetrahydrobiopterin cofactor
and reduced affinity for the end-product inhibitor
dopamine. The rate of tyrosine hydroxylation can
be attenuated by (1) activation of synthesis-
modulating autoreceptors, which may function by
reversing the kinetic activation of TH, and (2) end-
product inhibition by intraneuronal dopamine,
 product inhibition by intraneuronal dopamine,
which competes with cofactor for a binding site on
TH. Release- and synthesis-modulating
autoreceptors may represent distinct receptor
sites. Alternatively, one site may regulate both
functions through distinct transduction
mechanisms. The plasma membrane dopamine
transporter is a unique component of the
dopamine terminal that serves an important
physiological role in the inactivation and recycling
of dopamine release into the synaptic cleft. The
vesicular monoamine transporter (VAT) transports
cytoplasmic dopamine into storage vesicles,
decreasing the cytoplasmic concentration of
dopamine and preventing metabolism by
monoamine oxidase. VAT modulates the
concentration of free dopamine in the nerve
terminals.

DOPAMINE SYNTHESIS
Dopamine synthesis originates from tyrosine, and its rate-
limiting step is the conversion of L-tyrosine to L-DOPA by the
enzyme tyrosine hydroxylase (TH). DOPA is subsequently
converted to dopamine by L-aromatic amino acid
decarboxylase at rates so rapid that DOPA levels in the brain
are negligible under normal conditions. Because endogenous
levels of DOPA are normally low in the brain, the formation of
dopamine can be enhanced dramatically by providing this
enzyme with increased amounts of L-DOPA. Since the levels
of tyrosine in the brain are relatively high and above the Km
for TH, under normal conditions it is not feasible to augment
dopamine synthesis significantly by increasing brain levels of
this amino acid. Endogenous mechanisms for regulating the
rate of dopamine synthesis in dopamine neurons primarily
involve modulation of TH activity through four major
regulatory influences:

1. Dopamine and other catecholamines function as end-

product inhibitors of TH by competing with a
tetrahydrobiopterin (BH4) cofactor for a binding site on the
enzyme.

2. The availability of BH4 may also play a role in regulating

TH activity. Since endogenous levels of BH4 are controlled by
guanosine triphosphate (GTP) cyclohydrolase activity, this
rate-limiting enzyme in BH4 synthesis can indirectly influence
tyrosine hydroxylation. TH can exist in two kinetic forms,
which exhibit different affinities for BH4. Conversion from low-
to high-affinity forms is thought to involve direct
phosphorylation of the enzyme, and the proportion of TH
molecules in the high-affinity state appears to be a function of
the state of neuronal firing.

3. Presynaptic dopamine receptors also modulate the rate of

tyrosine hydroxylation. These receptors are activated by
dopamine released from the nerve terminal, resulting in
feedback inhibition of dopamine synthesis. Autoreceptors can
modulate both synthesis and release of dopamine and
represent important sites for the pharmacological
manipulation of dopaminergic function by dopamine agonists
and antagonists.

4. Dopamine synthesis in the striatum also depends on the

rate of impulse flow in the nigrostriatal pathway. During
periods of increased impulse flow, the rate of tyrosine
hydroxylation is increased primarily through kinetic activation
of TH, which increases its affinity for BH4 and decreases its
affinity for the normal end-product inhibitor dopamine. Under
conditions of increased impulse flow, tyrosine hydroxylation is
also more susceptible to precursor regulation by tyrosine
availability.
Calcium-dependent release of dopamine from the nerve
terminal is thought to occur in response to invasion of the
terminal by an action potential. The extent of dopamine
release appears to be a function of the rate and pattern of
firing. The burst-firing mode leads to enhanced release of
dopamine. Dopamine release is also modulated by
presynaptic release-modulating autoreceptors. In general,
dopamine agonists inhibit while dopamine antagonists
enhance the evoked release of dopamine.

DOPAMINE UPTAKE AND THE DOPAMINE

TRANSPORTER
Dopamine nerve terminals possess high-affinity dopamine-
uptake sites, which are important in terminating transmitter
action and maintaining transmitter homeostasis. Uptake is
accomplished by a membrane carrier, the dopamine
transporter (DAT), which can transport dopamine into and out
of the terminal depending on the existing concentration
gradient. Substantial progress in the 1980s led to the
development of a new class of very potent and selective
dopamine-uptake inhibitors, the GBR series (Fig. 9-3). With
these drugs, the stage was set for the isolation and molecular
characterization of the DAT, successfully achieved in 1991 by
several groups.

The DNA encoding the rat DAT exhibits high sequence

similarity with the previously cloned norepinephrine and -
aminobutyric acid transporters. The DAT is a 619-amino acid
protein with 12 putative hydrophobic membrane-spanning
domains and is a member of the family of Na+/Cl--dependent
plasma membrane transporters. Using the energy provided by
the Na+ gradient generated by the Na+/K--transporting
adenosine triphosphatase (ATPase), the DAT recaptures
dopamine soon after its release, thereby modulating its
concentration in the synapse and its time-dependent
interaction with both pre- and postsynaptic receptors.
Molecular characterization and cloning of rat, bovine, and
human DATs have shown that these proteins are highly
conserved between species with similar orientation in the
plasma membrane and potential sites of glycosylation and
phosphorylation (Fig. 9-4).

A number of studies have suggested that DAT is a useful

phenotypic marker for dopamine neurons and their nerve
terminals and perhaps even better in some cases than TH.
Nevertheless, caution should be exercised in the use of this
DAT marker since its expression varies significantly among
dopamine cell groups. The tuberoinfundibular dopamine
neurons (A12, see Fig. 9-1), which release dopamine into the
pituitary portal blood system, lack demonstrable DAT mRNA
and protein. Because dopamine released from
tuberoinfundibular neurons is carried away rapidly in the
vascular system, the existence of a transporter protein on
these dopamine neurons seems superfluous.

During development, embryonic midbrain dopamine neurons

express dopamine and TH well before they express DAT.
Although the catecholamine transporters have highly similar
molecular features, they exhibit important differences in their
selectivity for their catecholamines and for neurotoxins like 1-
methyl-4-phenylpyridinium (MPP+) and very distinct
pharmacologies (cf. Figs. 9-3 and 8-6).

Immunohistochemical studies of the subcellular localization of

these transporters led to an unexpected finding. The use of
antibodies generated against DAT revealed that the
transporter is typically expressed outside of the synapse, in
the extrasynaptic region of the axon terminal. This suggests
that the transporter may be used to inactivate (accumulate)
dopamine that has escaped from the synaptic cleft and, thus,
that diffusion is the initial process by which dopamine is
removed from the synapse. This is consistent with recent in
situ studies indicating that perisynaptic concentrations of
dopamine can reach approximately 1.0 mM. Receptors for
dopamine and many other transmitters are also found
extrasynaptically (indeed, along the length of axons); this
observation coupled with the presence of catecholamine
transporters in extrasynaptic regions suggests that
extrasynaptic ("paracrine" or "volume") neurotransmission
may be of considerable importance for catecholaminergic
signaling.

Mesolimbic dopamine neurons are implicated in the

reinforcing properties of a variety of drugs of abuse, including
psychomotor stimulants such as cocaine and amphetamine.
Cocaine and related drugs bind to DAT and prevent dopamine
transport in a fashion that correlates well with their behavioral
reinforcing and psychomotor-stimulating properties. In fact,
DATs have often been referred to as the brain's principal
"cocaine receptors."

Receptor-binding studies have demonstrated that compounds

that bind to the DAT also inhibit dopamine uptake, with a rank
order of potency proportional to the affinity demonstrated in
binding studies. This relationship of uptake inhibitory potency
and binding potency suggests that the two processes may be
intimately linked and that any compound that binds to DAT
will also block dopamine transport. However, point mutation
studies of the cloned DAT indicate that reuptake inhibition and
binding potency may be distinct processes that, under certain
conditions, are separable, not inextricably linked. These and
other studies on chimeric DAT proteins have revealed that the
cocaine-binding site on DAT is distinct from the substrate-
recognition site. These observations suggest that it may be
possible to develop agents that can prevent binding of
stimulants like cocaine to the DAT while still allowing normal
dopamine transport to ensue, thus supporting the feasibility
of developing cocaine antagonists for the treatment of drug
overdose, withdrawal, or addiction.

The DAT has also assumed importance in the study of 1-

methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-induced
and idiopathic Parkinson's disease. The selectivity of
dopamine neurotoxins like MPTP seems to depend on their
high affinity for the DAT. In primates, MPTP toxicity can be
prevented by pretreatment with DAT inhibitors, but once
transported into the neuron, the toxin destroys the dopamine
neurons, ultimately producing parkinsonism (see Chapter 13).
Expression of the cloned DAT in COS cells confers sensitivity
to MPP+ toxicity, while expression of a vesicular transporter
clone confers resistance to MPP+ in sensitive CHO cells. Thus,
the levels of vesicular transporter and DAT expression in
combination could conceivably dictate the response to
exogenous or endogenously generated neurotoxins.
Interestingly, regional differences in the levels of DAT
expression appear to correlate with the extent of dopamine
cell loss after MPTP treatment or in Parkinson's disease.

Figure 8-6. Representative compounds that

selectively inhibit norepinephrine reuptake.

Figure 9-3. Chemical structures of some

dopamine-uptake inhibitors.

Figure 9-4. Schematic diagram of the primary

amino acid sequence of the human dopamine
transporter. Amino acid residues in black are those
that differ between rat and human sequences. Y-
shaped symbols represent potential sites for N-
glycosylation. Boxed residues represent consensus
phosphorylation sites for protein kinase A (PKA),
 protein kinase C (PKC), or Ca2+-calmodulin-
dependent kinase II (CaM kinase II). Disposition
of the transporter protein with respect to the
plasma membrane is putative. Two leucine zipper
motifs are indicated. (From Giros and Caron,
1993.)

DOPAMINE METABOLISM
Released dopamine is converted to dihydroxyphenylacetic acid
(DOPAC) by intraneuronal monoamine oxidase (MAO) after
reuptake by the nerve terminal. Released dopamine is also
converted to homovanillic acid (HVA), probably at an
extraneuronal site, through the sequential action of catechol-
O-methyltransferase (COMT) and MAO. In rat brain, DOPAC is
the major metabolite and considerable amounts of DOPAC
and HVA are present in sulfate-conjugated as well as free
forms. In humans and other primates, however, the major
brain metabolite is HVA, of which only a small amount is
found in the conjugated form.

The primary metabolites of dopamine in the CNS are HVA,

DOPAC, and a small amount of 3-methoxytyramine (3-MT). In
dopamine systems, in contrast to norepinephrine systems
(see Chapter 8), the acidic rather than the neutral
metabolites appear to predominate. Accumulation of HVA in
the brain or cerebrospinal fluid (CSF) has often been used as
an index of the functional activity of dopaminergic neurons in
the brain. 3-MT is also a useful index, provided precautions
are taken to minimize postmortem increases in this
metabolite. Antipsychotic drugs, which increase the turnover
of dopamine (in part because of their ability to increase the
activity of dopaminergic neurons and to augment dopamine
release), also increase the amount of HVA and 3-MT in the
brain and CSF. In addition, electrical stimulation of the
nigrostriatal pathway increases brain levels of HVA and 3-MT
(normally quite low) as well as the release of HVA into
ventricular perfusates. In Parkinson's disease, where
substantia nigra dopamine neurons die, reduced HVA is
observed in the CSF. Similar changes are observed in MPTP-
induced parkinsonism in humans and nonhuman primates.

In rat brain, short-term accumulation of DOPAC in the

striatum can be taken as an accurate reflection of activity in
nigrostriatal dopaminergic neurons. Cessation of impulse flow
after the placement of acute lesions in the nigrostriatal
pathway leads to a rapid decrease in striatal DOPAC.
Conversely, electrical stimulation of the nigrostriatal pathway
results in a frequency-dependent increase in DOPAC within
the striatum. Drugs that increase impulse flow in the
nigrostriatal pathway, such as the antipsychotic
phenothiazines and butyrophenones, anesthetics, and
hypnotics, also increase striatal DOPAC. Drugs that block or
decrease impulse flow, such as -hydroxybutyrate, ( )3-
amino-1-hydroxypyrrolid-2-one [( )HA-966], apomorphine,
and amphetamine, reduce DOPAC levels. Thus, there appears
to be an excellent correlation between changes in impulse
flow in dopaminergic neurons, which are induced either
pharmacologically or mechanically, and changes in the
steady-state levels of DOPAC. In primates, DOPAC is a minor
brain metabolite. Not only is it difficult to measure DOPAC in
CSF but in primate brain this metabolite, in contrast to HVA,
is unresponsive to drug treatments that cause large changes
in dopamine metabolism.

By means of the sensitive and specific technique of gas

chromatography-mass fragmentography, it has been possible
to measure DOPAC, HVA, and their conjugates accurately in
rat and human plasma. Studies in rats have demonstrated
that stimulation of the nigrostriatal pathway and
administration of antipsychotic drugs increase plasma levels
of DOPAC and HVA. Several studies in humans have also
indicated that dopaminergic drugs can influence plasma levels
of HVA, although the effect is quite modest in comparison to
that observed in rodents, limiting their clinical usefulness.

FUNCTIONAL REGULATION
The synthesis and release of dopamine are clearly influenced
by the activity of dopaminergic neurons, but these neurons
behave differently from peripheral or central noradrenergic
neurons; indeed, differences exist between dopamine
neurons. Increased impulse flow in the nigrostriatal or
mesolimbic dopamine system does lead to both an increase in
dopamine synthesis and turnover and a frequency-dependent
increase in the accumulation of dopamine metabolites in the
striatum and olfactory tubercle. This parallels other
monoamine systems, where an increase in impulse flow
causes an increase in the synthesis and turnover of
transmitter.

Short-term stimulation of central dopaminergic neurons

increases tyrosine hydroxylation by kinetic alterations in TH,
with an increased affinity for pteridine cofactor and a
decreased affinity for the natural end-product inhibitor
dopamine. As in central noradrenergic systems, it seems that
a finite period of time is necessary for this activation to occur
and that, once activated, the enzyme remains in this altered
physical state for a short period after the stimulation ends.
The activation appears to involve TH phosphorylation.

However, if impulse flow is interrupted in the nigrostriatal or

mesolimbic dopamine system, either mechanically or
pharmacologically by treatment with -hydroxybutyrate, the
neurons respond in a rather peculiar fashion, by rapidly
increasing the concentration of dopamine in the nerve
terminals of the respective dopamine systems. Not only do
the terminals accumulate dopamine by reducing release but
there is also a dramatic increase in the rate of dopamine
synthesis. This increase occurs despite the steadily increasing
concentration of endogenous dopamine within the nerve
terminal.

The actual mechanisms whereby a cessation of impulse flow

initiates changes in the properties of TH are unclear, although
they may involve a decrease in the availability of intracellular
calcium. Similar changes in the activity or properties of TH
are not observed in central noradrenergic neurons or in other
dopamine neurons (e.g., the mesoprefrontal dopamine
neurons) lacking synthesis-modulating autoreceptors. At
present, the physiological significance of this paradoxical
response to a cessation of impulse flow is unclear. However, it
is conceivable that these neurons achieve some operational
advantage by increasing their supply of transmitter rapidly
during periods of quiescence.

POTENTIAL SITES OF DRUG ACTION ON

DOPAMINERGIC NEURONS

Introduction
There are many sites at which drugs can influence the
function of dopamine neurons. The potential sites for
modulation are illustrated in Figure 9-5 and summarized in
Table 9-1. For the purpose of this discussion, drug effects can
be divided into three broad categories:

1. Nonreceptor-mediated effects on presynaptic function

2. Dopamine receptor-mediated effects

3. Effects mediated indirectly as a result of drug interaction

with other neurotransmitter systems that interact with
dopamine neurons

The relative importance of each of these potential sites of

drug action will vary among different dopamine systems,
depending on factors such as the presence or absence of
autoreceptors, the efficiency of postsynaptic receptor-
mediated neuronal feedback pathways, and the nature of the
afferent inputs impinging on the dopamine neurons in
question.

Figure 9-5. Schematic model of striatal

dopaminergic nerve terminal. Drugs which alter
the dopamine (DA) life cycle include (1) -
methyltyrosine, a competitive inhibitor of tyrosine
hydroxylase; (2) NSD-1015, an inhibitor of DOPA
decarboxylase; (3) reserpine, which irreversibly
damages DA uptake/storage mechanisms and
produces long-lasting depletion of DA; (4)
tetrabenazine, which also interferes with DA
uptake/storage but the effects are of shorter
duration than those of reserpine and do not
appear to be irreversible; (5) amphetamine, which
increases synaptic DA through a number of
mechanisms, including induction of release of DA
and blockade of DA reuptake; (6) cocaine, which
also blocks DA reuptake and induces DA release;
(7) nomifensine and GBR, which also block DA
reuptake but lack DA-releasing ability; (8)
pargyline, an inhibitor of monoamine oxidase
(MAO); (9) tropolone, an inhibitor of catechol-O-
methyltransferase (COMT). HVA, homovanillic
acid; MT, 3-methoxytyramine. DOPA,
dihydroxyphenylalanine; DOPAC
dihydrophenylacetic acid.

Nonreceptor-Mediated Effects
There are several stages in the life cycle of dopamine
(synthesis, storage, and release) where drugs can influence
transmitter dynamics, as illustrated in Figure 9-5. There are
many useful pharmacological tools for modifying
dopaminergic activity and manipulating dopaminergic function
at these sites, but most of these agents are not very selective
for dopaminergic synapses and will interact with other
catecholamine (norepinephrine and epinephrine) systems (in
some cases, with other monoamine [5-hydroxytryptamine]
systems as well). Some drugs which interact with the plasma
membrane transporter do not have a high degreee of
specificity. For example, amphetamine, cocaine, benztropine,
and nomifensine interact with the plasma membrane
transporter that normally functions in the reuptake of
released dopamine. However, these drugs also have an
appreciable affinity for noradrenergic (and in some cases
serotonergic) uptake sites.

Nevertheless, drugs that are highly selective for the dopamine

transport complex have been developed and employed as
valuable experimental tools for visualization of the integrity of
dopamine systems in vivo. In fact, striking results have been
obtained with several new cocaine derivatives, such as 3 -(4-
fluorophenyl)tropane-2 -carboxylate (CFT) and 3 -(4-iodo-
phenyl)tropane-2 -carboxylate ( -CIT), which exhibit high
affinity for the DAT (see Fig. 9-3). These agents have been
used in autoradiographic experiments and in positron
emission tomography (PET) and single-photon emission
computed tomography (SPECT) studies to image the striatal
DAT in both normal and parkinsonian monkeys and humans.
These studies have demonstrated (1) loss of striatal DAT in
both experimental and idiopathic Parkinson's disease and (2)
restoration of DAT density and improvement of behavioral
functions following nigral grafts in the caudate of transplanted
MPTP monkeys. DAT ligands used for in vivo imaging in the
future may be routinely employed in the diagnosis of certain
diseases like parkinsonism and for following the progression
of the disease and the response to treatment. Imaging of DAT
may also prove very useful for monitoring the viability of
dopamine grafts and their outgrowth once transplanted into
parkinsonian recipients.

Dopamine Receptor-Mediated Effects

Originally, it was thought that all drugs that affect dopamine
activity, including the neuroleptics, worked through
nonreceptor-mediated mechanisms such as those described
above. However, it is now clear that many therapeutically
important drugs interact with dopamine receptors. Drugs that
affect dopamine receptors can be classified into two groups
(Table 9-2):

1. Receptors on nondopamine cell types, which are usually

referred to as "postsynaptic" receptors since they are
postsynaptic to a dopamine-releasing cell

2. Receptors on dopamine cells, which are referred to as

"autoreceptors" to indicate their sensitivity to the neuron's
own transmitter

Postsynaptic Dopamine Receptors Postsynaptic

dopamine receptors can be classified as either D1 or D2,
based on the functional and pharmacological criteria
described below. Both types of receptor have been found in
the projection areas of midbrain dopamine neurons, although
it is unclear whether they are located on distinct subsets of
dopamine-receptive cells in various projection fields. In the
striatum, postsynaptic dopamine receptors regulate the
activity of neuronal feedback pathways by which striatal
neurons can communicate with dopamine cell bodies in the
substantia nigra. This enables dopamine-innervated cells in
the striatum to modulate the physiological activity of
nigrostriatal dopamine neurons. In general, increased
postsynaptic receptor stimulation results in decreased
nigrostriatal dopamine activity.
Following chronic exposure to dopamine agonists or
antagonists, postsynaptic dopamine receptors exhibit
adaptive changes. For example, chronic exposure to
dopamine antagonists or chemical denervation with 6-
hydroxydopamine (6-OHDA) produces an increase in the
number of dopamine-binding sites measured in receptor-
binding assays. This may be related to the behavioral
supersensitivity to dopamine agonists that also develops as a
result of chronic antagonist administration or denervation.
Conversely, repeated administration of dopamine agonists
decreases the number of dopamine-binding sites and
produces subsensitivity to subsequent administration of
dopamine agonists in behavioral as well as biochemical and
electrophysiological models. Changes such as these may be
relevant to understanding the state of dopamine receptors in
diseases believed to involve chronic dopaminergic hyper- or
hypoactivity.

Autoreceptors Autoreceptors can exist on most portions of

dopamine cells, including the soma, dendrites, and nerve
terminals. Stimulation of dopamine autoreceptors in the
somatodendritic region slows the firing rate of dopamine
neurons, while stimulation of autoreceptors on the nerve
terminals inhibits dopamine synthesis and release. Thus,
somatodendritic and nerve terminal autoreceptors work in
concept to exert feedback on dopaminergic transmission. Both
somatodendritic and nerve terminal autoreceptors can be
classified as D2 receptors and exhibit similar pharmacological
properties. Like postsynaptic receptors, somatodendritic and
nerve terminal autoreceptors develop supersensitivity after
chronic antagonist treatment or prolonged decreases in
dopamine release and desensitize in response to repeated
administration of dopamine agonists. Interestingly, the
autoreceptors are more readily desensitized than postsynaptic
dopamine receptors. This has been suggested to play a role in
the "on-off" effects observed during chronic L-DOPA therapy
in Parkinson's disease.

Dopamine autoreceptors can be defined functionally in terms

of the events they regulate and are therefore divided into
three categories: synthesis-modulating, release-modulating,
and impulse-modulating autoreceptors. However, it is not yet
known whether distinct receptor proteins modulate each of
these functions or whether the same receptor protein is
coupled to each function through distinct transduction
mechanisms. It is clear, however, that autoreceptor-mediated
pathways for the regulation of dopamine release from the
nerve terminal are distinct from autoreceptor-mediated
pathways for the regulation of dopamine synthesis since
dopamine terminals in the prefrontal and cingulate cortices
possess autoreceptors that regulate release but lack
functional synthesis-modulating autoreceptors.

Autoreceptors Versus Postsynaptic Receptors:

Pharmacological and Functional Considerations
Autoreceptors and postsynaptic dopamine receptors differ in
several ways. The most clear-cut difference is that
autoreceptors are 5-10 times more sensitive to the effects of
dopamine and apomorphine than postsynaptic dopamine
receptors in behavioral, biochemical, and electrophysiological
models. In the low-dose range, therefore, autoreceptor-
mediated effects of dopamine agonists predominate, resulting
in diminished dopaminergic function, while higher doses also
stimulate postsynaptic receptors, leading to enhanced
dopaminergic function.

Autoreceptors also differ from postsynaptic receptors in their

pharmacological profile. Dopamine agonists that are relatively
selective for autoreceptors have been synthesized. As would
be predicted, autoreceptor-selective agonists inhibit dopamine
release, synthesis, and impulse flow in dopamine neurons and
elicit behavioral responses associated with diminished
dopaminergic function. These agonists are very useful
experimental probes for studying dopamine receptor function
and may prove useful in diseases thought to involve excessive
dopaminergic activity. Dopamine antagonists that selectively
block autoreceptors have also been synthesized. By blocking
dopamine autoreceptors, they enhance dopamine function.
Some of these agents appear to have a built-in ceiling on
their response since as the dose is increased, they exert an
antagonistic action on postsynaptic dopamine receptors. The
structures of several of these autoreceptor-selective agents
are illustrated in Figure 9-6.

Dopamine agonists and antagonists may act on several types

of dopamine receptor to elicit biochemical changes in
dopamine metabolism and alter the functional output of
dopaminergic systems (Table 9-2). A drug's net effect on
dopaminergic activity will depend on both its pre- and
postsynaptic effects and the selectivity with which it acts at
these different sites.

Figure 9-6. Structures of some selective

dopamine autoreceptor ligands.

DRUG INTERACTIONS AT D1 AND D2

RECEPTORS
In the preceding discussion, dopamine receptors were broadly
divided into presynaptic and postsynaptic categories. A
second popular dopamine receptor classification that received
increasing attention in the 1980s is based on the presence or
absence of positive coupling between the receptor and
adenylate cyclase activity. On the basis of biochemical,
physiological, and pharmacological studies, it is now well
established that dopamine can act on at least two types of
brain receptor, termed D1 and D2 receptors (Table 9-3).
These two classes of receptors are clearly distinguished by
their biochemical characteristics. D1 receptors mediate the
dopamine-stimulated increase of adenylate cyclase activity.
D2 receptors are thought to mediate effects that are
independent of D1-mediated effects and to exert an opposing
influence on adenylate cyclase activity. The D2 site is further
characterized by picomolar affinity for antagonist, while the
D1 site is characterized by millimolar affinity for antagonist.
The arrangements of D1 and D2 receptors and the nerve
terminal autoreceptor are diagrammed in Figure 9-7.

With the development of D1- and D2-selective agonists and

antagonists, however, it has become common to rely on
pharmacological characteristics when determining whether an
effect is mediated by D1 or D2 receptors. The distinction
between pharmacological and functional definitions is
important because it is becoming clear that dopamine
receptors with the same pharmacological characteristics do
not necessarily have the same functional characteristics. For
example, dopamine receptors with D2 pharmacology are
present in both the striatum and nucleus accumbens but are
coupled to inhibition of adenylate cyclase only in the striatum.
Regional differences in coupling between dopamine receptors
and GTP-binding proteins have also been reported.
Furthermore, dopamine receptors can influence cellular
function through mechanisms other than stimulating or
inhibiting adenylate cyclase (Table 9-3). These may include
direct effects on potassium and calcium channels as well as
modulation of inositol phosphate production. It therefore
seems unrealistic to expect equivalence between
pharmacological and biochemical classifications of dopamine
receptor subtypes.

Figure 9-7. Schematic diagram depicting the

 anatomical arrangement of D1 and D2 receptors
and the nerve terminal autoreceptor. The D1 site
is positively coupled with adenylate cyclase via a
Gs protein. The D2 postsynaptic site is negatively
coupled with adenylate cyclase via a Gi protein,
stimulation of which leads to hyperpolarization.
The autoreceptors appear to exert their dopamine
synthesis- and release-modulating effects via a Gi
regulatory protein. Dopamine autoreceptors are
sensitive to very low (nanomolar) concentrations
of dopamine. Although it is not known whether the
D1 and D2 receptors are situated on the same
neuron, they appear to be functionally linked, as
depicted by the arrows. Innervated postjunctional
dopamine receptors operate in the low-affinity
state (D1 low and D2 low) in contrast to the D2
autoreceptor, which operates in the high-affinity
state (D2 high).

DOPAMINE SIGNALING
Although D1 and D2 receptors can have opposite effects on
adenylate cyclase activity, it is apparent that the physiological
significance of their interaction is more complex. While D1
and D2 agonists can have opposite effects on oral
movements, they also produce a synergistic increase in
locomotor activity and behavioral stereotypes under certain
circumstances. Electrophysiological experiments have
suggested that D1 receptor activation is required for full
postsynaptic expression of D2 effects. The interaction of the
D1 receptor with other neurotransmitter systems is still being
explored. Despite widespread interest in dopamine signaling,
little was known until recently about the molecular and
cellular basis for the action of dopamine on its target cells. A
phosphoprotein named DARPP-32 (dopamine and cAMP-
regulated phosphoprotein of 32 kDa) plays a key role in the
biology of dopaminoceptive neurons (Fig. 9-8). By acting on
the D1 receptors, dopamine stimulates adenylyl cyclase via a
G protein to increase cAMP formation and the activity of
cAMP-dependent protein kinase (protein kinase A, PKA),
leading to phosphorylation of DARPP-32 on a single threonine
residue. Phosphorylation converts this phosphoprotein into a
potent inhibitor of protein phosphatase-1. At least two
intracellular pathways that decrease DARPP-32
phosphorylation are involved in the modulation of dopamine
signaling via D2 receptors. One mechanism involves inhibition
of adenylyl cyclase, a decrease in cAMP, a decrease in the
activity of PKA, and a decrease in DARPP-32 phosphorylation.
The other D2-mediated effect involves an increase in
intracellular calcium and activation of calcineurin. One of the
actions of calcineurin is to dephosphorylate DARPP-32 and
thus relieve the inhibition of protein phosphatase-1.
Striatonigral neurons receive glutamate input from the
cerebral cortex as well as a rich dopamine innervation from
the substantia nigra. In these neurons, glutamate acting on
N-methyl-D-aspartate (NMDA) receptors also gives rise to a
large influx of calcium (Fig. 9-8). Thus, in this system,
stimulation of NMDA receptors also results in activation of
calcineurin and enhanced dephosphorylation of DARPP-32,
producing an effect very similar to the stimulation of D2
receptors.

Figure 9-8. Postulated pathways by which

dopamine and glutamate may regulate dopamine
and cAMP-regulated phosphoprotein of 32 kDa
(DARPP-32) phosphorylation. Dopamine, by acting
on the D1 class of receptors, stimulates adenylyl
cyclase via a G protein to increase cyclic
adenosine monophosphate (cAMP) formation and
 the activity of cAMP-dependent protein kinase
(protein kinase A, PKA), leading to
phosphorylation of DARPP-32. Phosphorylation
converts this phosphoprotein into a potent
inhibitor of protein phosphatase-1. Inhibition of
protein phosphatase-1 increases the
phosphorylation state of numerous
phosphoproteins involved in the regulation of
important physiological processes. Dopamine, via
D2 receptors, decreases DARPP-32
phosphorylation by two intracellular signaling
pathways. One mechanism involves inhibition of
adenylyl cyclase, a decrease in cAMP, a decrease
in the activity of PKA, and decreased
phosphorylation of DARPP-32. The other D2-
mediated effect involves an increase in
intracellular calcium, activation of calcineurin, and
increased dephosphorylation of phospho-DARPP-
32. Glutamate acting on N-methyl-D-aspartate
(NMDA) receptors also gives rise to a large influx
of calcium. Thus, stimulation of NMDA receptors
can also lead to activation of calcineurin and
enhanced dephosphorylation of DARPP-32,
producing an effect very similar to stimulation of
D2 receptors.

DYNAMICS OF DOPAMINE RECEPTORS

Destruction of the nigrostriatal dopamine systems has clear
and reproducible behavioral consequences. Unilateral lesions
of this system produce rotational behavior. Behavioral studies
in lesioned rats indicate that dopamine receptors in the
denervated striatum are supersensitive. Administration of
dopamine agonists (e.g., apomorphine) that selectively
stimulate dopamine receptors produces rotational behavior in
rats with unilateral lesions of the nigrostriatal dopamine
neurons. The degree of receptor sensitivity can be quantified
by measuring the amount of rotational behavior. The number
of dopamine receptors in the striatum ipsilateral to the lesion
increases markedly, and this increase appears to correlate
with the extent of the behavioral supersensitivity reflected by
the rotational behavior. Thus, an increase in dopamine
receptor density appears to be related to the behavioral
supersensitivity observed following unilateral destruction of
the nigrostriatal dopamine system.

Changes in the number of dopamine receptors are also

observed in the striatum following chronic administration of
dopamine antagonists. This led to the speculation that serious
side effects, such as tardive dyskinesia, following chronic
treatment with a neuroleptic drug might be due to
supersensitivity of dopamine receptors that have been
chronically blocked.

Dopamine receptors have been observed to change in disease

states. In schizophrenia, the density of the DAT and of the D1
dopamine receptor is normal. However, the D2 receptor
density is consistently elevated in postmortem studies of
brain regions such as caudate and putamen, even in tissue
obtained from neuroleptic-free individuals. Some preliminary
evidence indicating abnormal D2 structure as well as reduced
linkage between D1 and D2 receptors is available, warranting
a detailed study of the genes for these two receptors in
schizophrenia. Loss of midbrain dopamine in Parkinson's
disease is accompanied by a matching loss of the DAT and a
rise in density of both D1 and D2 receptors. These alterations
are found in the caudate nucleus and putamen tissues from
unmedicated patients. Long-term treatment with L-DOPA
appears to revert the receptor densities back toward normal
levels. D1 and D2 receptors are decreased in the striatum of
patients with Huntington's chorea, and there appears to be
reduced or absent linkage between them.
MOLECULAR BIOLOGY OF DOPAMINE
RECEPTORS
D1 and D2 receptors are distinct molecular entities, utilize
different transducing units (Table 9-3), and have a different
distribution in the brain (Table 9-2).

Developments in molecular biology, including cloning of the

cDNA and/or genes for several members of the large family of
G protein-coupled receptors, have revealed that
heterogeneity in the biochemical characteristics or
pharmacology of individual receptors often indicates the
presence of previously unsuspected molecular subtypes (see
Chapter 11). For dopamine systems, even though the D1/D2
receptor classification is widely accepted, biochemical,
pharmacological, and behavioral approaches have produced
data that are increasingly difficult to reconcile with the
existence of only two dopamine receptor subtypes and
suggested the presence of several novel subtypes of both
receptor types. Cloning studies have already identified four
subtypes of the D2 receptor and two subtypes of the D1
receptor (Table 9-4). Two forms of the D2 receptor, D2(short)
and D2(long), were identified by gene cloning and shown to
be derived from alternative splicing of a common gene. These
two subtypes appear to have an identical pharmacology. A
third subtype of the D2 receptor, the D3 receptor, was isolated
by screening rat brain cDNA and genomic libraries by reverse
transcription-polymerase chain reaction. This new D3 receptor
exhibits several novel characteristics. It has a different
anatomical distribution, with the highest levels found in limbic
brain structures, and its pharmacological profile, although
similar to the D2(short) and D2(long) forms, shows some
distinct differences; the D3 receptor exhibits about a 100-fold
increase in affinity for the dopamine agonist quinpirol.

The fourth subtype of the D2 receptor, cloned in 1991, the D4

receptor gene, has high homology to the human D2 and D3
receptor genes. The pharmacological profile of this receptor
resembles that of the D2 and D3 receptors, but its affinity for
the atypical antipsychotic drug clozapine is an order of
magnitude higher. The D4 RNA has an interesting regional
distribution in monkey brain, with high levels observed in the
frontal cortex, midbrain, amygdala, and medulla and lower
levels detected in the basal ganglia. The function of these D2
receptor subtypes is presently unknown. All known varieties
of the D2 receptor have seven membrane-spanning domains,
similar to the structure originally proposed for -adrenergic
receptors. Differences in ligand-binding and transduction
mechanisms are presumably related to variations in the
sequence of the receptor. The D1 receptor of humans and rats
has also been cloned, expressed, and characterized by several
laboratories; this work in conjunction with other studies is
consistent with the idea that other D1 receptor subtypes may
also exist. In fact, a gene encoding a 477-amino acid protein
has been cloned that has a striking homology to the cloned
D1 receptor. This D1 receptor subtype, called D5, has a
pharmacological profile similar to that of the cloned D1
receptor but displays a 10-fold higher affinity for the
endogenous agonist dopamine. Similar to the D1 receptor, the
D5 receptor stimulates adenylate cyclase activity. This
receptor is neuron-specific and located primarily in the limbic
areas of the brain but is absent from the parathyroid, kidney,
and adrenal gland.

DISTRIBUTION OF SUBTYPE-SPECIFIC
DOPAMINE RECEPTOR MRNA IN BRAIN
Advances in molecular biology have made it feasible to
determine which specific cells express a given gene, thus
allowing the anatomical determination not only of what
population of cells expresses a given gene but also of how
these genes may be regulated in normal and pathological
conditions. At least five genes encoding dopamine receptors
have been discovered. Currently, there are few
pharmacological or immunological tools for accurately
measuring the distribution of the receptor proteins of the new
dopamine receptors. Thus, our knowledge concerning the
tissue distribution of these receptors, especially in cases
where very specific ligands or selective antibodies are just
beginning to be developed, has come primarily from in situ
hybridization experiments. The tissue distribution and
characteristics of the mRNAs of the five different dopamine
receptors are illustrated in Tables 9-4 and 9-5. These
dopamine receptors appear to have overlapping as well as
some unique anatomical distributions and, in some cases,
distinct pharmacological profiles. In general, the distribution
patterns found in rodents parallel those observed in primates,
with several notable exceptions alluded to below. The D1 and
D2 receptor mRNAs are present in all dopaminoceptive
regions of the rat brain. In brain regions such as the
substantia nigra and ventral tegmental area, high levels of
D2, but not D1, mRNA are detected. The absence of D1 and
D5 receptor mRNA in the substantia nigra and ventral
tegmental area argues against these receptors playing a role
as autoreceptors. Receptor mRNAs for D3, D4, and D5 are
largely present in tissues where D1 and/or D2 receptor
mRNAs are also expressed. However, in most cases, the
relative abundance of mRNA for these receptors is one or
several orders of magnitude lower than that found for D1
and/or D2 receptors. While the primate substantia nigra
contains high levels of D2 receptor mRNA, the ventral
tegmental area in the primate brain does not contain readily
measurable levels of D2 or D3 receptor mRNA, suggesting
that the primate ventral tegmental area may not contain
appreciable numbers of dopamine autoreceptors.

Anatomically, D5 receptor mRNA has a rather discrete

distribution in rat brain: it is found only in the hypothalamus,
hippocampus, and parafascicular nucleus of the thalamus. In
the primate, this distribution extends to other temporal lobe
structures as well. The consensus from a number of studies is
that the only brain region that expresses all five dopamine
receptors is the hippocampus. Dopamine receptor mRNA has
also been found outside the CNS. D2 receptor mRNA is
abundant in the pituitary and adrenal glands and in retina.
Northern blot analyses have shown that neither D1 nor D3
receptor mRNA is detected outside the CNS, although in
kidney and heart D1- and D2-like activities have been
described. Since low levels of D5 receptor mRNA are
expressed in the kidney, this could account for the D1-like
activity. The D4 receptor mRNA found in rat heart may
account for the previously described D2-like activity. The level
of D4 receptor mRNA found in rat brain is about 20-fold lower
than the level in heart. In peripherally innervated tissue,
heart seems to be the exception since no D4 mRNA has been
found in adrenal, kidney, or liver.

PHARMACOLOGY OF DOPAMINE RECEPTOR

SUBTYPES
At present, no selective ligands have been developed that can
distinguish between D1 and D5 receptors. Pharmacologically,
the only characteristic that distinguishes D1 from D5
receptors is the increased affinity of the D5 receptor for
dopamine. Most neuroleptic drugs exhibit a higher affinity for
D2 receptors than for D3 or D4 receptors. The affinities of the
five dopamine receptor subtypes for selected clinically
relevant dopamine antagonists are summarized in Table 9-6.

Uniquely among the subtypes, D4 receptors respond to low

concentrations of norepinephrine and epinephrine as well as
to dopamine. The most interesting feature of the human D4
receptor is its apparent high affinity for clozapine (an atypical
neuroleptic) and its unique distribution in primate brain
(frontal cortex midbrain amygdala striatum), differing
markedly from D2 and D3 receptor mRNA. This interesting
pharmacology and unique distribution in brain has generated
a great deal of excitement, particularly from a clinical
standpoint. The possibility that clozapine exerts its
therapeutic effects via a D4 receptor mechanism was seen
immediately as offering a new and rational target for drug
development. Seeman and colleagues provided tantalizing
evidence that there may be an increase in the number of D4
receptors in schizophrenic patients, further fueling the
impetus to develop D4-selective antagonists as potential
antipsychotic drugs; but this provocative, albeit indirect,
study has not been replicated using more direct measures to
assess D4 receptor numbers in brains of normal and
schizophrenic subjects. Also, the significance of this finding,
even if replicated, would still be uncertain. The neuroleptics
taken by patients throughout the course of their disease could
modify dopamine receptor density, and the overabundance of
D4 receptors observed in the autopsied brains of
schizophrenic patients could be a result of drug treatment
rather than a cause of the disease. Future clinical research
efforts might profitably be directed to the use of in vivo
imaging techniques (i.e., SPECT and PET) to evaluate
dopamine receptor subtypes in schizophrenia when
appropriate D4- and D3-selective ligands become available.
However, the low density of these receptors, especially D4
receptors, may present an insurmountable obstacle.

The identification of novel dopamine receptor subtypes has

already had a dramatic impact on our understanding of
dopaminergic systems. Studies of the human D4 receptor
indicate that its DNA sequence is highly polymorphic at both
the DNA and amino acid levels, exhibiting a least 25 alleles. A
novel polymorphism of the D4 receptor was observed within
the putative third cytoplasmic loop of the protein, suggesting
that some polymorphic variants may display different
pharmacological properties. This high frequency of variation in
the coding region of a functional receptor protein is
unprecedented and could confer differences in efficacy of drug
treatment and/or predispose an individual to the development
of dopamine-dependent neuropsychiatric disorders. In fact,
the D4 receptor gene (DRD4) has been implicated in the
pathophysiology of several common neuropsychiatric
disorders, including mood disorders, attention-
deficit/hyperactivity disorder, Parkinson's disease, and specific
personality traits. The evidence is particularly strong for
attention-deficit/hyperactivity disorder.

The availability of receptor clones, receptor antibodies, and

expressed receptor proteins has permitted gene mapping as
well as in-depth studies of the circuitry of the dopaminergic
systems and the mechanisms regulating them at both the
genomic and cytoplasmic levels. It has also allowed the
physical structure of the receptors to be ascertained and
should permit the design and development of highly specific
ligands. It is hoped that these new selective agents will be
helpful not only in studying the function of dopamine systems
in normal and pathological states but also in the therapeutic
management of disorders associated with malfunction of
specific dopaminergic systems. Strides toward this goal had
already begun with the successful development of selective
D4 antagonists by several pharmaceutical companies when
the last edition of this text was completed. Some 5 years
later, these D4 antagonists have not yet provided the magic
bullet for the treatment of schizophrenia. However, these
agents have been useful in studying the localization and
function of D4 receptors in animals, including monkeys and
humans, and may someday find a therapeutic use in the
treatment of specific dopamine-dysregulated states.

PHARMACOLOGY OF DOPAMINERGIC
SYSTEMS

Nigrostriatal and Mesolimbic Dopamine

Systems
The nigrostriatal and mesolimbic dopamine neurons appear to
respond in a similar manner to drug administration (Table 9-
6). Acute administration of dopamine agonists (dopamine
receptor stimulators) decrease dopamine cell activity,
turnover, and catabolism. Acute administration of
antipsychotic drugs (dopamine receptor blockers) increases
dopaminergic cell activity, turnover, catabolism, and
biosynthesis. The increase in dopamine biosynthesis occurs at
the TH step and is in part a result of the ability of
antipsychotic drugs to block postsynaptic receptors and to
increase dopaminergic activity via a neuronal feedback
mechanism (Table 9-1). Also, some of the observed effects
are enhanced as a result of interaction with nerve terminal
autoreceptors. Blockade of nerve terminal autoreceptors
increases both the synthesis and the release of dopamine.
These systems respond to MAO inhibitors (MAOIs) with an
increase in dopamine and a decrease in dopamine synthesis,
as do the other dopamine systems discussed below.

Long-term treatment with antipsychotic drugs produces a

different spectrum of effects on central dopaminergic
neurons. For example, following long-term treatment with
haloperidol, nigrostriatal dopamine neurons in the rat become
quiescent and dopamine metabolite levels and dopamine
synthesis and turnover in the striatum return to normal limits.
The kinetic activation of striatal TH, which occurs following an
acute dose of an antipsychotic drug, also subsides following
long-term treatment. These results are usually interpreted as
indicative of the development of tolerance in the nigrostriatal
dopamine system. In contrast (see below), tolerance to the
biochemical effects observed following acute administration of
antipsychotic drugs does not appear to develop in the
mesoprefrontal and mesocingulate cortical dopamine
pathways after chronic administration.

Mesocortical Dopamine System

The response of the mesocortical dopamine systems to
dopaminergic drugs in most instances is qualitatively similar
to that of the nigrostriatal and mesolimbic systems (Table 9-
7), although some notable exceptions have been observed.
The mesotelencephalic dopamine neurons, which were once
believed to be three relatively simple and homogeneous
systems, have been found to be an anatomically,
biochemically, and electrophysiologically heterogeneous
population of cells with differing pharmacological
responsiveness. For example, although a great majority of
midbrain dopamine neurons appear to possess autoreceptors
on their cell bodies, dendrites, and nerve terminals, dopamine
cells that project to the prefrontal and cingulate cortices
appear either to have a greatly diminished number of these
receptors or to lack them entirely. The absence (or
insensitivity) of impulse-regulating somatodendritic as well as
synthesis-modulating nerve terminal autoreceptors on the
mesoprefrontal and mesocingulate cortical dopamine neurons
may, in part, explain some of the unique biochemical,
physiological, and pharmacological properties of these two
subpopulations of midbrain dopamine neurons (Table 9-8).
For example, the mesoprefrontal and mesocingulate
dopamine neurons appear to have a faster firing rate and a
more rapid turnover of transmitter than the nigrostriatal,
mesolimbic, and mesopiriform dopamine neurons. Transmitter
synthesis is also more readily influenced by altered availability
of precursor tyrosine in midbrain dopamine neurons lacking
autoreceptors (mesoprefrontal and mesocingulate) than in
those possessing autoreceptors. This may be related to the
enhanced rate of physiological activity in this subpopulation of
midbrain dopamine neurons, making them more responsive to
precursor regulation. Mesoprefrontal and mesocingulate
dopamine neurons also show diminished biochemical and
electrophysiological responsiveness to dopamine agonists and
antagonists. Low doses of apomorphine or autoreceptor-
selective dopamine agonists, in contrast to their inhibitory
effect on other midbrain dopamine neurons, are ineffective at
decreasing the activity or dopamine metabolite levels in these
two cortical dopamine projections. Dopamine receptor-
blocking drugs, such as haloperidol, produce large increases
in the synthesis and accumulation of dopamine metabolites in
nigrostriatal, mesolimbic, and mesopiriform dopamine
neurons but have only a modest effect in mesoprefrontal and
mesocingulate dopamine neurons.

Heterogeneity among midbrain dopamine neurons is also

found when one studies the effects of chronic antipsychotic
drug administration. When classic antipsychotic drugs are
administered repeatedly over time, the great majority of
dopamine cells cease to fire due to the development of a state
of depolarization inactivation. However, some midbrain
dopamine cells appear to be unaffected by repeated
antipsychotic drug administration. These dopamine cells are
the neurons projecting to the prefrontal and cingulate
cortices. Parallel observations have been made biochemically.
Following chronic administration of antipsychotic drugs,
tolerance develops to the metabolite-elevating effects of these
agents in the midbrain dopamine systems that possess
autoreceptors but not in the systems that lack autoreceptors.
When the atypical antipsychotic drug clozapine (which
possesses therapeutic efficacy but lacks Parkinson-like side
effects and an ability to produce tardive dyskinesia) is
administered repeatedly, dopamine neurons in the ventral
tegmental area develop depolarization inactivation but
neurons in the substantia nigra do not. The reason for this
differential effect is unknown. Foot shock, swim stress, and
conditioned fear cause selective (benzodiazepine-reversible)
metabolic activation of mesoprefrontal dopamine neurons
without causing a marked or consistent effect on other
midbrain dopamine neurons, including the mesocingulate
dopamine neurons. Thus, this selective activation does not
appear to be due solely to the absence of autoreceptors. The
anxiogenic benzodiazepine receptor ligands, such as the -
carbolines, also produce a selective dose-dependent
activation of mesoprefrontal dopamine neurons without
increasing dopamine metabolism in other midbrain dopamine
neurons.

In summary, certain mesotelencephalic dopamine systems,

namely, the mesoprefrontal and mesocingulate dopamine
neurons, possess many unique characteristics compared to
the nigrostriatal, mesolimbic, and mesopiriform dopamine
systems (Table 9-8). Many of these unique characteristics
may be the consequence of a lack of impulse-regulating
somatodendrite and synthesis-modulating nerve terminal
dopamine autoreceptors. However, some, such as the
response to stress and the anxiogenic -carbolines, are clearly
dependent on other regulatory influences and not solely
related to the absence of autoreceptors. These findings
suggest that dopamine action at autoreceptors may be one of
the more critical ways that dopamine cells modulate their
function. If valid, how do midbrain dopamine systems that
lack autoreceptors regulate themselves? Perhaps it is through
afferent systems by neuronal feedback. Some studies have
suggested that a substance P/substance K innervation of the
ventral tegmental area (A10) may influence the functional
activity of mesocortical and mesolimbic dopamine neurons.

A number of studies have demonstrated the importance of

NMDA receptors and of the glutamatergic input to the ventral
tegmental area in the regulatory control of mesoprefrontal
dopamine neurons. This input is believed to be at least
partially responsible for converting pacemaker-like firing in
dopamine cells into burst-firing patterns. NMDA receptors in
the ventral tegmental area appear to modulate differentially
the dopamine projections to the prefrontal cortex and nucleus
accumbens. The NMDA receptor is selectively activated by
NMDA and regulated at several pharmacologically distinct
sites, including a high-affinity, strychnine-insensitive glycine-
binding site (see Chapter 6). Competitive antagonists of this
strychnine-insensitive glycine site, which cross the blood-
brain barrier, have made possible the in vivo pharmacological
modulation of the NMDA receptor via this site. In behavioral
paradigms (restraint stress and conditioned fear) that cause
metabolic activation of mesoprefrontal and mesaccumbens
dopamine neurons, these agents (e.g., [+]-HA-966)
selectively abolish the activation of mesoprefrontal dopamine
neurons. The stress-induced activation of serotonin neurons
in the prefrontal cortex and the dopaminergic activation of the
nucleus accumbens are not altered by (+)-HA-966. Activation
of mesoprefrontal dopamine neurons elicited by acute
administration of phencyclidine and/or -9-
tetrahydrocannabinol, the active ingredient in marijuana, is
also attenuated by HA-966. These data indicate that under
certain perturbed states the NMDA receptor complex and the
associated glycine-modulatory site play an important role in
the afferent control of the dopamine neurons in the prefrontal
cortex and provide a potential target for pharmacological
regulation of this important dopamine projection.

The observation that central dopamine systems are quite

heterogeneous from both a biochemical and a functional point
of view holds promise that it will soon be possible to develop
drugs targeted to modify or restore function to selective
dopamine systems that are abnormal in various behavioral or
pathological states. Some progress has already been achieved
in developing agents that appear to act at selective dopamine
receptor sites (Fig. 9-6). Whether these agents will be useful
in selectively modifying the function of subsets of midbrain
dopamine neurons remains to be determined.

Tuberoinfundibular and Tuberohypophysial

Dopamine Systems
The tuberoinfundibular dopamine system responds to
pharmacological and endocrinological manipulations in a
manner that is qualitatively different from the other three
dopamine systems (nigrostriatal, mesolimbic, and
mesocortical) described above (Table 9-7). Tuberoinfundibular
neurons appear to be regulated in part by circulating levels of
prolactin. Prolactin increases the activity of these neurons by
acting within the medial basal hypothalamus, possibly directly
on the tuberoinfundibular neurons. These neurons in turn
release dopamine, which then inhibits prolactin release from
the anterior pituitary. Haloperidol and other antipsychotic
drugs have no effect on dopamine turnover in the
tuberoinfundibular dopamine system until about 16 hours
after drug administration, whereas the biochemical effects in
other systems are observed within minutes and are maximal
in several hours. The absence of an acute response to
dopamine antagonists and agonists may be related to the lack
of autoreceptors in this system. While less is known about the
pharmacology of the tuberohypophysial dopamine neurons,
this system seems to respond to drugs in a manner
qualitatively similar to the better-studied dopamine systems
(Table 9-7).

SPECIFIC DRUG CLASSES

Antipsychotic Drugs
For many years, it has been known that antipsychotic drugs of
both the phenothiazine and butyrophenone classes can
increase the turnover of dopamine in the CNS. Since these
drugs appear to have potent dopamine receptor-blocking
capabilities, it has been suggested that the increased
dopamine turnover results from blockade of both dopamine
autoreceptors and postsynaptic dopamine receptors and a
consequent feedback activation of the dopaminergic neurons,
presumably by some sort of neuronal feedback loop. This
speculation has been verified in part by direct extracellular
recording techniques.

The antagonism of central dopamine receptors by

antipsychotic drugs has been postulated as a critical
determinant of the therapeutic efficacy of this class of drugs.
It is clear, however, that drugs with clinical antipsychotic
effects influence dopamine transmission at several levels,
including transmitter synthesis, release, and metabolism.
Indeed, the "dopamine hypothesis" of schizophrenia was
originally based on studies of alterations in brain dopamine
metabolism produced by haloperidol and chlorpromazine in
mice. The dopamine hypothesis later gained support from the
finding that classical antipsychotic drugs such as haloperidol
and chlorpromazine occupy dopamine D2-like receptors in the
brain.

After the superior efficacy of clozapine was demonstrated, a

new generation of "atypical" antipsychotic drugs followed. The
greater therapeutic efficacy of these drugs does not appear to
be related to actions at the D2 receptor alone (see Chapter
13). Moreover, the atypical antipsychotic drugs profoundly
affect cortical dopamine metabolism and release after acute
and chronic administration. Indeed, a variety of novel drugs
have already been characterized as atypical antipsychotic
drugs (clozapine, olanzapine, risperidone, amperozide,
ziprasidone), and it has become customary to attribute their
beneficial actions to a preferential increase in dopamine
release in the frontal cortex relative to the striatum. Each of
these drugs profoundly increases cortical, versus subcortical,
dopamine release in a pattern unlike that found for typical
antipsychotic drugs. Moreover, this increase in cortical
dopamine efflux is directly related to increased impulse flow
of mesocortical dopamine neurons. The argument that this
cortical dopamine effect may be related to the superior
efficacy of this class of drugs arises, in part, from the notion
that the schizophrenic disease process may include a
component of frontal cortical dopaminergic hypofunction,
which these agents might reverse. The pharmacological
mechanisms by which atypical antipsychotic drugs increase
cortical dopamine transmission are currently unknown.

Compared to typical antipsychotic drugs, chronic

administration of atypical antipsychotic drugs appears to
affect dopaminergic systems in a distinct way. For example,
basal cortical, but not subcortical, dopamine efflux is
increased after chronic clozapine administration. Thus, after
chronic clozapine administration, there may be a functional
disconnection between impulse flow and release, possibly due
to an emphasis on terminal level-dependent regulation of
transmitter release. In any case, it appears that atypical
antipsychotic drugs facilitate cortical dopamine transmission
after acute or chronic exposure, and this relative activation of
cortical versus subcortical dopamine release may underlie the
strong therapeutic effects in the absence of observable
extrapyramidal side effects.

Stimulants
The therapeutic use of this class of drugs is becoming less
and less common as awareness of their abuse potential
increases. At present, their use is largely restricted to the
treatment of narcolepsy and of hyperkinetic children and as
general anorectic agents. The principal drugs in this category
are the various analogues and isomers of amphetamine and
methylphenidate.

For many years it has been known that ingestion of large

amounts of amphetamine often leads to a state of paranoid
psychosis that may be hard to distinguish from the paranoia
associated with schizophrenia. It now appears that this
paranoid state can be readily and reproducibly induced in
humans given large amounts of amphetamine, so the drug
may provide a convenient "model psychosis" for experimental
study. It is of interest in this regard that antipsychotic drugs
such as chlorpromazine can readily reverse amphetamine-
induced psychosis.

On the biochemical level, it was no surprise to learn that

amphetamine and related compounds interact with
catecholamine-containing neurons since amphetamine is a
close structural analogue of the catecholamines. However,
there was no clear evidence that amphetamine produced its
CNS effects through a catecholamine mechanism until it was
demonstrated that -methyl-tyrosine (a potent inhibitor of
TH) prevented most of the behavioral effects of D-
amphetamine. The question as to which catecholamine,
norepinephrine or dopamine, is involved in the behavioral
effects of amphetamine remains unanswered, but it is
generally believed that the so-called stereotypic behaviors in
animals (i.e., compulsive gnawing, sniffing) induced by
amphetamine are associated with a dopaminergic mechanism
and that the increase in locomotor activity involves a
noradrenergic mechanism or both.

Many classes of psychotropic drugs interact in one way or

another with catecholamine-containing neurons. Figure 9-9
outlines the life cycle of the transmitters of dopaminergic and
noradrenergic neurons in the CNS and indicates possible sites
at which drugs may intervene in this cycle. This schematic
model also provides examples of drugs or chemical agents
that interfere at the various sites within the life cycle of the
transmitter substances. These numerous sorts of interaction
ultimately result in an increase, decrease, or no change in the
functional activity of the catecholamine neuron in question.

Only recently did it become clear exactly how various

pharmacological agents alter activity in defined catecholamine
neuronal systems in the brain. In most cases, the turnover of
monoamines depends essentially on impulse flow in the
neuron. An increase in impulse flow usually causes an
increase in turnover, and a decrease in impulse flow causes a
reduction in turnover. As mentioned above, however, this is
not always the case in the dopamine system, if synthesis is
used as an index of turnover. Turnover measurement has
been used to gain some insight into the activity of various
types of monoamine-containing neuron during different
behavioral states or after administration of different
psychotropic drugs. As might be predicted, psychotropic drugs
can have a variety of effects, and these effects can alter the
turnover of a given transmitter substance without necessarily
altering impulse flow in the neuronal system under study. For
example, a drug can have a direct effect on the synthesis,
degradation, uptake, or release of a given transmitter that will
then alter the turnover of the transmitter in question but will
not necessarily lead to an increase or decrease in the activity
of the neuronal system that utilizes that substance as a
transmitter. Thus, an alternation in turnover of a transmitter
is not necessarily a clear indication that there has been a
change in impulse flow in a given neuronal pathway.
Therefore, the most direct way to determine if a drug alters
impulse flow in a chemically defined neuronal system is to
measure the activity of that system while the animal is under
the influence of the drug.
Drugs can alter impulse flow in several ways. For example, a
drug can act directly on the nerve cell body; it can act on
other neurons, which then influence impulse flow in the
neuron under study; or it can act at the postsynaptic receptor
to cause stimulation or blockade, which then results in some
sort of feedback influence on the presynaptic neuron. This
feedback information could be either neuronal or perhaps
transsynaptically mediated by release of some local chemical
from the postsynaptic membrane. The combined
histochemical-neurophysiological identification of
dopaminergic and noradrenergic neurons has made possible
the direct study of the effects of various drugs on the firing of
these chemically defined neurons.

Figure 9-9. Schematic model of a central

dopaminergic neuron indicating possible sites of
drug action.

Site 1: Enzymatic synthesis: Tyrosine hydroxylase

reaction blocked by the competitive inhibitor -
methyltyrosine and other tyrosine hydroxylase
inhibitors.

Site 2: Storage: Reserpine and tetrabenazine

interfere with the uptake-storage mechanism of
the amine granules. The depletion of dopamine
(DA) produced by reserpine is long-lasting, and
the storage granules appear to be irreversibly
damaged. Tetrabenazine also interferes with the
uptake-storage mechanism of the granules, except
that the effects of this drug do not appear to be
irreversible.

Site 3: Release: -Hydroxybutyrate and HA966

effectively block the release of DA by blocking
impulse flow in dopaminergic neurons.

Amphetamine administered in high doses releases

 Amphetamine administered in high doses releases
DA, but most of the releasing ability of
amphetamine appears to be related to its ability to
effectively block DA reuptake.

Site 4: Receptor interaction: Apomorphine is an

effective DA receptor-stimulating drug, with both
pre- and postsynaptic sites of action. Both 3-PPP
and EMD-23-448 (an indolebutylamine) are
autoreceptor agonists. Perphenazine and
haloperidol are effective DA receptor-blocking
drugs.

Site 5: Reuptake: DA has its action terminated by

being taken up into the presynaptic terminal.
Amphetamine as well as benztropine, the
anticholinergic drug, are potent inhibitors of this
reuptake mechanism.

Site 6: Monoamine oxidase (MAO): DA present in

a free state within the presynaptic terminal can be
degraded by the enzyme MAO, which appears to
be located in the outer membrane of the
mitochondria. Dihydroxyphenylacetic acid
(DOPAC) is a product of the action of MAO and
aldehyde oxidase on DA. Pargyline is an effective
inhibitor of MAO. Some MAO is also present
outside the dopaminergic neuron.

Site 7: Catechol-O-methyltransferase (COMT): DA

can be inactivated by the enzyme COMT, which is
believed to be localized outside the presynaptic
neuron. Tropolone is an inhibitor of COMT.

IMAGING DOPAMINE TRANSMISSION AND

INTEGRITY IN NEUROLOGICAL AND
PSYCHIATRIC DISORDERS

Introduction
Advances in neuroimaging techniques have made it possible
to visualize dopamine transmission in neuropsychiatric
disorders. Radiotracer imaging with PET and SPECT can be
used to measure pre-, post-, and intrasynaptic aspects of
dopaminergic transmission. Presynaptic sites can be labeled
with radiotracers for the DAT, VMAT or the synthetic enzyme
aromatic L-amino acid decarboxylase. Postsynaptic sites can
be labeled with radiotracers for the D1 or D2 receptor.
Estimates of synaptic endogenous dopamine release can be
made indirectly by measuring the displacement of receptor
tracers by endogenous dopamine. Pharmacological agents
that either release (i.e., amphetamine) or deplete ( -methyl-
p-tyrosine) dopamine tissue stores are used to assess
alterations in synaptic dopamine in normal and disease states.

Parkinson's Disease
Parkinson's disease is a progressive neurodegenerative
disorder of the basal ganglia that is characterized by tremor,
muscular rigidity, difficulty in initiating motor activity, and loss
of postural reflexes. It is observed in approximately 1% of the
population over age 55. It has been known for over 70 years
that Parkinson's disease is characterized pathologically by loss
of pigmented cells in the substantia nigra, but only since 1960
has it been appreciated that there is a substantial loss of
dopamine in the striatum. It is now clear that Parkinson's
disease can be defined in biochemical terms as primarily a
dopamine-deficiency state resulting from degeneration or
injury to dopamine neurons. The most striking degenerative
loss of dopamine neurons is observed in the nigrostriatal
system. Even in patients with mild symptoms, a striatal
dopamine loss of 70%-80% is observed, while severely
impaired subjects have striatal dopamine depletions in excess
of 90%. Since the DAT is heavily expressed in the terminals of
dopamine neurons that are lost in Parkinson's disease, it is
not surprising that striatal binding of agents that label this
site (cocaine, nomifensine, GBR, and mazindol) is lost in the
parkinsonian striatum. This alteration corresponds well with
the loss of functional dopamine uptake visualized in vivo by
PET using (18F)-L-DOPA uptake or nomifensine. Although
striatal dopamine loss represents the primary neurochemical
abnormality in the Parkinson's disease brain, typical
parkinsonism is accompanied by loss of other dopamine
systems and other monoamine neurons as well. Some
degeneration of dopamine-containing neurons is also
apparent in the mesolimbic, mesocortical, and hypothalamic
systems; the norepinephrine-containing neurons in the locus
ceruleus; and the serotonin neurons in the raphe nucleus.
Nonmonoamine systems are also affected, with depletions
observed in somatostatin, neurotensin, substance P,
enkephalin, and cholecystokinin-8. Since many of these
nondopamine systems indirectly interact with
mesotelencephalic dopamine systems, changes in some of
them are bound to influence in a complex way the function of
dopamine neurons.

The strategy for treating Parkinson's disease has been to

restore the dopamine deficit in the brain by pharmacological
means or, more recently, by neural grafting of dopamine-
containing cells. There are a number of theoretical strategies
for drug therapy in Parkinson's disease, including substrate
supplementation with direct and indirect dopamine agonists,
metabolic inhibitors (MAOIs), and uptake inhibitors. The most
successful treatment has been the use of L-DOPA. Direct
dopamine agonists also have some benefit in patients whose
responsiveness to L-DOPA is greatly reduced or erratic. So far,
the only direct-acting dopamine agonist that has found
extensive use is bromocriptine, primarily a D2 agonist. Other
agents belonging to this class will no doubt prove useful in the
future as supplements or alternatives to L-DOPA.

In view of the behavioral and electrophysiological studies that

suggest that D1 receptor activation is necessary for the
effects of D2 receptor stimulation to be maximally expressed
in normal animals as well as in animals with supersensitive
dopamine receptors, the functional interaction between D1
and D2 receptors could have important implications in
Parkinson's disease, where stimulation of postsynaptic
dopamine receptors confers symptomatic benefit. Knowledge
of the optimal ratio of relative drug activity at D1 and D2
receptors that is required to elicit effective stimulation of
dopamine-mediated function may provide a basis for the
design of new drugs. Also, as more knowledge accumulates
concerning the distribution and function of various dopamine
receptor subtypes, this should facilitate the development of
new agents to treat dopamine-deficiency states.

The MPTP-treated parkinsonian primate has provided a very

useful animal model in which to examine therapeutic
strategies for treatment (see Chapter 13). In fact, this model
has already been successfully exploited to design, refine, and
evaluate neural transplantation and gene therapy techniques
and to test new pharmacological strategies for the therapeutic
management of Parkinson's disease.

Schizophrenia
The etiology of schizophrenia has also been linked to defective
dopamine neurotransmission. The growing conviction that
antipsychotic agents act therapeutically by decreasing central
dopaminergic transmission led to the formulation of the
dopamine theory of schizophrenia. This theory has been
revised in recent years to take into account the heterogeneity
of midbrain dopamine systems, their differential response to
antipsychotic drugs, and data from neuroimaging of
schizophrenic subjects and primate models of this disorder. In
its simplest revised form, this hypothesis states that
schizophrenia may be related to a relative excess of
mesolimbic and a deficit of mesoprefrontal dopaminergic
activity (see Chapter 13).

SELECTED REFERENCES
Bannon, M. J., J. G. Granneman, and G. Kapatos (1995). The
dopamine transporter: potential involvement in
neuropsychiatric disorders. In Psychopharmacology: The
Fourth Generation of Progress (F. E. Bloom and D. J. Kupfer,
eds.). Raven Press, New York, pp. 179-188.

Bannon, M. J. and R. H. Roth (1983). Pharmacology of

mesocortical dopamine neurons. Pharmacol. Rev. 35, 53.

Bjorklund, A. and O. Lindvall (1984). Dopamine-containing

systems in the CNS. In Handbook of Chemical Neuroanatomy.
Classical Transmitters in the CNS, vol. 2, part I (A. Bjorklund
and T. Hokfelt, eds.). Elsevier, Amsterdam, pp. 55-122.

Chiodo, L. A. and A. S. Freeman (eds.) (1987).

Neurophysiology of Dopaminergic Systems: Current Status
and Clinical Perspectives. Lakeshore Publishing, Gross Pointe,
MI.

Chiodo, L. A., A. S. Freeman, and B. S. Bunney (1995).

Dopamine autoreceptor signal transduction and regulation. In
Psychopharmacology: The Fourth Generation of Progress (F.
E. Bloom and D. J. Kupfer, eds.). Raven Press, New York, pp.
221-226.

Civelli, O. (1995). Molecular biology of the dopamine receptor

subtypes. In Psychopharmacology: The Fourth Generation of
Progress (F. E. Bloom and D. J. Kupfer, eds.). Raven Press,
New York, pp. 155-162.
Dahlstrom, A. and K. Fuxe (1964). Evidence for the existence
of monoamine-containing neurons in the central nervous
system. I: Demonstration of monoamines in the cell bodies of
brain stem neurones. Acta Physiol. Scand., 62. Suppl. 232.1-
55.

Elsworth, J. D., M. Al-Tikriti, J. R. Sladek, Jr., J. R. Taylor, R. B.

Innis, D. E. Redmond, Jr., and R. H. Roth (1994). Novel
radioligands for the dopamine transporter demonstrate the
presence of nigral grafts in caudate nucleus of MPTP treated
monkey with improved behavioral function. Exp. Neurol. 126,
300-304.

Elsworth, J. D. and R. H. Roth (1995). Dopamine autoreceptor

pharmacology and function: recent insights. In The Dopamine
Receptors (K. A. Neve, ed.). Human Press, Totowa, NJ, pp.
223-265.

Fuxe, K. and T. Hokfelt (1971). Histochemical fluorescence

detection of changes in central monoamine neurones
provoked by drugs acting on the CNS. Triangle, 10(3): 73-84.

Gainetdinov, R. R., S. R. Jones, and M. G. Caron (1999).

Functional hyperdopaminergia in dopamine transporter knock-
out mice. Biol. Psychiatry 46, 303-311.

Gardner, E. L. and C. R. Ashby, Jr. (2000). Heterogeneity of

the mesotelencephalic dopamine fibers: physiology and
pharmacology. Neurosci. Biobehav. Rev. 24, 115-118.

Gingrich, J. A. and M. G. Caron (1993). Recent advances in

the molecular biology of dopamine receptors. Annu. Rev.
Neurosci. 16, 299-321.

Giros, B. and M. G. Caron (1993). Molecular characterization

of the dopamine transporter. Trends Pharmacol. Sci. 14, 43-
49.

Grace, A. (2002). Dopamine. In Neuropsychopharmacology

The Fifth Generation of Progress (K. L. Davis, D. Charney, J. T.
Cole and C. Nemeroff, eds.). Lippincott Williams Wilkins,
Philadelphia, PA, pp. 119-132.

Greengard, P., P. A. Allen, and A. C. Nairn (1999). Beyond the

dopamine receptor: the DARPP-32/protein phosphatase-1
cascade. Neuron 23, 435-447.

Grenhoff, J., S. W. Johnson (1995). Electrophysiological

Effects of Dopamine Receptor Stimulation. In The Dopamine
Receptors (K. A. Neve, ed.) Human Press, Totowa, N.J., pp.
267-304.

Jentsch, J. D. and R. H. Roth (2000). Effects of antipsychotic

drugs on dopamine release and metabolism in the central
nervous system. In Neurotransmitter Receptors in Actions of
Antipsychotic Medications, vol. 3 (M. Lidow, ed.). CRC Press,
Boca Raton, FL, pp. 31-41.

Jentsch, J. D., J. R., Taylor, and R. H. Roth (2000).

Phencyclidine model of frontal cortical dysfunction in non-
human primates. Neuroscientist 6, 268-275.

Kitai, S. T., P. D. Shepard, J. C. Callaway, and R. Scroggs

(1999). Afferent modulation of dopamine neuron firing
patterns. Curr. Opin. Neurobiol. 9, 690-697.

Kim, D. S., M. S., Szczypka, and R. D. Palmiter (2000).

Dopamine-deficient mice are hypersensitive to dopamine
receptor argonists. J. Neurosci. 20, 4405-4413.

McCracken, J. T., S. L. Smalley, J. J. McGough, L. Crawford, M.

Del'Homme, R. M. Cantor, A. Liu, and S. F. Nelson (2000).
Evidence for linkage of a tandem duplication polymorphism
upstream of the dopamine D4 receptor gene (DRD4) with
attention deficit hyperactivity disorder (ADHD). Mol.
Psychiatry 5, 531-536.

Meador-Woodruff, J. H. (1994). Update on dopamine

receptors. Ann. Clin. Psychiatry 6, 79-90.

Miller, G. W., R. R. Gainetdinov, A. I. Levey, and M. C. Caron

(1999). Dopamine transporters and neuronal injury. Trends
Pharmacol Sci. 20, 424-429.

Missale, C., S. R., Nash, S. W., Robinson, M. Jaber, and M. G.

Caron (1998). Dopamine receptors: from structure to
function. Physiol. Rev. 78, 189-225.

Moore, K. E. and K. J. Lookingland (1995). Dopaminergic

neuronal systems in the hypothalamus. In
Psychopharmacology: The Fourth Generation of Progress (F.
E. Bloom and D. J. Kupfer, eds.). Raven Press, New York, pp.
245-256.

O'Malley, K. L., S. Haron, L. Tang, and R. D. Todd (1992). The

rat dopamine D4 receptor: sequence, gene structure and
demonstration of expression in the cardiovascular system.
New Biol. 4, 137-146.

Pani, L., A. Porcella, and G. L. Gessa (2000). The role of

stress in the pathophysiology of the dopaminergic system.
Mol. Psychiatry 5, 14-21.

Roth, R. H. and J. D. Elsworth (1995). Biochemical

pharmacology of midbrain dopamine neurons. In
Psychopharmacology: The Fourth Generation of Progress (F.
E. Bloom and D. J. Kupfer, eds.). Raven Press, New York, pp.
277-244.

Schwartz, J.-C., D., Levesque, M.-P. Martres, and P. Sokoloff

(1993). Dopamine D3 receptor: basic and clinical aspects.
Clin. Neuropharmacol. 16, 295-314.

Seeman, P. (1995). Dopamine receptors: clinical correlates.

In Psychopharmacology: The Fourth Generation of Progress
(F. E. Bloom and D. J. Kupfer, eds.). Raven Press, New York,
pp. 295-302.

Seeman, P. and H. H. M. Van Tol (1994). Dopamine receptor

pharmacology. Trends Pharmacol. Sci. 15, 264-270.

Sharma, A., M. L. Kramer, P. F., Wick, D. Liu, S. Chari, S.

Shim, W. Tan, D. Ouellette, M. Nagata, C. J. DuRand, M. Kotb,
and R. C. Deth (1999). D4 dopamine receptor-mediated
phospholipid methylation and its implications for mental
illnesses such as schizophrenia. Mol. Psychiatry 4, 235-246.

Sibley, D. R. and F. J. Monsma, Jr. (1992). Molecular biology

of dopamine receptors. Trends Pharmacol. Sci. 13, 61-69.

Sunahara, R. K., H. C. Guan, B. F. O'Dowd, P. Seeman, L. G.

Laurier, G. Ng, S. R. George, J. Torchia, H. H. M. Van Tol, and
H. B. Niznik (1991). Cloning of the gene for a human
dopamine D5 receptor with higher affinity for dopamine than
D1. Nature 350, 614-619.

Tang, L., R. D. Todd, A. Heller, and K. L. O'Malley (1994).

Pharmacological and functional characterization of D2, D3 and
D4 dopamine receptors in fibroblast and dopaminergic cell
lines. J. Pharmacol. Exp. Ther. 268, 495-502.

Tarazi, F. I. and R. J. Baldessarini (1999). Dopamine D4

receptors: significance for moloecular psychiatry at the
millennium. Mol. Psychiatry 4, 529-538.

Van Tol, H. H. M., J. R. Bunsow, H. C. Guan, R. K. Sunahara,

P. Seeman, H. B. Niznik, and O. Civelli (1991). Cloning of the
gene for a human dopamine D4 receptor with high affinity for
the antipsychotic clozapine. Nature 350, 610-614.

Verhoeff, N. P. L. G. (2001). Imaging of dopaminergic

transmission in neuropsychiatric disorders. Curr. Opin.
Psychiatry 3, 227-239.

Wolf, M. E., A. Y. Deutch, and R. H. Roth (1987).

Pharmacology of central dopaminergic neurons. In Handbook
of Schizophrenia. Neurochemistry and Neuropharmacology of
Schizophrenia, vol. 2 (F. A. Henn and L. E. DeLisi, eds.).
Elsevier, Amsterdam, pp. 101-147.

Wolf, M. E. and R. H. Roth (1990). Autoreceptor regulation of

dopamine synthesis. Ann. N.Y. Acad. Sci. 604, 323-343.

Xu, M., X.-T. Hu, D. C. Cooper, R. Moratalla, A. M. Graybiel, F.

J. White, and S. Tonegawa (1994). Elimination of cocaine-
induced hyperactivity and dopamine-mediated
neurophysiological effects in dopamine D1 receptor mutant
mice. Cell 79, 945-955.

(+/-) Show / Hide Bibliography

Send Feedback Teton Server (5.0.0) - ©2007 Teton Data Title Updates
Customer Service Systems User
800.901.5494 Send Us Your Comments Responsibilities
Training Center
What's New
10. Serotonin (5-
Hydroxytryptamine), Histamine, and
Adenosine
SEROTONIN
Of all the neurotransmitters discussed in this book, serotonin
remains historically the most intimately involved with
neuropsychopharmacology. From the mid-nineteenth century,
scientists had been aware that a substance found in serum
caused powerful contraction of smooth muscle organs, but
over 100 years passed before scientists at the Cleveland Clinic
succeeded in isolating this substance as a possible cause of
high blood pressure.

At the same time, investigators in Italy were characterizing a

substance found in high concentrations in chromaffin cells of
the intestinal mucosa. This material also constricted smooth
muscular elements, particularly those of the gut. The material
isolated from the bloodstream was given the name serotonin,
while that from the intestinal tract was called enteramine.
Subsequently, both materials were purified, crystallized, and
shown to be 5-hydroxytryptamine (5-HT), which could then
be prepared synthetically and shown to possess all of the
biological features of the natural substance. The indole nature
of this molecule bore many resemblances to the psychedelic
drug lysergic acid diethylamide (LSD), with which it could be
shown to interact on smooth muscle preparations in vitro. 5-
HT is also structurally related to other psychotropic agents
(Fig. 10-1).

When 5-HT was first found within the mammalian central

nervous system (CNS), the theory arose that various forms of
mental illness could be due to biochemical abnormalities in its
synthesis. This line of thought was extended when the
tranquilizing substance reserpine was observed to deplete
brain 5-HT; throughout the duration of the depletion,
profound behavioral depression occurred. As we shall see,
many of these ideas are still maintained, although we now
have much more evidence with which to evaluate them.

Figure 10-1. Structural relationships of the various indolealky

Compound
Tryptamine R
Serotonin Trypt
Melatonin 5-Met
DMT* R
DET* R
Bufotenine* 5-Hyd
Szara psychotrope* 6-Hyd
Psilocin* 4-Hyd
Harmaline* 6-Met
5-MT 5-Met
5,6-DHT 5,6-D
5,7-DHT 5,7-D
*Psychotropic or behavioral effects

BIOSYNTHESIS AND METABOLISM OF

SEROTONIN

Introduction
Serotonin is found in many cells that are not neurons, such as
platelets, mast cells, and the enterochromaffin cells
mentioned above. In fact, only about 1%-2% of the serotonin
in the whole body is found in the brain. Nevertheless, because
5-HT cannot cross the blood-brain barrier, brain cells must
synthesize their own.
For brain cells, the first important step is uptake of the amino
acid tryptophan, which is the primary substrate for synthesis.
Plasma tryptophan arises primarily from the diet, and unlike
the catecholamine precursor tyrosine, elimination of dietary
tryptophan can profoundly lower the levels of brain serotonin.
In addition, an active uptake process is known to facilitate the
entry of tryptophan into the brain, and this carrier process is
open to competition from large neutral amino acids, including
the aromatic amino acids (tyrosine and phenylalanine), the
branched-chain amino acids (leucine, isoleucine, and valine),
and others (e.g., methionine and histidine). The competitive
nature of the large neutral amino acid carrier means that
brain levels of tryptophan will be determined not only by the
plasma concentration of tryptophan but also by the plasma
concentration of competing neutral amino acids. Thus, dietary
protein and carbohydrate content can specifically influence
brain tryptophan and serotonin levels by effects on plasma
amino acid patterns. Because plasma tryptophan has a daily
rhythmic variation in its concentration, it seems likely that
this concentration variation could also profoundly influence
the rate and synthesis of brain serotonin.

The first step in the synthetic pathway is hydroxylation of

tryptophan at the 5 position (Fig. 10-2) to form 5-
hydroxytryptophan (5-HTP). The enzyme responsible for this
reaction, tryptophan hydroxylase, occurs in low
concentrations in most tissues, including the brain; and it was
very difficult to isolate for study. After purifying the enzyme
from mast cell tumors and determining the characteristic
cofactors, however, it became possible to characterize this
enzyme in the brain. (Students should investigate the
ingenious methods used for the initial assays of this
extremely minute enzyme activity.) As isolated from brain,
the enzyme appears to have an absolute requirement for
molecular oxygen, for reduced pteridine cofactor, and for a
sulfhydryl-stabilizing substance, such as mercaptoethanol, to
preserve activity in vitro. With this fortified assay system,
there is sufficient activity in the brain to synthesize 1 mg of 5-
HTP per gram of brain stem in 1 hour. The pH optimum is
approximately 7.2, and the Km for tryptophan is 3 10-4 M.
Additional research into the nature of the endogenous
cofactor tetrahydrobiopterin yielded a Km for tryptophan of 5
10-5 M, which is still above normal tryptophan
concentrations. Thus, the normal plasma tryptophan content
and the resultant uptake into brain leave the enzyme
normally unsaturated with available substrate. Tryptophan
hydroxylase appears to be a soluble cytoplasmic enzyme, but
the procedures used to extract it from the tissues may greatly
alter the natural particle-binding capacity. Investigators
examining the relative distribution of particulate and soluble
tryptophan hydroxylase have reported that the particulate
enzyme may be associated with 5-HT-containing synapses,
while the soluble form is more likely to be associated with the
perikaryal cytoplasm. The particulate form of the enzyme
shows the lower Km and bears an absolute requirement for
tetrahydrobiopterin.

Purified tryptophan hydroxylase has a molecular weight of

52,000-60,000. It is a multimer of identical subunits that can
be activated by phosphorylation, Ca2+ phospholipids, and
partial proteolysis.

Cloning and sequencing of cDNAs for tryptophan hydroxylase

have been accomplished. Comparison of the rabbit tryptophan
hydroxylase sequence with the sequences of phenylalanine
hydroxylase and tyrosine hydroxylase demonstrates that
these three pterin-dependent aromatic amino acid
hydroxylases are highly homologous, reflecting a common
evolutionary origin from a single primordial genetic locus. The
pattern of sequence homology supports the hypothesis that
the C-terminal two-thirds of the molecules constitute the
enzymatic activity cores and the N-terminal one-third
constitutes domains for substrate specificity.

The tryptophan hydroxylase step in the synthesis of 5-HT can

be specifically blocked by p-chlorophenylalanine, which
competes directly with the tryptophan and binds irreversibly
to the enzyme. Therefore, recovery from tryptophan
hydroxylase inhibition with p-chlorophenylalanine appears to
require the synthesis of new enzyme molecules. In the rat, a
single intraperitoneal injection of 300 mg/kg of this inhibitor
lowers the brain serotonin content to less than 20% within 3
days, and complete recovery does not occur for almost 2
weeks.

Considerable attention has been directed to the overall

regulation of this first enzymatic step of serotonin synthesis,
especially in animals and humans treated with psychoactive
drugs alleged to affect the serotonin systems as a primary
mode of action. These studies have made an important
general point that seems to apply to the brain's response to
drug exposure in many other cell systems as well as to
serotonin: because transmitter synthesis, storage, release,
and response are dynamic processes, the acute imbalances
produced initially by drug treatments are soon counteracted
by the built-in feedback nature of synthesis regulation. Thus,
if a drug reduces tryptophan hydroxylase activity, the nerve
cells may respond by increasing their synthesis of the enzyme
and transporting increased amounts to the nerve terminals.

Mandell and colleagues have provided evidence, for example,

that short-term treatment with lithium will initially increase
tryptophan uptake, resulting in increased amounts being
converted to 5-HT. After 14-21 days of chronic treatment,
however, repetition of the measurements shows that while
tryptophan uptake is still increased, the activity of the
enzyme is decreased so that normal amounts of 5-HT are
being made. In this new equilibrium state, the neurochemical
actions of drugs like amphetamine and cocaine on 5-HT
synthesis rates are greatly reduced, as are their behavioral
actions. In this way, the 5-HT system, by shifting the
relationship between uptake and synthesis during Li
exposure, can be viewed as more stable. This factor may be
more fully appreciated when one considers that in the
treatment of manic-depressive psychosis a minimum of 7-10
days is usually required before the therapeutic action of Li
begins, a period during which the reequilibration of the 5-HT-
synthesizing process might undergo restabilization.

Figure 10-2. The metabolic pathways available

for the synthesis and metabolism of serotonin.

Decarboxylation
Once synthesized from tryptophan, 5-HTP is almost
immediately decarboxylated to yield serotonin. The enzyme
responsible for this conversion is identical with the enzyme
that decarboxylates dihydroxyphenylalanine (DOPA, i.e.,
aromatic amino acid decarboxylase, AADC [EC 4.1.128] or
DOPA-decarboxylase). Since this decarboxylation reaction
occurs so rapidly and since its Km (5 10-6 M) requires less
substrate than the preceding steps, tryptophan hydroxylase is
the rate-limiting step in the synthesis of serotonin. Because of
this kinetic relationship, drug-induced inhibition of serotonin
by interference with the decarboxylation step is not feasible.

It is possible to increase serotonin formation by administering

5-HTP and bypassing the rate-limiting tryptophan hydroxylase
step. AADC is widespread in distribution; it is found in the
peripheral and central nervous systems associated with
catecholamine- and serotonin-containing neurons and in the
adrenal and pineal glands. It is also found in the kidney, liver,
and various other tissues in which little or no monoamine
transmitter is normally produced. Thus, unlike tryptophan
administration, which can result in a selective increase in
serotonin in serotonin-containing neurons, 5-HTP
administration will result in the nonspecific formation of
serotonin at all sites containing AADC, including the
catecholamine-containing neurons.

Catabolism
The only effective route of continued metabolism for serotonin
is deamination by monoamine oxidase. The product of this
reaction, 5-hydroxyindoleacetaldehyde, can be further
oxidized to 5-hydroxyindoleacetic acid (5-HIAA) or reduced to
5-hydroxytryptophol, depending on the ratio of the oxidized
to the reduced form of nicotinamide adenine dinucleotide
(NAD+/NADH) in the tissue. Enzymes have been described in
the liver and brain that could catabolize 5-HT without
deamination through formation of a 5-sulfate ester. This could
then be transported out of the brain, possibly by the acid
transport system handling 5-HIAA. The brain contains an
enzyme that catalyzes the N-methylation of 5-HT using S-
adenosylmethionine as the methyl donor.

Control of Serotonin Synthesis and

Catabolism
Although there is a relatively brief sequence of synthetic and
degradative steps involved in serotonin turnover, there is still
much to be learned regarding the physiological mechanisms
for controlling this pathway. Since tryptophan hydroxylase
depends on molecular oxygen, the rate of 5-HT formation
could also be influenced by the tissue level of oxygen. In fact,
rats permitted to breathe 100% oxygen greatly increase their
synthesis of 5-HT. Also, 5-hydroxytryptophan does not inhibit
the activity of tryptophan hydroxylase.
If the situation for serotonin were similar to that previously
described for the catecholamines, we might also expect that
the concentration of 5-HT itself could influence the levels of
activity at the hydroxylation step. However, when the
catabolism of 5-HT is blocked by monoamine oxidase
inhibitors, the brain 5-HT concentration accumulates linearly
to levels three times greater than controls, thus suggesting
that end-product inhibition by serotonin is, at best, trivial.
Similarly, if the efflux of 5-HIAA from the brain is blocked by
the drug probenecid (which appears to block all forms of acid
transport), 5-HIAA levels also continue to rise linearly for
prolonged periods of time, again suggesting that the initial
synthesis step is not affected by the levels of any of the
subsequent metabolites. Two possibilities, therefore, remain
open: the initial synthesis rate may be limited only by the
availability of required cofactors or substrate such as oxygen,
pteridine, and tryptophan from the bloodstream or the initial
synthesis rate may be limited by the other more subtle
control features, more closely related to brain activity.
Indeed, the evidence suggests that impulse flow may, as in
the catecholamine systems, initiate changes in the physical
properties of the rate-limiting enzyme tryptophan
hydroxylase. Several mechanisms have been postulated for
the physiological regulation of tryptophan hydroxylase
induced by alterations in neuronal activity within serotonergic
neurons. The majority of evidence currently supports the
involvement of calcium-dependent phosphorylation in this
impulse-coupled regulatory process.

Serotonin Uptake and the Serotonin

Transporter
As with the catecholamine-containing neurons, reuptake
serves as a major mechanism for the termination of the
action of synaptic serotonin. Serotonin nerve terminals
possess high-affinity serotonin uptake sites that play an
important role in terminating transmitter action and in
maintaining transmitter homeostasis. This reuptake of
released serotonin is accomplished by a plasma membrane
carrier that is capable of transporting serotonin in either
direction, depending on the concentration gradient. Although
the involvement of transporters in norepinephrine (NE) and
serotonin clearance has been appreciated for several decades
(see Chapter 9), progress in understanding transporter
structure and regulation has been slow, mainly because of
difficulties associated with transporter protein purification.
However, this has changed with the successful expression and
homology-based cloning of the monoamine transporters (NE,
dopamine, and serotonin) and the realization that they are
members of a large gene family comprised of carriers for
other transmitters including -aminobutyric acid (GABA) and
glycine (see Chapter 6). Expression of these transporters in
nonneuronal cells has established useful model systems for
analyzing the structural basis of transporter specificity for
transmitters and antagonists. The availability of transporter
protein has also enhanced the feasibility of obtaining
transporter-specific antibodies and nucleic acid probes.
Antibodies raised against the cloned serotonin transporter
(SERT) as well as DNA and RNA probes derived from it have
enabled the localization and expression of SERT in the CNS.
The SERT protein is ubiquitous in the CNS, which is consistent
with its transport to the nerve terminals of the extensive
projections of the serotonin neurons throughout the brain and
spinal cord. This is in contrast to SERT mRNA expression,
which occurs almost exclusively in the serotonergic cell bodies
in the raphe nuclei; especially high levels are found in the
median and dorsal raphe. SERT mRNA expression is absent
from other brain stem nuclei, including the substantia nigra
and locus ceruleus. The availability of transporter-specific
antibodies and nucleic acid probes has made it feasible to
investigate the endogenous mechanism that acutely regulates
monoamine transporters in vivo and to determine whether
chronic alterations in transporter genes underlie
neuropsychiatric disorders. To date, however, the cloning of
SERT has not had as big an impact on the field as the cloning
of serotonin receptor subtypes. So far, there is little evidence
for SERT heterogeneity. However, this transporter might be
subject to regulation. In fact, SERT uptake capacity (Vmax) is
regulated by kinase-linked pathways, particularly those
involving protein kinase C (PKC), resulting in transporter
phosphorylation and sequestration. Ligand occupancy of the
transporter significantly impacts both SERT phosphorylation
and sequestration. Ligands such as serotonin and
amphetamine, which permeate the transporter, prevent PKC-
dependent SERT phosphorylation. In contrast, nontransported
antagonists like cocaine and antidepressants are permissive
for SERT phosphorylation but effectively block serotonin's
effects. Hints of altered SERT gene regulation following
hormonal stimulation also suggest that significant information
might be acquired from systematic analysis of genomic
regulatory elements that control transporter expression.
Clearly, a goal for the future is to determine whether
hereditary genetic variations in such systems contribute to
psychiatric disorders.

PINEAL BODY

Introduction
The pineal organ is a tiny gland (1 mg or less in the rodent)
contained within connective tissue extensions of the dorsal
surface of the thalamus. While physically connected to the
brain, the pineal is cytologically isolated for all intents since,
as one of the circumventricular organs, it is on the
"peripheral" side of the blood-permeability barriers (see
Chapter 2) and its innervation arises from the superior
cervical sympathetic ganglion. The pineal is of interest for two
reasons. First, it contains all of the enzymes required for the
synthesis of serotonin plus two enzymes for further
processing of serotonin, which are not so pronounced in other
organs. The pineal contains more than 50 times as much 5-
HT (per gram) as the whole brain. Second, the metabolic
activity of the pineal 5-HT enzymes can be controlled by
numerous external factors, including the neural activity of the
sympathetic nervous system operating through release of
norepinephrine. As such, the pineal appears to offer a
potential model for the study of brain 5-HT. The pineal also
appears to contain many neuropeptides, however, and thus its
secretory role remains as clouded as ever.

Actually, the 5-HT content of the pineal was discovered after

the isolation of a pineal factor, melatonin, known to induce
pigment-lightening effects on skin cells. When melatonin was
crystallized and its chemical structure determined to be 5-
methoxy-N-acetyltryptamine, an indolealkylamine, a
reasonable extension was to analyze the pineal for 5-HT itself.
The production of melatonin from 5-HT requires two
additional enzymatic steps. The first is the N-acetylation
reaction to form N-acetylserotonin. This intermediate is the
preferred substrate for the final step, the 5-hydroxyindole-O-
methyltransferase reaction, requiring S-adenosylmethionine
as the methyl donor.

The melatonin content, and its influence in supressing the

female gonads, is reduced by environmental light and
enhanced by darkness. The established cyclic daily rhythm of
both 5-HT and melatonin in the pineal is driven by
environmental lighting patterns through sympathetic
innervation. In animals made experimentally blind, the pineal
enzymes and melatonin content continue to cycle but with a
rhythm uncoupled from lighting cycles. The adrenergic
receptors of the pineal are of the type, and, as is
characteristic of such receptors, their effect on the pineal is
mediated by the postjunctional formation of 3 ,5 -cyclic
adenosine monophosphate (cAMP). Elevated levels of cAMP
occur within minutes of the dark phase and lead to an almost
immediate activation of 5-HT-N-acetyltransferase. The same
receptor action also appears to be responsible on a longer
time scale for tonic enzyme synthesis. Thus, the proposed use
of the pineal as a model applicable to brain 5-HT loses its
luster since this regulatory step does not seem to be of
functional importance in the CNS. Furthermore, the
adrenergic sympathetic nerves also can accumulate 5-HT
(which leaks out of pinealocytes) just as they accumulate and
bind NE. It remains to be shown whether this is a functional
mistake (i.e., secretions of an endogenous false transmitter)
or simply a case of mistaken biological identity. It seems
likely, however, that it is the NE whose release is required to
pass on the intended communications from the sympathetic
nervous system since only NE activates the pinealocyte
cyclase to start the enzyme regulation cascade.

Localizing Brain Serotonin to Nerve Cells

Serotonin-containing neurons are restricted to clusters of cells
lying in or near the midline or raphe regions of the pons and
upper brain stem (Fig. 10-3). In addition to the nine 5-HT
nuclei (B1-B9) originally described by Dahlstrom and Fuxe,
the immunocytochemical localization of 5-HT has detected
reactive cells in the area postrema and the caudal locus
ceruleus, as well as in and around the interpeduncular
nucleus. The more caudal groups, studied by electrolytic or
chemically induced lesions, project largely to the medulla and
spinal cord. The more rostral 5-HT cell groups (raphe dorsalis,
raphe medianus, and centralis superior, or B7-B9 [Fig. 10-4])
are thought to provide the extensive 5-HT innervation of the
telencephalon and diencephalon. The intermediate groups
may project into both ascending and descending groups, but
since lesions here also interrupt fibers of passage, discrete
mapping has required analysis of the orthograde and
retrograde methods. Immunocytochemical studies have also
revealed a far more extensive innervation of the cerebral
cortex, which, unlike the noradrenergic cortical fibers, is quite
patternless in general.

In part, these studies could be viewed as disappointing in that

most raphe neurons appear to innervate overlapping terminal
fields and thus are more NE-like than dopamine-like in their
lack of obvious topography. Exceptions to this generalization
are that the B8 group (raphe medianus) appears to furnish a
very large component of the 5-HT innervation of the limbic
system, while B7 (or dorsal raphe) projects with greater
density to the neostriatum, cerebral and cerebellar cortices,
and thalamus (Fig. 10-5).

In the past, attempts to localize 5-HT-containing terminals

relied primarily on the uptake of reactive 5-HT analogues or
radiolabeled 5-HT. The use of labeled 5-HT, electron-dense
analogues, or 5-HT-selective toxins (e.g., the
dihydroxytryptamines) depended for specificity on the
selectivity of the uptake process. This situation has been
rectified by immunocytochemistry of endogenous 5-HT (Figs.
10-5, 10-6), employing antibodies directed against serotonin.

With this more sensitive technique, it has become clear that

the cerebral cortex in many mammals is innervated by two
morphologically distinct types of 5-HT axon terminal. Fine
axons with small varicosities originate from the dorsal raphe
nuclei, and beaded axons with large spherical varicosities
arise from the median raphe nuclei. These two types of 5-HT-
containing axon have different regional and laminar
distributions and appear to be differentially sensitive to the
neurotoxic effects of certain amphetamine derivatives,
including 3,4-methylenedioxymethamphetamine (MDMA),
referred to more commonly as "ecstasy." The fine axons are
much more sensitive to the neurotoxic effects than the
beaded axons, and the loss of fine axons lasts for months and
may be permanent. Beaded axons appear to be resistant and
remain unaffected following neurotoxic treatment with MDMA.
This finding may be relevant to human studies, which have
indicated that individuals using MDMA as a recreational drug
may be exposed to dosages approximating those shown to
exhibit serotonin neurotoxicity in nonhuman primates. A 26%
decrease in the serotonin metabolite 5-HIAA was observed in
the cerebrospinal fluid of MDMA users. This indirect evidence
of a decrease in serotonin turnover in the brain perhaps
reflects destruction or compromised function of this fine
serotonin-containing axon system. Further studies of MDMA
users seem warranted and could provide important
information on the effects of selective loss of this fine axon
system in humans. At present, the functional roles played by
the fine and beaded axon systems and whether the functions
are distinct or similar remain unclear. In serial section analysis
of 5-HT terminals in the primate visual cortex, the fine and fat
boutons appeared to coexist in the same axon, arguing
against distinct 5-HT innervation of this brain region.

Figure 10-3. Schematic diagram illustrating the

distribution of the main serotonin-containing
pathways in the rat central nervous system.
(Modified from Breese, 1975.)

Figure 10-4. Fluorescence micrograph of raphe

cell bodies in the mid-brain of the rat. This rat was
pretreated with l-tryptophan (100 mg/kg) 1 hour
prior to death (Courtesy of G. K. Aghajanian,
1995).

Figure 10-5. Photomicrograph of the serotonergic

neurons of the caudal portion of the dorsal raphe.
Section was stained using an antibody directed
against serotonin. The serotonergic innervation of
the dorsal tegmental nucleus (DTN) can also be
seen. V marks the third ventricle. (Courtesy of A.
Y. Deutch.)
 Figure 10-6. Darkfield photomicrograph
illustrating the serotonergic innervation of the bed
nucleus of the stria terminalis, as revealed by
immunohistochemical staining with antibody
directed against serotonin. The innervation of the
bed nucleus of the anterior commissure (BAC) can
also be seen above the anterior commissure (AC).
LV marks the lateral ventricle. (Courtesy of A. Y.
Deutch.)

CELLULAR EFFECTS OF 5-HT

Introduction
From the biochemical and morphological data discussed
above, we can be relatively certain that the 5-HT of the brain
occurs not only within the nerve cells but also within specific
tracts or projections of nerve cells. We must now inquire into
the effects of serotonin when applied at the cellular level (see
Chapter 2). In those brain areas where
microelectrophoretically administered 5-HT has been tested
on cells that exhibit spontaneous electrical activity, the
majority of cells decrease their discharge rate. The effects
observed typically last much longer than the duration of the
microelectrophoretic current. In other regions, however, 5-HT
also causes pronounced activation of discharge rate.

Electrophysiological analyses of 5-HT have focused on

neurons of the facial motor nucleus (cranial nerve VII), where
innervation by 5-HT fibers has been well documented. With
intracellular recordings from these cells, 5-HT is found to
produce a slow, depolarizing action accompanied by a modest
increase in membrane resistance. As might then be
anticipated on biophysical grounds, the combination of
depolarization and increased membrane resistance facilitates
the response of these neurons to other excitatory inputs. In a
rigorous sense, such effects are not exactly in keeping with
the emerging characteristics of modulatory actions (see
Chapter 2) since here iontophoretic 5-HT changes both
membrane potential and resistance on its own. It will be of
interest to determine if activation of a 5-HT pathway to these
cells at levels that do not in themselves directly change
membrane properties will nevertheless modify responses of
the target cells to other inputs, analogous to the effects
described for noradrenergic connections.

Several specific 5-HT-containing tracts, investigated

electrophysiologically, indicate that 5-HT produces mainly, if
not exclusively, inhibitory effects. While the pathways do
conduct slowly, as would be expected for such fine-caliber
axons, the synaptic mediation process appears to be relatively
prompt.

Characterization of 5-HT Receptors

The existence of multiple receptors for serotonin in the CNS
has been suggested by physiological studies. Radioligand
binding studies demonstrating that H3-5-HT and H3-
spiperone label separate populations of high-affinity binding
sites for 5-HT, termed 5-HT1 and 5-HT2, respectively, have
also provided evidence for multiple receptors in brain tissue.
A high correlation between the potencies of 5-HT antagonists
in displacing the binding of H3-spiperone and inhibiting
serotonin-induced behavioral hyperactivity, coupled with a
lack of such a correlation of H3-5-HT binding sites, has led to
the proposal that inhibition and excitation induced by
serotonin are mediated at 5-HT1 and 5-HT2 receptors,
respectively. Additional binding studies have further
subdivided the 5-HT1 class of recognition sites into 5-HT1A
and 5-HT1B subtypes, and compounds selective for these
receptor subtypes have been identified. Autoreceptors for 5-
HT appear to fall into the 5-HT1A category. Drugs such as 8-
hydroxydipropylaminotetralin (8-OH DPAT) and the long-chain
substituted piperazines appear to have selective 5-HT1A
binding activity, and these 5-HT autoreceptor agonists are
effective at inhibiting midbrain raphe neurons. Ketanserin
blocks 5-HT receptors that belong to the 5-HT2 category
without having any significant effect on 5-HT1 receptors.
Because of this discrimination, it has been described as a
selective 5-HT2 blocker. However, it is important to note that
ketanserin is nonspecific in the sense that it has appreciable
affinity for both 1-adrenergic receptors and histamine H1
receptors as well. It does have the advantage that its action
on 5-HT receptors is as a pure antagonist.

5-HT Receptors
It is more than 40 years since Gaddum and Picarelli (1957),
basing their work on the pharmacological properties of
serotonin agonists and antagonists, reported evidence for two
separate serotonin receptors in peripheral smooth muscle
preparations studied in vitro. One they named the D receptor
because it was blocked by dibenzyline, and the other they
called the M receptor because the indirect contractile
response to 5-HT mediated by the release of acetylcholine
from cholinergic nerves in the myenteric plexus could be
antagonized by morphine. The advent of receptor-binding
studies in the 1970s revealed the existence of multiple
binding sites, but while a correlation existed between the
pharmacological profile of the 5-HT2 binding site and the D
receptor, no such correlation existed between the 5-HT1
receptor subtypes identified by receptor-binding techniques
and the M receptor. Since the M receptor originally described
in the nerves of the guinea pig ileum is pharmacologically
distinct from the 5-HT1 and 5-HT2 receptors and their
numerous subtypes, it was renamed the 5-HT3 receptor. With
the introduction of several extremely potent and highly
selective 5-HT3 antagonists, attention has shifted to this
receptor and its possible functions in both the periphery and
the CNS, as well as the therapeutic potential for these newly
developed 5-HT3 compounds.

At least eight subtypes of serotonin receptor in brain tissue

have been defined and characterized, based on radioligand
binding studies. As noted in Chapter 5, it is premature to
characterize as a receptor a binding site defined in this
manner. Nevertheless, the majority of the original
speculations about serotonergic receptor subtypes generated
from radioligand binding experiments appear to have been
substantiated by many other types of experiment, including,
most recently, the cloning of three of these subtypes. Table
10-1 lists the characteristics of these 5-HT receptor subtypes.

Molecular Biology
Since the mid-1980s, a vast amount of new information has
become available concerning the various 5-HT receptor
subtypes and their functional and structural characteristics.
This derives from two main research approaches, operational
pharmacology employing selective agonists and antagonists
and molecular biology. With the advent of the latter
technique, the field of 5-HT receptors has experienced
exceptionally rapid growth over the past 5 years and the
existence of multiple 5-HT receptors has been unequivocally
confirmed. Because medicinal chemistry has lagged behind
molecular biology in 5-HT neuropharmacology, there are very
few highly selective receptor agonists and antagonists for the
individual 5-HT receptor subtypes.

Serotonin receptors are highly heterogeneous, and cloning

not only has led to the discovery and recognition of previously
unknown receptors but has greatly facilitated their
classification. As of the new millenia, there are at least 15
molecularly identified 5-HT receptors, some with splice
varients and others with isoforms created by mRNA editing.
The majority of the 5-HT receptors belong to the large family
of receptors interacting with G proteins, except for the 5-HT3
receptors, which are ligand-gated ion channel receptors (see
Table 10-2). The 5-HT receptors belonging to the G protein
receptor superfamily are characterized by the presence of
seven transmembrane domains and the ability to alter G
protein-dependent processes. The amino acid sequence of the
membrane-spanning domains shows the least amount of
variability compared with other cloned biogenic amine
receptors.

This group of 5-HT receptors can be divided into distinct

families based on their coupling to second messengers and
their amino acid sequence homology. The 5-HT1 family
contains receptors that are negatively coupled to adenylyl
cyclase. The 5-HT2 family contains three receptors that have
striking amino acid homology and the same coupling with
second messenger, that is, activation of phospholipase C. The
5-HT receptors that are positively coupled to adenylyl cyclase
are a heterogeneous group, including the 5-HT4, 5-HT6, and
5-HT7 subtypes. The 5-HT5 group contains two types, 5HT5A
and 5-HT5B, and represents a new family of 5-HT receptors
that do not resemble receptors of the 5-HT1 and 5-HT2
families in terms of amino acid sequence, pharmacological
profile, and transduction system. They are probably coupled
to a different effector system. In contrast to the G protein-
coupled 5-HT receptors that modulate cell activities via
second-messenger systems, 5-HT3 receptors directly activate
a 5-HT-gated cation channel that rapidly and transiently
depolarizes a variety of neurons. Like other ligand-gated ion
channels, the 5-HT3 receptor consists of four transmembrane
segments and a large extracellular N-terminal region
incorporating a cysteine-cysteine loop and potential N-
glycosylation sites. Other members of this molecular receptor
family include GABA, glutamate, glycine, and the nicotinic
cholinergic receptors.

A third major molecular recognition site for 5-HT is the

transporter proteins. These transporter proteins consist of 12
membrane-spanning proteins and represent a large gene
family encoding Na+ and Cl--dependent transport proteins.
The first 5-HT transporter was identified in rat brain, but 5-HT
transporters have been characterized in other tissues,
including platelets, placenta, lung, and basophilic leukemia
cells. A number of selective 5-HT uptake blockers have been
developed, such as fluoxetine, sertraline, citalopram, and
paroxetine (see Fig. 10-7), and several have exhibited clinical
utility in the treatment of depression and obsessive-
compulsive disorders.

The careful categorization of old and new subtypes of 5-HT

receptors should be an important foundation for defining their
function. The development of selective agents targeting these
receptors, coupled with the molecular approaches highlighted
in Chapter 3, such as antisense oligonucleotides to decrease
steady-state levels of target proteins and transgenic animals
in which specific receptors have been "knocked out," will
undoubtedly enhance our understanding of the function of
central serotonergic systems. The existence of a large number
of receptors with distinct signaling properties and expression
patterns might enable a single substance like 5-HT to
generate simultaneously a large array of effects in many
discrete brain structures.

Figure 10-7. Selective inhibitors of the serotonin

transporter.
Behavioral Aspects of Serotonin Function
5-HT neurons in the brain stem raphe nuclei exhibit
spontaneous monotonic activity, discharging in a clock-like
manner with an intrinsic frequency of 1 to 5 spikes/second. In
the rodent, these monotonic properties are manifested early
in development (3 to 4 days before birth). The 5-HT neurons
appear to possess a negative feedback mechanism that limits
their neuronal activity. As the physiological activity of the 5-
HT neuron increases, the local release of 5-HT from dendrites
or axonal collaterals acts on somatodendritic 5-HT
autoreceptors to inhibit neuronal activity. This autoregulatory
mechanism seems to function only under physiological
conditions and to be inoperative during periods of low-level
activity, but it becomes functional as neuronal activity
increases. Dysfunction of this autoregulatory mechanism has
been implicated in some forms of human neuropathology, so
autoreceptors have become a potentially important site for
drug-targeted therapeutic intervention. The combined use of
5-HT1B autoreceptor antagonists and selective 5-HT uptake
blockers holds some promise.

In view of the extraordinarily widespread projections and

highly regulated pacemaker pattern of activity that is
characteristic of serotonin neurons, a broad homeostatic
function has been suggested for serotonergic systems. By
exerting simultaneous modulatory effects on neuronal
excitability in diverse regions of the brain and spinal cord, the
serotonergic system is in a strategic position to coordinate
complex sensory and motor patterns during varied behavioral
states. Single-unit electrophysiological recordings from
serotonergic neurons in unanesthetized animals have shown
that serotonergic activity is highest during periods of waking
arousal, reduced in quiet waking, reduced further in slow-
wave sleep, and absent during rapid-eye-movement sleep. An
increase in tonic activity of serotonergic neurons during
waking arousal would enhance motor neuron excitability via
descending projections to the ventral horn of the spinal cord.
Suppression of sensory input, conversely, would screen out
distracting sensory cues. Cessation of serotonergic neuronal
activity during rapid-eye-movement sleep would tend to
impede motor function in this paradoxical state in which
internal arousal is associated with diminished motor output.
Altered function of serotonergic systems has been reported in
several psychopathological conditions, including affective
illness, hyper-aggressive states, and schizophrenia. There is
mounting evidence for impaired serotonergic function in
major depressive illness and suicidal behavior. In this
connection, several effective antidepressant drugs appear to
act by enhancing serotonergic transmission.

The pathophysiology of major affective illness is poorly

understood, but several lines of clinical and preclinical
evidence indicate that enhancement of 5-HT-mediated
neurotransmission might underlie the therapeutic response to
different types of antidepressant treatment (see Chapter 13).
Table 10-3 shows the effects of long-term administration of
different types of antidepressant on the 5-HT system
assessed with electrophysiological techniques. Treatment with
all of these drugs appears to cause a net increase in 5-HT
neurotransmission. Several clinical observations have also
provided strong evidence of a pivotal role for 5-HT
neurotransmission in depression. For example, a large
number of selective 5-HT reuptake inhibitors examined
clinically have been found to be effective in major depression.
These drugs belong to different chemical families but appear
to share a single common property, the ability to inhibit the 5-
HT reuptake carrier. In addition, clinical studies show that
inhibition of 5-HT synthesis in drug-remitted depressed
patients, using either the tryptophan hydroxylase inhibitor p-
chlorophenylalanine or the tryptophan depletion paradigm,
produces a rapid relapse of depression. In the latter
paradigm, the symptomatology reactivated by the tryptophan
depletion was nearly identical to that present before the
response to the antidepressant treatment, suggesting a
causal relationship.

Signal-Transduction Pathways
Only recently has it been possible to examine the
mechanisms of signal transduction in central 5-HT receptor
systems, and evidence is emerging that major transducing
systems are linked to different 5-HT receptors in the
mammalian brain.

In general, two major 5-HT receptor-linked signal-

transduction pathways exist: direct regulation of ion channels
and a multistep enzyme-mediated pathway. Both pathways
require a guanine nucleotide triphosphate (GTP)-binding
protein (G protein) to link the receptor to the effector
molecule. The 5-HT1 family of receptors is negatively coupled
to adenylyl cyclase via the G1 family of G proteins. The 5-
HT1A receptor is the best characterized of this family. This
receptor, in addition to coupling with adenylyl cyclase, is
linked directly to voltage-sensitive K+ channels via Gi-like
proteins. This direct coupling with both adenylyl cyclase and
the K+ channel is a recognized characteristic of Gi-linked
receptors. Direct coupling to L-type Ca2+ channels has also
been described as an additional transduction pathway for the
Gi-linked family of receptors. The other members of the 5-
HT1 family of receptors have also been shown to be
negatively coupled to adenylyl cyclase (Table 10-2). The 5-
HT2 family of receptors, in contrast to the 5-HT1 family, is
coupled to phospholipase C. The G protein involved has not
been identified but is assumed to be a member of the Gq
family. Phospholipase C activation induces diverse changes in
the cell, leading to regulation of numerous cellular processes.
All members of this family appear to be coupled primarily to
phospholipase C, leading to phosphoinositide hydrolysis.
Although stimulation of adenylyl cyclase was the first signal-
transduction pathway to be linked to 5-HT, the specific
receptors mediating activation of adenylyl cyclase, the 5-HT4,
5-HT6, and 5-HT7 receptors, were identified only recently.
The 5-HT4 receptor is found in rodent brain (hippocampus)
and peripheral tissue, including the guinea pig ileum and
human atrium. The 5-HT7 receptor is also found in the brain
and heart (see Table 10-1). Another novel receptor, 5-HT6, is
the most recent serotonin receptor to be identified by
molecular cloning. The 5-HT6 receptor has also been shown
conclusively to couple positively to adenylyl cyclase and to
have a high affinity for tricyclic antidepressant drugs. The
interesting distribution of this receptor in the brain coupled
with its high affinity for atypical antipsychotic and tricyclic
antidepressant drugs has led to significant efforts to learn
more about its function and possible role in psychiatric
disorders. While our knowledge is far from complete,
emerging data suggest that 5-HT6 receptors regulate
cholinergic neurotransmision in the brain rather than the
anticipitated modulation of dopaminergic function.

The 5-HT3 receptor differs from other 5-HT receptors by

forming an ion channel that regulates ion flux in a G protein-
independent manner. This receptor is a member of a large
family of ligand-gated ion channels and thus shares more
similarities with the nicotinic cholinergic receptor, which is the
prototype of this superfamily, than with the 5-HT1 and 5-HT2
receptor families. The 5-HT3 receptors were first found on
peripheral sensory, autonomic, and enteric neurons, where
they mediate excitation. The direct demonstration that 5-HT3
receptors in the guinea pig submucosal plexus are ligand-
gated ion channels implies a role for 5-HT as a "fast" synaptic
transmitter and fits with their function in the periphery.
Molecular studies indicated that the cloned 5-HT3 receptor
protein forms a homomeric subunit, which regulates the
gating of cations and thus presumably mediates the rapid and
transient depolarization that occurs following 5-HT3 receptor
activation. Immunocytochemical studies have revealed that 5-
HT3 receptor-immunoreactive neurons are broadly distributed
throughout the rat brain and spinal cord and suggest that this
receptor can subserve significant participation in CNS
neurotransmission. However, the neuronal circuits in which
this receptor might participate and the functions subserved by
it remain to be established.

In general, the pharmacological and electrophysiological

characteristics of the cloned receptor are largely consistent
with the properties of the native receptors. Despite this
consistency, however, a number of biochemical,
pharmacological, and electrophysiological studies have
suggested that CNS 5-HT3 receptors exhibit heterogeneity
with respect to subtypes and intracellular signal-transduction
mechanisms. Thanks to the pharmacological similarities
between the 5-HT M receptors and the 5-HT3 receptors, we
have numerous pharmacological agents that exert specific
effects on 5-HT3 receptors, and many studies using these
agents have addressed the possible functions of 5-HT3
receptors. Results obtained from these in vivo studies have
also suggested the presence of multiple 5-HT3-like receptors,
spurring on the search for other subunit proteins that might
explain this pharmacological heterogeneity. The possibility
should be entertained that 5-HT3 receptors are analogous to
the glutamate receptors, which were first characterized as
inotropic receptors and later shown to exist also as G protein-
linked metabotropic receptors. Clearly, additional studies are
required to determine whether the 5-HT3 receptor
heterogeneity can be explained by subtypes of 5-HT3 receptor
from both the inotropic and metabotropic superfamilies. The
discovery of multiple subtypes of 5-HT3 may help to clarify
the electrophysiological observation that 5-HT3-like receptors
in the prefrontal cortex elicit a slow depression of cell firing
rather than the fast activation expected for a ligand-gated 5-
HT3 receptor.

Recent data suggest that 5-HT3 receptors are coupled to an

ion channel, probably a calcium channel. Thus, they share
more similarities with the nicotinic cholinergic receptor than
with the 5-HT1 or 5-HT2 receptors. The direct demonstration
that 5-HT3 receptors in guinea pig submucosal plexus are
ligand-gated ion channels implies a role for 5-HT (and
perhaps for other biogenic amines) as a "fast" synaptic
transmitter.

Adaptive Regulation
5-HT2A and 5-HT2C receptors adapt to chronic activation by
reducing response sensitivity or receptor density as expected,
but chronic inactivation does not elicit the opposite adaptive
response. Central 5-HT2 and 5-HT2C receptors seem to be
relatively resistent to upregulation. For example, 5-HT2A
receptors are not upregulated after denervation of 5-HT
neurons or chronic administration of 5-HT antagonists.
Instead, chronic administration of 5-HT antagonists results in
a paradoxical downregulation of both 5-HT2A and 5-HT2C
receptors.

Physiology
Neurophysiological and behavioral studies have benefited
from the development and characterization of selective agents
for 5-HT receptors. Electrophysiological studies have clearly
demonstrated that 5-HT1A receptors mediate inhibition of the
raphe nuclei. The firing of serotonergic neurons is tightly
regulated by intrinsic ionic mechanisms (e.g., calcium-
activated potassium conductance), which accounts for the
well-known tonic pacemaker pattern of activity of these cells.
The intrinsic pacemaker is modulated by at least two
neurotransmitters: (1) NE acting through adrenergic
receptors accelerates the pacemaker and (2) 5-HT acting
through somatodendritic 5-HT1A autoreceptors slows the
pacemaker. In the hippocampus, which is another anatomical
structure containing a high density of 5-HT1A sites, 5-HT1A
agonists hyperpolarize CA1 pyramidal cells by opening
potassium channels via a pertussis toxin-sensitive G protein.
Electrophysiological studies carried out in Xenopus oocytes,
which express the 5-HT1C receptor, revealed that application
of 5-HT causes a detectable inward current that is blocked by
mianserin. Activation of the 5-HT1C receptor apparently
liberates inositol phosphates, raising intracellular Ca2+ levels
and leading to the opening of Ca2+-dependent chloride
channels. In mammalian systems, two specific
neurophysiological effects have been attributed to activation
of the 5-HT2 receptor. 5-HT facilitates the excitatory effects of
glutamate in the facial motor nucleus, an action that is
antagonized by 5-HT2 antagonists. Similar data were
obtained from intracellular studies that showed that activation
of 5-HT receptors caused a slow depolarization of facial motor
neurons and increased input resistance, leading to increased
excitability of the cell, probably through a decrease in resting
membrane conductance to potassium. These data suggest
that 5-HT2 receptors mediate the 5-HT-induced excitation of
facial motor neurons.

It has also been shown that 5-HT causes a slow depolarization

of cortical neurons that is associated with decreased
conductance. The effect can be desensitized by repeated
applications of 5-HT and blocked by the selective 5-HT2
antagonist ritanserin. Thus, the effects of 5-HT on cortical
pyramidal neurons share many similarities to the depolarizing
effects observed in the facial motor nucleus and appear to be
mediated by 5-HT2 receptors.

5-HT3 receptors in peripheral nervous tissue mediate

excitatory responses to 5-HT and are involved in modulating
transmitter release. These receptors are also found in the
CNS, where they are present in high density in the entorhinal
cortex and area postrema. Release of endogenous dopamine
by stimulation of 5-HT3 receptors in the striatum occurs in a
calcium-dependent fashion. Even though the physiological
function of 5-HT3 receptors in the CNS is unclear, the
observation that selective 5-HT3 antagonists possess central
activity in rats and primates in anxiolytic and
antidopaminergic-like behavioral models has generated
considerable excitement and speculation about the potential
therapeutic use of 5-HT3 antagonists and agonists. From
animal experiments, it has been speculated that these
compounds may be useful in treating schizophrenia, pain,
anxiety, drug dependence, and cytotoxic drug-induced
emesis; but so far only the analgesic and antiemetic effects
have been demonstrated in humans.

Electrophysiology of 5-HT Receptors

Knowledge of the molecular biology of 5-HT receptors has
revolutionized electrophysiological approaches to investigating
the 5-HT systems in brain since studies can now be directed
toward neurons that express specific 5-HT receptor subtypes
based on in situ hybridization maps of receptor mRNA
expression. This has enabled the diverse electrophysiological
actions of 5-HT in the CNS to be categorized according to
receptor subtypes and their respective effects or mechanisms
of action. Following are a few generalizations from these
studies: (1) inhibitory effects of 5-HT are mediated by 5-HT1
receptors linked to the opening of K+ channels or to the
closing of Ca2+ channels, both via pertussis toxin-sensitive G
proteins; (2) facilitative effects of 5-HT that are mediated by
5-HT2 receptors and involve the closing of K+ channels can
be modulated by the phosphatidylinositol second-messenger
system and PKC acting as a negative feedback loop; (3) other
facilitative effects of 5-HT appear to be mediated by 5-HT4
and 5-HT7 receptors by a reduction in certain voltage-
dependent K+ currents mediated through the protein kinase A
phosphorylation pathway and thus involving positive coupling
of the 5-HT response to adenylyl cyclase; (4) fast excitations
are mediated by 5-HT3 receptors through a ligand-gated
cationic ion channel that does not require coupling with a G
protein or a second messenger. Thus, it is clear that the
electrophysiological actions of 5-HT encompass the two major
neurotransmitter gene superfamilies, the G protein-coupled
receptors and the ligand-gated cationic channels. The end
effects are determined by the receptor subtype and its
anatomical location (Fig. 10-8).

Figure 10-8. Model of a serotonin (5-HT)

synapse. Tryptophan is taken up into the neuron
by an active transport mechanism and converted
to 5-OH tryptophan (5-HTP) by the rate-limiting
enzyme tryptophan hydroxylase. Aromatic amino
acid decarboxylase (AADC) converts 5-HTP to 5-
HT. 5-HT is taken up and stored by the VMAT.
Once released, 5-HT can interact with as many as
15 different receptors. 5-HT autoreceptors (5-
HT1B rodent, and 5-HT1D human), which
modulate the stimulus-induced release of 5-HT,
are located on the presynaptic terminals. 5-HT1A
autoreceptors are located on the 5-HT cell bodies
 and dendrites where they modulate impulse flow.
A number of G protein-coupled 5-HT receptors are
located postsynaptically. The 5-HT3 receptor is
also localized postsynaptically but in contrast to
the other 5-HT receptors is a ligand-gated ion
channel similar to the cholinergic nicotinic
receptor. The action of 5-HT is terminated by
reuptake into the neuron by SERT. AC, adenylyl
cyclase; DAG, diacylglycerol; IP3, inositol
triphosphate; PLC, phospholipase C; SERT, plasma
membrane serotonin transporter; VMAT, vesicular
monoamine transporter. (Modified from Nesther et
al., Molecular Neuropharmacology.)

Relevance to Clinical Disorders and Drug

Actions
The diversity of receptors and transduction pathways that
underlie the actions of 5-HT, together with the differential
expression of 5-HT receptor subtypes in different neuronal
and effector cell populations, helps to explain how it is
possible for a single transmitter to be linked to such a large
array of behaviors, clinical conditions, and drug actions.
Alterations in 5-HT function have been implicated in a host of
clinical states, including affective disorder, obsessive-
compulsive disorder, schizophrenia, anxiety states, phobic
disorders, eating disorders, migraine, and sleep disorders
(see Table 10-4). There is also a wide range of psychotropic
drugs that affect 5-HT neurotransmission, including
antidepressants (selective 5-HT uptake blockers, e.g.,
fluoxetine [see Fig. 10-7]), hallucinogens (e.g., LSD, psilocin
[see Fig. 10-1]), anxiolytics (e.g., buspirone), atypical
antipsychotics (e.g., clozapine), antiemetics (e.g.,
ondansetron), appetite-suppressants (e.g., fenfluramine), and
antimigraine drugs (e.g., sumatriptan).
Behavioral Effects of 5-HT Systems
One of the first drugs found empirically to be effective as a
central tranquilizing agent, reserpine, was employed mainly
for its action in the treatment of hypertension. In the early
1950s, when both brain serotonin and the central effects of
reserpine were first described, great excitement arose when
the dramatic behavioral effects of reserpine were correlated
with loss of 5-HT content. However, it was soon found that NE
and dopamine content could also be depleted by reserpine
and that all of these depletions lasted longer than sedative
actions. Brain levels of all three amines remained down for
many days, while the acute behavioral effects ended in 48-72
hours. Hence, it became difficult to determine whether it was
the loss of brain catecholamine or serotonin that accounted
for the behavioral depression after reserpine.

More extensive experiments into the nature of this problem

were possible when the drugs that specifically block the
synthesis of catecholamines or serotonin were discovered.
After depleting brain serotonin content with p-
chlorophenylalanine, which effectively removed 90% of brain
serotonin, investigators observed that no behavioral
symptoms reminiscent of the reserpine syndrome appeared.
Moreover, when p-chlorophenylalanine-treated rats, which
were already devoid of measurable serotonin, were treated
with reserpine, typical reserpine-induced sedation arose.
These results again favor the view that loss of catecholamines
could be responsible for the reserpine-induced syndrome.
Furthermore, when -methyl-p-tyrosine was given and
synthesis of NE was blocked for prolonged periods of time,
the animals were behaviorally sedated and their condition
resembled the depression seen after reserpine. Students will
find it profitable to review in detail the original papers
describing the results just summarized and the other theories
currently in vogue.
Hallucinogenic Drugs
One of the more alluring aspects of the study of brain
serotonin is the possibility that it is this system of neurons
through which the hallucinogenic drugs cause their effects. In
the early 1950s, the concept arose that LSD might produce its
behavioral effects in the brain by interfering with the action of
serotonin there as it did in smooth muscle preparations, such
as the rat uterus. However, this theory of LSD action was not
supported by the finding that another serotonin-blocking
agent, 2-bromo-LSD, produced minimal behavioral effects in
the CNS. It was subsequently shown that very low
concentrations of LSD itself, rather than blocking serotonin
action, could potentiate it. However, none of these data could
be considered particularly pertinent since all of the research
was done on the peripheral nervous system and all of the
philosophy was applied to the CNS.

Shortly thereafter, Freedman and Giarman initiated a

profitable series of experiments investigating the basic
biochemical changes in the rodent brain following injection of
LSD. Although their initial studies required them to use a
bioassay for changes in serotonin, they detected a small (on
the order of 100 ng/g or less) increase in the serotonin
concentration of the rat brain shortly after injection of very
small doses of LSD. Subsequent studies have shown that a
decrease in 5-HIAA accompanies the small rise in 5-HT.
Although the biochemical effects are similar to those that
would be seen from small doses of monoamine oxidase
inhibitors, no direct monoamine oxidase inhibitory effect of
LSD has been described. This effect was generally interpreted
as indicating a temporary decrease in the rate at which
serotonin was metabolized, and it could also be seen with
higher doses of less effective psychoactive drugs. In related
studies by Costa and co-workers, who estimated the
biochemical turnover of brain serotonin, prolonged infusion of
somewhat larger doses of LSD clearly promoted a decrease in
the turnover rate of brain serotonin.

The next advance in the explanation of the LSD response was

made when Aghajanian and Sheard reported that electrical
stimulation of the raphe nuclei selectively increased the
metabolism of 5-HT to 5-HIAA. This finding suggested that
the electrical activity of 5-HT cells could be directly reflected
in the metabolic turnover of the amine. Subsequently, the
same authors recorded single raphe neurons during
parenteral administration of LSD and observed that they
slowed down, with a time course similar to the effect of LSD.
Thus, following LSD administration, both decreased electrical
activity of these cells and decreased transmitter turnover
occur.

Of the several explanations originally proposed for this effect,

the data now support the view that LSD depresses directly 5-
HT-containing neurons at receptors, which may be the sites
where raphe neurons feed back 5-HT messages to each other
through recurrent axon collaterals or dendrodendritic
interactions; at these receptive sites on raphe neurons, LSD is
a 5-HT agonist. In other tests on raphe neurons, iontophoretic
LSD antagonizes the activity of raphe neurons that show
excited responses to either NE or 5-HT. However, at sites in 5-
HT terminal fields, when LSD is evaluated as an agonist or
antagonist of 5-HT, its effects are considerably weaker than
on the raphe neurons. These observations have led
Aghajanian to suggest that LSD begins its sequence of
physiological and neurochemical changes by acting on 5-HT
neurons. In this view, the psychedelic actions of LSD must
entail a primary or perhaps total reliance upon decreased
efferent activity of the raphe neurons.

As Freedman has indicated, however, several points need

further clarification before the description of LSD-induced
changes in cell firing and cell chemistry can be incorporated in
an explanation of the pharmacological and behavioral effects
of LSD. If the effects of LSD could be simply equated with
silencing of the raphe and subsequent disinhibition of raphe
synaptic target cells, then many aspects of LSD-induced
behavioral changes should be detectable in raphe-lesioned
animals; but they are not. Typical LSD effects are also seen
when raphe-lesioned animals are administered LSD. When
normal animals are pretreated with p-chlorophenylalanine (in
dose schedules that antagonize raphe-induced synaptic
inhibition) and given LSD, the effect is accentuated. In
contrast to the predictions of the 5-HT silencing effect, the
LSD response is further accentuated by concomitant
treatment with 5-HTP. The physiological manipulation that
most closely simulates the behavioral actions of LSD is
stimulation of the raphe, leading to decreased habituation to
repetitive sensory stimuli. These data are difficult to reconcile
with the view that LSD-induced behavior results from
inactivation of tonic 5-HT-mediated postsynaptic actions. An
important aspect to be evaluated critically in the continued
search for the cellular changes that produce the behavioral
effects is the issue of tolerance: LSD and other indole
psychotomimetic drugs show tolerance and cross-tolerance in
humans, but these properties have not been seen in animal
studies at the cellular level.

It is extremely difficult to track down all of the individual

cellular actions of a potent drug like LSD and to fit these
effects together in a jigsaw puzzle-like effort to solve the
question of how LSD produces hallucinations. Similar jigsaw
puzzles lie just below the surface of every simple attempt to
attribute the effects of a drug or the execution of a complex
behavioral task like eating, sleeping, mating, and learning (no
rank ordering of author's priorities intended) to a single family
of neurotransmitters like 5-HT. While it is clearly possible to
formulate hypothetical schemes by which divergent inhibitory
systems like the 5-HT raphe cells can become an integral part
of such diverse behavioral operations as pain suppression,
sleeping, thermal regulation, and corticosteroid receptivity, a
very wide chasm of unacquired data separates the concept
from the documentary evidence needed to support it. The gap
is even wider than it seems since we do not at present have
the slightest idea of the kinds of method that can be used to
convert correlational data (raphe firing associated with
sleeping-stage onset or offset) into proof of cause and effect.
While previous editions of this guide to cellular
neuropharmacology have provided a superficial overview of
the behavioral implication of 5-HT neuronal actions, we now
relinquish that effort until the cellular bases become more
readily perceptible.

Actions of Other Psychotropic Drugs

Subsequent studies by Aghajanian and colleagues have
indicated that other drugs that can alter 5-HT metabolism can
also disturb the discharge pattern of the raphe neurons. Thus,
monoamine oxidase inhibitors and tryptophan, which would
be expected to elevate brain 5-HT levels, also slow raphe
discharge, while clinically effective tricyclic drugs
(antidepressants such as imipramine, chlorimipramine, and
amitriptyline) also slow the raphe neurons and elevate
synaptic 5-HT levels locally by inhibition of 5-HT reuptake. At
present, it remains unclear whether these metabolic changes
in 5-HT after psychoactive drugs are administered reflect
changes in the cells that receive 5-HT synapses or in the
raphe neurons themselves. A significant question is whether
these circuits mediate the therapeutic responses of the drugs
or are the primary pathological site of the diseases the drugs
treat. One etiological view of schizophrenia, for example,
might be based on the production of an abnormal indole, such
as the hallucinogenic dimethyltryptamine, a compound that
can be formed in human plasma from tryptamine. Some
evidence indicates that schizophrenic patients have
abnormally low monoamine oxidase activity in their platelets,
which could permit the formation of abnormal amounts of
plasma tryptamine and consequently abnormal amounts of
dimethyltryptamine. Figure 10-9 summarizes possible sites of
drug interaction in a hypothetical serotonin synapse in the
CNS.

In this chapter, we have encountered one of the more striking

examples of an intensively studied brain biogenic amine, and
there is every reason to believe that it is an important
synaptic transmitter. Still to be determined are the precise
synaptic connections at which this substance transmits
information and the functional role these connections play in
the overall operation of the brain with respect to both
effective and other multicellular interneuronal operations. The
central pharmacology of serotonin, however, remains poorly
resolved. Specific inhibition of uptake and synthesis are
possible, but truly effective and selective postsynaptic
antagonists remain to be developed.

Figure 10-9. Schematic model of a central

serotonergic neuron indicating possible sites of
drug action.

Site 1: Enzymatic synthesis: Tryptophan is taken

up into the serotonin-containing neuron and
converted to 5-OH-tryptophan by the enzyme
tryptophan hydroxylase. This enzyme can be
effectively inhibited by p-chlorophenylalanine and
-propyldopacetamide. The next synthetic step
involves the decarboxylation of 5-OH-tryptophan
to form serotonin (5-hydroxytryptamine [5-HT]).

Site 2: Storage: Reserpine and tetrabenazine

interfere with the uptake-storage mechanism of
the amine granules, causing a marked depletion of
serotonin.

Site 3: Release: At present, there is no drug

available which selectively blocks the release of
available which selectively blocks the release of
serotonin. However, lysergic acid diethylamide
(LSD), because of its ability to block or inhibit the
firing of serotonin neurons, causes a reduction in
the release of serotonin from the nerve terminals.

Site 4: Receptor interaction: LSD acts as a partial

agonist at serotonergic synapses in the central
nervous system (CNS). A number of compounds
have also been suggested to act as receptor-
blocking agents at serotonergic synapses, but
direct proof of these claims at the present time is
lacking.

Site 4a: Autoreceptor interaction: Autoreceptors

on the nerve terminal appear to play a role in the
modulation of serotonin release. 8-
Hydroxydiproplaminotetraline (8-OH-DPAT) acts as
an autoreceptor agonist at serotonergic synapses
in the CNS.

Site 5: Reuptake: Considerable evidence suggests

that serotonin may have its action terminated by
being taken up into presynaptic terminals. The
tricyclic drugs with a tertiary nitrogen, such as
imipramine and amitriptyline, appear to be potent
inhibitors of this uptake mechanism.

Site 6: Monoamine oxidase (MAO): Serotonin

present in a free state within the presynaptic
terminal can be degraded by the enzyme MAO,
which appears to be located in the outer
membrane of mitochondria. Iproniazid and
clorgyline are effective inhibitors of MAO. 5-HIAA,
5-hydroxyindoleacetic acid.
HISTAMINE

Introduction
Since the beginning of this century, when Henry Dale
demonstrated that histamine is an endogenous tissue
constituent and a potent stimulator of a variety of cells, this
substance has been thought to act as a transmitter or
neuromodulator. However, direct evidence in support of this
idea has accumulated relatively slowly.

The challenge posed by histamine to neuropharmacologists

has led to a vigorous chase across meadows of enticing
hypotheses surrounded by bogs of confusion and dubious
methodology. At last, more than 70 years after its isolation
from the pituitary by J. J. Abel, the role of histamine in the
brain seems to be approaching a resolution that this amine
occurs in two types of cells: mast cells and magnocellular
neurons in the posterior hypothalamus.

Most of our previous understanding of the synthesis and

degradation of histamine in the brain was based on attempts
to simulate in brain tissue data obtained from more or less
homogeneous samples of peritoneal mast cells (Fig. 10-10).
However, since mast cells were not supposed to be found in
healthy brains and since histamine does not cross the blood-
brain barrier, it had long been assumed that histamine could
also be formed by neurons and, hence, be considered as a
neurotransmitter. In fact, attempts to develop drugs that were
histamine antagonists, in the mistaken belief that battlefield
shock was compounded by histamine release, were a key to
the subsequent development of antipsychotic drugs.
Furthermore, as every hay-fever sufferer knows, present-day
antihistamine drugs clearly produce substantial CNS actions
such as drowsiness and hunger. Finally, in retrospect, the
increased content of histamine found in transected
degenerating sensory nerve trunks was likely an artifact of
mast cell accumulation rather than a peculiar form of
neurochemistry. Viewing the other data on CNS histamine
through that same retrospectroscope, as they have been
illuminated by the innovative experiments of Schwartz and
associates, we now have a rather compelling case that
histamine qualifies as a putative central neurotransmitter in
addition to its role in mast cells.

The major obstacles in elucidating the functions of histamine

in the brain have been the absence of a sensitive and specific
method to demonstrate putative histaminergic fibers in situ,
the lack of suitable methods to measure histamine and its
catabolites, and problems in the characterization of histamine
receptors in the nervous system. Some progress has been
made in overcoming these impediments, and evidence has
accumulated to support the hypothesis that histamine
functions as a neurotransmitter or neuromodulator in the
brain.

Figure 10-10. Metabolism of histamine. (1)

Histidine decarboxylase. (2) Histamine
methyltransferase. This is the major pathway for
inactivation in most mammalian species. (3)
Monoamine oxidase. (4) Diamine oxidase
(histaminase). (5) Minor pathway of histamine
catabolism.

Histamine Synthesis and Catabolism

Because histamine penetrates the blood-brain barrier so
poorly, brain histamine arises from histamine synthesis in situ
from histidine. Active transport of histidine by brain slices has
been demonstrated, and, because histidine loading has been
shown to elevate brain histamine, histidine transport could be
a controlling factor in brain histamine synthesis.
Two enzymes are capable of decarboxylating histidine in vitro:
L-AADC (i.e., DOPA-decarboxylase) and the specific histidine
decarboxylase. The pH optimum, affinity for histidine, effects
of selective inhibitors, and regional distribution of histamine-
synthesizing activity indicate that the specific histidine
decarboxylase is responsible for histamine biosynthesis in the
brain (Fig. 10-10). The instability of histidine decarboxylase
and its low activity in adult brain have precluded purification
of the enzyme from this source, but fetal liver histidine
decarboxylase has been purified to near homogeneity. Studies
of the pH optimum, cofactor (pyridoxal phosphate),
requirements, inhibitor sensitivity, and antigenic properties
have demonstrated that the brain and fetal enzymes have
similar properties. Antibodies to histidine decarboxylase have
been used to map the distribution of this enzyme in the brain
by immunohistochemical techniques.

Estimates of the Km of histidine for brain histidine

decarboxylase are much higher than the concentration of
histidine in plasma or brain, suggesting that it is not saturated
with substrate in vivo and consistent with the observations
that administration of L-histidine increases brain histamine
levels.

There are surprisingly few selective inhibitors of histidine

decarboxylase. Most of the effective inhibitors act at the
cofactor site, and although they reduce histamine formation,
they also inhibit other pyridoxal-requiring enzymes.

A selective inhibitor of histidine decarboxylase, -

fluoromethylhistidine, has been identified. This agent acts
selectively and irreversibly by formation of a covalent linkage
with the serine residue at the active site of the enzyme. It is
more selective and potent than other decarboxylase inhibitors
and does not inhibit DOPA and glutamate decarboxylase or
other histamine-metabolizing enzymes, such as histamine N-
methyltransferase and N-acetylhistamine deacetylase. This
agent may prove useful for manipulating in vivo actions of
histamine.

Histamine is metabolized by two distinct enzymatic systems in

mammals. It is oxidized by diamine oxidase to
imidazoleacetaldehyde and then to imidazoleacetic acid and
methylated by histamine methyltransferase to produce
methylhistamine. Mammalian brain lacks the ability to oxidize
histamine and nearly quantitatively methylates it.
Methylhistamine undergoes oxidative deamination by either
diamine oxidase or monoamine oxidase. The affinity is higher
for diamine oxidase, but in brains that lack diamine oxidase,
methylhistamine is oxidized primarily by monoamine oxidase
type B.

The major route of catabolism of orally injested histamine is

via N-acetylation of bacteria in the gastrointestinal tract to
form N-acetylhistamine.

Histamine-Containing Cells
Although there are fluorogenic condensation reactions that
can detect histamine in mast cells by cytochemistry, the
method has never been able to demonstrate histamine-
containing nerve fibers or cell bodies. Because the histamine
content of mast cells is quite substantial, however, it has been
possible to use the cytochemical method to detect mast cells
in brain and peripheral nerve. From such studies, Schwartz et
al. has estimated that the histamine content of brain regions
and nerve trunks that show about 50 ng histamine/g (i.e.,
every place except the diencephalon) could be explained on
the basis of mast cells. Moreover, the once inexplicable rapid
decline of brain histamine in early postnatal development can
also be attributed to the relative decline in the number of
mast cells in the brain during development.
Mast cell histamine shows some interesting differences from
what we shall presume is neuronal histamine. In mast cells,
histamine levels are high, turnover is relatively slow, and the
activity of histidine decarboxylase is relatively low; moreover,
mast cell histamine can be depleted by mast cell-
degranulating drugs (48/80 and polymyxin B). In brain, only
about 50% of the histamine content can be released with
these depletors and, for that which remains, the turnover is
quite rapid. The activity of histidine decarboxylase is also
much greater. Even better separation of the two types of
cellular histamine dynamics comes from studies of brain and
mast cell homogenates. In the adult brain, a significant
portion of histamine is found in the crude mitochondrial
fraction that is enriched in nerve endings and released from
these particles upon hypo-osmotic shock. In cortical brain
regions and in mast cells, most histamine is found in large
granules, which sediment with the crude nuclear fraction; this
histamine, unlike that in the hypothalamic nerve endings, has
the slow half-life characteristic of mast cells.

Despite the encouraging result that histamine may be present

in fractions of brain homogenates containing, among other
broken cellular elements, nerve endings, it is difficult to
parlay that information into a direct documentation of
intraneuronal storage. More promising steps in that direction
were taken in studies where specific brain lesions were made.
Lesions of the medial forebrain bundle region produced a loss
of forebrain NE and 5-HT along a time course parallel to nerve
fiber degeneration in experiments that were done when our
understanding of CNS monoaminergic systems was not as far
along as it is for histamine today. When such lesions were
placed unilaterally and specific histidine decarboxylase activity
followed, a 70% decline in forebrain histidine decarboxylase
activity was found at the end of 1 week. The decline was not
due to the concomitant loss of monoaminergic fibers since
lesions made by hypothalamic injections of 6-
hydroxydopamine or 5,7-dihydroxytryptamine did not result
in histidine decarboxylase loss. The most reasonable
explanation would be that the lesion interrupted a histamine-
containing diencephalic tract; the histamine and
decarboxylating activity on the ipsilateral cortex was reduced
only about 40%, suggesting that the pathway may make
diffuse contributions to nondiencephalic regions. However,
subsequent studies of completely isolated cerebral cortical
"islands" showed complete loss of histamine content. As
mentioned earlier, antibodies to histidine decarboxylase have
also been utilized to map the distribution of this enzyme in
rodent brain by immunohistochemical techniques. Steinbusch
and coworkers have developed an immunohistochemical
method for the visualization of histamine in the CNS using an
antibody raised against histamine coupled to a carrier protein.
Results obtained with this technique are in general agreement
with studies on the immunohistochemical localization of
histidine decarboxylase. The highest density of histamine-
positive fibers is found in brain regions known to contain high
histamine levels, such as the median eminence and the
premammillary regions of the hypothalamus.

Lesion and immunohistochemical studies indicate the

presence of a major histamine-containing pathway emanating
from neurons in the posterior basal hypothalamus and
reticular formation and ascending through the medial
forebrain bundle to project ipsilaterally to broad areas of the
telencephalon (Fig. 10-11). A descending pathway originating
from the hypothalamus and projecting to the brain stem and
spinal cord has also been suggested.

The majority of histamine-containing perikarya are comprised

of a continuous group of magnocellular neurons numbering
about 2000 in the rat and confined primarily to the tuberal
region of the posterior hypothalamus, collectively named the
tuberomammillary nucleus. A similar organization has also
been described in humans, although histamine-containing
neurons are more abundant ( 64,000) and occupy a greater
portion of the hypothalamus. Like other monoaminergic
neurons, the histaminergic system consists of long, highly
divergent, mostly unmyelinated, varicose fibers projecting in
a diffuse manner to many telencephalic, mesencephalic, and
cerebellar structures (Fig. 10-11). Individual neurons appear
to project to widely divergent areas but make few typical
synaptic contacts. Histamine neurons are unique in that they
are characterized by the presence of an unusually large
number of biological markers normally associated with other
neurotransmitter systems, including glutaminic acid
decarboxylase, adenosine deaminase, substance P, galanin,
met-enkephalin, and brain natriuretic peptide.
Tuberomammillary neurons also contain monoamine oxidase
B, the major enzyme responsible for deamination of N-
methylhistamine, a major metabolite of histamine in the CNS.
Unraveling the possible roles of such a large number of
putative cotransmitter peptides in a single population of
magnocellular neurons in the posterior hypothalamus (the
tuberomammillary nucleus) poses an exciting challenge.

Figure 10-11. Schematic illustration of the

distribution of histamine-containing neurons in
brain. M.R.F., mesencephalic reticular formation;
M.B., mammillary bodies; M.F.B., medial forebrain
bundle. (Modified from Schwartz and Arrang,
2002.)

Histamine Receptors
Three classes of histamine receptors have been identified in
vertebrates and named in the order of their discovery: H1,
H2, and H3. Both H1 and H2 receptor DNA have been cloned.
They belong to the superfamily of receptors with seven
transmembrane domains and are coupled to
guanylnucleotide-sensitive G proteins. The autoradiographic
distribution of the three histamine receptors has been
mapped in primate brain. H1 receptors are particularly
abundant in the neocortex, while H2 and H3 receptors,
although present in the cerebral cortex, are enriched in
caudate and putamen. H3 receptors are predominant in basal
ganglia, with the highest density found in the globus pallidus.
Histamine receptors are also found in the hippocampus and
cortical areas. Although the presence of H3 receptors in brain
has been appreciated since the mid-1980s, the molecular
identity of this receptor has remained elusive. Thus, this
receptor has been characterized in the CNS largely on the
basis of radioligand binding and pharmacological studies with
subtype-specific agents. In general, it is conjectured that the
histamine H3 receptor functions as an autoreceptor coupled to
Gi proteins, analogous to dopamine D2, adrenergic 2, and
serotonergic 5HT1A autoreceptors. H3 receptors also appear
to function as heteroreceptors regulating the release of other
neurotransmitters from their nerve terminals.

The cloning and functional expression of the human H3

receptor were finally achieved in 1999. The molecular
structure of the H3 receptor demonstrates that it, like the
other histamine receptors, belongs to the superfamily of G
protein-coupled receptors. With the availability of the cDNA
for the H3 receptor, it will now be possible to develop
chemical and biological reagents to address a number of
questions concerning the possible role of H3 receptors in the
CNS using molecular biology.

A new histamine receptor, H4, has been reported and

characterized, adding a new chapter to the histamine story.
This gene encodes a 390-amino acid Gi-coupled receptor with
40% homology to the H3 receptor and a similar intron-exon
structure. The localization of this receptor is quite restricted,
and it does not appear to occur in the CNS. Expression
appears localized to spleen, thymus, intestine, and
immunologically active cells such as neutrophils, eosinophils,
and T cells, suggesting a role in the regulation of immune
function. If confirmed, this could offer a novel therapeutic
potential for histamine receptor ligands in allergic and
inflammatory diseases.

H1 receptors are coupled to inositol phospholipid hydrolysis

and involved in a variety of responses, including contraction
of smooth muscle, increased capillary permeability mediated
by contraction of terminal venules, hormone release, and
brain glycogenolysis induced by a rise in the intracellular
concentration of free calcium ions. H2 receptors are positively
coupled to adenylyl cyclase via a Gs protein and mediate most
of their effects by changes in the intracellular levels of cAMP.
The functional responses mediated by H2 receptors include
smooth muscle relaxation, gastric acid secretion, positive
chronotropic and inotropic actions on cardiac muscle, and
inhibitory effects on the immune system.

A third class of histamine receptors, H3, was initially identified

as an autoreceptor controlling histamine synthesis and
release from histamine-containing neurons in rat cerebral
cortex. Subsequently, the H3 receptor was shown to inhibit
the presynaptic release of other monoamines and peptides
from brain and peripheral tissue. Functional studies have
provided evidence for presynaptic release modulating H3
receptors on noradrenergic, serotonergic, dopaminergic,
cholinergic, and peptidergic neurons. H3 receptors are
believed to provide a major regulatory mechanism involved in
the control of histaminergic activity under physiological
conditions. In vivo microdialysis studies have demonstrated
that administration of selective H3 receptor agonists reduces
histamine release and turnover. Administration of H3 receptor
antagonists, in contrast, enhances histamine turnover and
release, suggesting that H3 autoreceptors are normally under
tonic stimulation by endogenous histamine.

H1 Receptor Agonists and Antagonists

Because of its high affinity and good receptor selectivity,
mepyramine remains the agent of choice as a selective H1
antagonist. A number of H1 antagonists exist as optical
isomers, and in many instances the (+) stereoisomer exhibits
higher potency than the (-) stereoisomer. (+)-
Chlorpheniramine is such an example. One of the most potent
H1 antagonists is the geometric isomer transtriprolidine,
which has a Kd of 0.1 nM. In humans, a major side effect
associated with administration of H1 antagonists is a high
degree of sedation, which has been attributed to the ability of
the antagonists to occupy H1 receptors in the brain. Most of
the classic H1 antihistamines (e.g., diphenhydramine,
mepyramine) readily cross the blood-brain barrier and elicit a
significant degree of sedation. A number of new compounds,
however, penetrate poorly into the CNS and thus are
relatively nonsedating. Terfenadine (Fig. 10-12) is an example
of such an agent. Highly selective agonists do not exist for H1
receptors. 2-Methylhistamine and 2-thiazoylethylamine show
some selectivity, but their relative potency at H1 receptors
differs by no more than an order of magnitude from that
exhibited at H2 receptors.

Figure 10-12. H1 agonists and antagonists.

H2 Receptor Antagonists

Following the successful use of H2 receptor antagonists like

cimetidine in the treatment of peptic ulcers, a number of
these agents have become available. The structures of several
H2 antagonists are illustrated in Figure 10-13. Although a
number of more potent agents have been developed, only
ranitidine and tiotidine lack antagonist activity on H3
receptors. These agents have limited access to the CNS. The
only selective H2 receptor antagonist that effectively
penetrates into the CNS is zolantidine. This compound has
been used extensively in experimental animals but has not
been marketed for clinical use. Dimaprit is a fairly selective
H2 agonist that discriminates well between H2 and H1 or H3
receptors. Impromidine is one of the most potent H2 receptor
agonists available. Like dimaprit, it has minimal agonist
effects on H1 receptors, but it has more potent H3 receptor
antagonist properties. Amthamine is also a more potent
agonist at H2 receptors than diamaprit and does not affect H1
receptors. However, this agent exerts weak agonist activity at
the H3 receptor and thus does not exhibit optimal subtype
selectivity.

Figure 10-13. H2 agonists and antagonists.

H3 Receptor Antagonists and Agonists

H3 receptor antagonists and agonists are illustrated in Figure

10-14. Thioperamide is a potent H3 receptor antagonist that
exhibits good receptor specificity. This drug, unlike other H3
antagonists, crosses the blood-brain barrier and has been
useful in evaluating the behavioral role played by H3
receptors in the CNS. A new antagonist, clobenpropit, is the
most potent H3 receptor antagonist available for investigating
these receptors. N -Methylhistamine and N -N -
dimethylhistamine are both potent H3 receptor agonists, but
neither shows any marked discrimination between the three
histamine receptors. However, substitution of methyl groups
in the or position of the side chain of the histamine
molecule produces agents that exhibit a high degree of
selectivity for the H3 receptors. For example, R -
methylhistamine and R ,S -dimethylhistamine are 15-20
times more potent than histamine as H3 agonists but possess
only about 1% of the activity of histamine on H1 and H2
receptors. Imetit is a potent and selective H3 receptor full
agonist that is effective both in vivo and in vitro and 60 times
more potent than histamine. Table 10-5 summarizes the
pharmacology of histamine receptors and their effector
pathways.

In summary, there is compelling evidence that histamine, in

addition to its presence in mast cells, exists in neurons, where
it probably functions as a neurotransmitter in the CNS. The
brain has a similar nonuniform distribution of histamine, a
specific histidine decarboxylase, and methylhistamine (the
major metabolite of brain histamine). The cerebellum has the
lowest levels, whereas the hypothalamus is most enriched.
Hypothalamic synaptosomes contain histamine and histidine
decarboxylase, suggesting a localization in nerve endings.
Depolarization of hypothalamic slices causes a calcium-
dependent release of histamine, and tetrodotoxin sensitive
histamine release from neurons has been demonstrated in
vivo. The turnover of brain histamine is quite rapid, as with
other biogenic amine transmitters.
The actions of histamine in the CNS appear to be mediated by
three classes of receptors. These three types, H1, H2, and
H3, can be distinguished in the CNS by their pharmacology,
their localization, and the intracellular responses they
mediate. Neuropharmacological studies of brain histaminergic
systems have been hampered by the lack of appropriate
pharmacological agents. In the last few years, parenterally
active drugs, H2 antagonists, H3 agonists, and antagonists
have become available; and their effects on behavioral,
neuroendocrine, and vegetative functions are being studied in
animals.

Lesions of the midbrain or caudal hypothalamus cause a

progressive reduction in forebrain histamine and histidine
decarboxylase, consistent with the existence of an ascending
histamine-containing system, which has now been directly
visualized by immunohistochemical techniques. Continued
application of these immunohistochemical techniques for
visualization of histidine decarboxylase-like and histamine-like
activity in the brain should provide more detailed maps of
histamine-containing neuron systems in the brain in the near
future. However, we are still far from being able to speak with
conviction of specific histaminergic circuits and their relation
to behavior; thus, the functions of histamine in brain are
highly speculative.

The high levels of histamine and the presence of histamine

receptors in the hypothalamus, together with the known
ability of histamine to alter food and water intake
thermoregulation, autonomic activity, and hormone release,
indicate a possible role for histamine in these vegetative
functions. Also, because of the presence of ascending
histamine-containing fibers originating from the reticular
formation, coupled with the sedative nature of H1 antagonists
and the activation effects of histamine on some brain cells, a
possible role for histamine in arousal has been suggested.

Despite numerous speculations based largely on observations

of the response to locally applied histamine, only a few
physiological roles for histamine are well documented. These
include arousal, control of appetite, regulation of pituitary
hormone secretion, and control of vestibular reactivity. The
therapeutic role for histaminergic drugs is fairly limited, and
the primary pharmacological actions of these agents are often
associated with side effects ascribed to other therapeutic
agents. For example, several antipsychotic (e.g., clozapine
and thioridazine) and antidepressant (e.g., doxepin and
amitriptyline) drugs with sedative and weight-gain side effects
display potent H1 antagonist properties. Histamine
antagonists have been used therapeutically for only a limited
number of conditions. H1 antagonists like meclizine are the
most commonly used anti-motion-sickness drugs. However, it
is still uncertain whether H1 blockade is primarily responsible
for the effectiveness of these agents. Some H1 antagonists,
notably meclizine and dimenhydrinate, are of benefit in
vestibular disturbances such as Meniere's disease. H1
receptor antagonists are also the ingredients of a number of
proprietary remedies sold over the counter as sleeping pills
for the treatment of insomnia. These preparations are
generally ineffective at the recommended dosages, although
some sensitive individuals may derive benefit. H1 antagonists
have a well-established and valued place in the symptomatic
treatment of hypersensitivity reactions (allergic rhinitis and
conjunctivitis). However, their effect is purely palliative, being
confined to suppression of the symptoms attributed to the
antigen-antibody-induced release of histamine. Despite
popular belief, H1 antagonists are without value in the
treatment of the common cold.

We would be remiss not to mention the therapeutic usefulness

of the H2 antagonists. The main therapeutic use of these
agents is in the treatment of gastric and duodenal ulcers. H2
antagonists are very effective at lowering the basal and
nocturnal secretion of gastric acid and that which is
stimulated by meals. They both reduce the pain and the
consumption of antacids and hasten the healing of ulcers. The
incidence of side effects with these H2 antagonists (cimetidine
and ranitidine) is low and can probably be attributed, in part,
to the limited function of H2 receptors in organs other than
the stomach and to their poor penetration of the blood-brain
barrier.

Figure 10-14. H3 agonists and antagonists.

ADENOSINE
Considerable interest has developed over the past two
decades in the possible role of adenosine and adenine
nucleotides, such as adenosine triphosphate (ATP), as
transmitters or neuromodulators in the CNS. While it was
known more than 70 years ago that adenosine and ATP
elicited potent pharmacological effects on the nervous system
(Drury and Szent-Gyorgy, 1929), evidence that endogenous
purines regulated the activity of the nervous system was not
available until recently. The seminal work of Rall, Sattin,
McIlwain, and others demonstrated that adenosine stimulated
the formation of cAMP in brain slices and that
methylxanthines such as a theophylline and caffeine were
competitive pharmacological antagonists of this response.
These findings were instrumental in developing much of the
interest in adenosine in the mid-1980s. Adenosine and
adenosine analogues are potent activators of cellular function,
altering neuronal activity and affecting behavior as well.
Highly specific, high-affinity binding sites for adenosine have
also been identified in the CNS, and it is thought that most, if
not all, of the physiological actions of adenosine are mediated
via the interaction of extracellular adenosine with these G
protein-coupled receptor sites. Nevertheless, despite the fact
that adenosine is a potent neuroactive substance, it meets
few of the criteria considered essential to identify an
endogenous substance as a candidate for a neurotransmitter.
Adenosine is neither stored nor released as a classical
neurotransmitter. In contrast to traditional neurotransmitters,
it has poorly defined pathways leading to its synthesis and
degradation and appears to act more as a neuromodulator to
fine-tune synaptic function. Adenosine is released by most
cells, including neurons and glial cells, and modulates the
activity of the nervous system by acting presynaptically
(facilitating or inhibiting transmitter release), postsynaptically,
and/or nonsynaptically. To date, four adenosine receptor
subtypes have been cloned, A1, A2A, A2B, and A3, each of
which exhibits a unique tissue distribution, ligand affinity, and
signal-transduction mechanism. Adenosine mediates it
physiological effects through activation of high-affinity
receptors (A1 and A2A; see Chapter 4). Activation of lower-
affinity receptors (A2B and A3) is believed to be involved in
mediating pathological actions since these subtypes become
activated only when adenosine levels increase into the
micromolar range during periods of hypoxia, ischemia, or
inflammation. Thus, the pathophysiological role of these
adenosine receptor subtypes may be very different from that
of A1 and A2 in that they may act as endogenous regulators
largely under conditions of more severe challenge.

In addition to exerting an effect on nerve cells directly,

adenosine receptor activation influences the action of other
neurotransmitters and neuromodulators indirectly. A
convincing case has been made for adenosine receptors
having a critical function in regulating the activation of
multiple receptors that affect neurotransmitter release and
synaptic function. In particular, they regulate N-methyl-D-
aspartate and metabotropic glutamate receptors, nicotinic
receptors, and peptidergic receptors such as calcitonin gene-
related peptide and vasoactive intestinal polypeptide
receptors. Thus, adenosine has been speculated to behave as
a modulator to fine-tune the actions of other transmitters and
neuromodulators.

SELECTED REFERENCES

Serotonin
Aghajanian, G. K. and E. Sanders-Bush (2002). Serotonin. In
Neuropsychopharmacology The Fifth Generation of Progress
(K. L. Davis, D. Charney, J. T. Cole and C. Nemeroff, eds)
Lippincott Williams and Wilkins, Philadelphia, PA, pp. 15-34.

Barker, E. L. and R. D. Blakely (1995). Norepinephrine and

serotonin transporters: molecular targets of antidepressant
drugs. In Psychopharmacology, Fourth Generation of Progress
(F. E. Bloom and D. J. Kupfer, eds.). Raven Press, New York,
pp. 321-333.

Barnes, N. M. and T. Sharp (1999). A review of central 5-HT

receptors and their function. Neuropharmacology 38, 1083-
1152.

Blakely, R. D., S. Ramamoorthy, S. Schroeter, Y. Quian, S.

Apparsondarum, A. Galliand and L. DeFelice (1998).
Regulated phosphorylation and trafficking of antidepressant-
sensitive serotonin transporter proteins. Biol. Psychiatry 44,
169-178.

Blier, P. and C. deMontigny (1994). Current advances and

trends in the treatment of depression. Trends Pharmacol. Sci.
15, 220-226.

Branchek, T. A. and T. P. Blackburn (2000). 5-HT6 receptors

as emerging targets for drug discovery. Annu. Rev.
Pharmacol. Toxicol. 40, 319-334.

Fitzpatrick, P. F. (2000). The aromatic amino acid

hydroxylases. Adv. Enzymol. 74, 235-294.

Glennon, R. A. and M. Dukat (1995). Serotonin receptor

subtypes. In Psychopharmacology, Fourth Generation of
Progress (F. E. Bloom and D. J. Kupfer, eds.), Raven Press,
New York, pp. 415-430.

Heninger, G. R. (1997). Serotonin, sex and psychiatric illness.

Proc. Natl. Acad. Sci. U.S.A. 94, 4823-4824.

Hoyer, D., D. E. Clarke, J. R. Fozard, P. R. Hartig, G. R. Martin,

E. J. Mylecharane, P. R. Saxena, and P.P.A. Humphrey (1994).
International Union of Pharmacology classification of
receptors for 5-hydroxytryptamine (serotonin). Pharmacol.
Rev. 46, 157-203.

Lovenberg, T. W., B. M. Baron, L. deLecea, J. D. Miler, R. A.

Porsser, M. A. Rea, P. E. Foye, M. Racke, A. L. Slone, B. W.
Siegel, P. E. Danielson, J. G. Sutcliffe, and M. G. Erlander
(1993). A novel adenylyl cyclase-activating serotonin receptor
(5-HT7) implicated in the regulation of mammalian circadian
rhythms. Neuron 11, 449-458.

Mamounas, L. A., C. A. Mullen, E. O'Hearn, and M. E. Molliver

(1992). Dual serotonergic projection to forebrain in the rat:
morphologically distinct 5HT axon terminals exhibit
differential vulnerability to neurotoxic amphetamine
derivatives. J. Comp. Neurol. 314, 558-586.
Maricq, A. V., A. S. Peterson, A. J. Brake, R. M. Myers, and D.
Julius (1991). Primary structure and functional expression of
the 5HT3 receptor, a serotonin-gated ion channel. Science
254, 432-437.

Morales, M., E. Battenberg, and F. E. Bloom (1998).

Distribution of neurons expressing immunoreactivity for the
5HT3 receptor subtype in the rat brain and spinal cord. The
Journal of Comparative Neurology 385, 385-401.

Murphy, D. L., C. Wichems, Q. Li, and A. Heils (1999).

Molecular manipulations as tools for enhancing our
understanding of 5-HT neurotransmission. Trends Pharmacol.
Sci. 20, 246-252.

Owens, M. J. and C. B. Nemeroff (1994). The role of serotonin

in the pathophysiology of depression: focus on the serotonin
transporter. Clin. Chem. 40, 288-295.

Peroutka, S. J. (1994). Molecular biology of serotonin (5-HT)

receptors. Synapse 18, 241-260.

Ramamoorthy, S. and R. Blakely (1999). Phosphorylation and

sequestration of serotonin transporters differentially
modulated by psychostimulants. Science, 285, 763-766.

Ricaurte, G. A., J. Yuan, and U. D. McCann (2000). ( )3,4-

Methylenediozymethamphetamine ("ecstasy")-induced
serotonin neurotoxicity: studies in animals.
Neuropsychobiology 42, 5-10.

Sanders-Bush, E. and H. Canton (1995). Serotonin receptors:

signal transduction pathways. In Psychopharmacology, Fourth
Generation of Progress (F. E. Bloom and D. J. Kupfer, eds.).
Raven Press, New York, pp. 431-442.
Saudou, F. and R. Hen (1994). 5-Hydroxytryptamine receptor
subtypes in vertebrates and invertebrates. Neurochem. Int.
25, 503-532.

Steinbusch, H. W. M. and A. H. Mulder (1984). Serotonin-

immunoreactive neurons and their projections in the CNS. In
Handbook of Chemical Neuroanatomy. Classical Transmitters
and Transmitter Receptors in the CNS, vol. 33, part II (A.
Bjorklund, T. Hokfelt, and M. Kuhar, eds.), Elsevier, New York,
pp. 68-118.

Histamine
Hill, S. J. (1990). Distribution, properties, and functional
characteristics of three classes of histamine receptor.
Pharmacol. Rev. 42, 45-83.

Honrubia, M. A., M. T. Vilaro, J. M. Palacios, and G. Mengod

(2000). Distribution of the histamine H2 receptor in monkey
brain and its mRNA localization in monkey and human brain.
Synapse, 38, 343-354.

Lovenberg, T. W., B. L. Roland, S. J. Wilson, X. Jiang, J. Pyati,

A. Huvar, M. R. Jackson and M. G. Erlander (1999). Cloning
and functional expression of the human histamine H3
receptor. Molecular Pharmacology, 55, 1101-1107.

Martinez-Mir, M. I., H. Pollard, J. Moreau, J. M. Arrang, M.

Ruat, E. Traiffort, J. C. Schwartz, and J. M. Palacios (1990).
Three histamines receptors (H1, H2 and H3) visualized in the
brain of human and non-human primates. Brain Research
526, 322-327.

Onodera, K., A. Yamatodani, T. Watanabe, and H. Wadas

(1994). Neuropharmacology of the histaminergic neuron
system in the brain and its relationship with behavioral
disorders. Prog. Neurobiol. 42, 685-702.

Panula, P. and M. S. Airaksinen (1991). The histaminergic

neuronal system as revealed with antisera against histamine.
In Histaminergic Neurons: Morphology and Function (T.
Watanabe and H. Wada, eds.). Boca Raton, FL, CRC Press, pp.
127-144.

Schwartz, J. C. and J. M. Arrang (2002). Histamine. In

Neuropsychopharmacology The Fifth Generation of Progress
(K. L. Davis, D. Charney, J. T. Cole and C. Nemeroff, eds.)
Lippincott Williams Wilkiins, Philadelphia, PA, pp. 179-190.

Schwartz, J. C. and H. L. Haas, eds. (1992). The Histamine

Receptor. New York, Wiley-Liss.

Staines, W. A. and J. I. Nagy (1991). Neurotransmitter

coexistence in the tuberomammillary nucleus. In
Histaminergic Neurons: Morphology and Function (T.
Watanabe and H. Wada, eds.). Boca Raton, FL, CRC Press, pp.
163-176.

Steinbusch, H. W. M. and A. H. Mulder (1984).

Immunohistochemical localization of histamine in neurons and
mast cells in brain. In Handbook of Chemical Neuroanatomy.
Classical Transmitters and Transmitter Receptors in the CNS,
vol. 3, part II. (A. Bjorklund, T. Hokfelt, and M. J. Kuhar,
eds.). Elsevier, New York, pp. 126-140.

Tohyama, M., R. Tamiya, N. Inagaki, and H. Takagi (1991).

Morphology of histaminergic neurons with histidine
decarboxylase as a marker. In Histaminergic Neurons:
Morphology and Function (T. Watanabe and H. Wada, eds.).
Boca Raton, FL, CRC Press, pp. 107-126.

Vizuete, M. L., E. Traiffort, M. L. Bouthenet, M. Ruat, E. Souil,

J. Tardivel-Lacombe, and J.-C. Schwartz (1997). Detailed
mapping of the histamine H2 receptor and its gene transcripts
in guinea-pig brain. Neuroscience 80(2), 321-343.

Wouterlood, F. G. and H.W.M. Steinbusch (1991). Afferent and

efferent fiber connections of histaminergic neurons in the rat
brain: comparison with dopaminergic, noradrenergic and
serotonergic systems. In Histaminergic Neurons: Morphology
and Function (T. Watanabe and H. Wada, eds.). Boca Raton,
FL, CRC Press, pp. 145-162.

Adenosine
Bischofberger, N., K. A. Jacobson and D. K. J. E. von Lubitz
(1997). Adenosine A1 receptor agonists as clinically viable
agents for treatment of ischemic brain disorders. Ann. N.Y.
Acad. Sci. 825, 23-29.

Krugel, U., H. Kittner, and P. Illes (2000). Mechanisms of

adenosine 5'-triphosphate-induced dopamine release in the
rat nucleus accumbens in vivo. Synapse 39, 222-232.

Nyce, J. W. (1999). Insight into adenosine receptor function

using antisense and gene knockout approaches. Trends
Pharmacol. Sci. 20, 79-83.

Ongini, E., M. Adami, C. Ferri, and R. Bertorelli (1997).

Adenosine A2A receptors and neuroprotection. Ann. N.Y.
Acad. Sci. 825, 30-48.

Williams, M. and M. F. Jarvis (2000). Purinergic and

pyrimidinergic receptors as potential drug targets. Biochem.
Pharmacol. 59, 1173-1185.

Zimmermann, H. (1994). Signalling via ATP in the nervous

system. Trends Neurosci. 10, 420-426.
 (+/-) Show / Hide Bibliography

Send Feedback Teton Server (5.0.0) - ©2007 Teton Data Title Updates
Customer Service Systems User
800.901.5494 Send Us Your Comments Responsibilities
Training Center
What's New
11. Neuroactive Peptides
INTRODUCTION
After more than 30 years of ever incisive examination,
neuropeptides remain dear to the hearts of
neuropharmacologists. For those hoping eventually to
understand the role of neuropeptides, the quest to devise new
medications to enhance or antagonize their actions remains
intense. Rapid advances in molecular biology and the
regulation of gene expression have accelerated the
determination of peptide amino acid sequences. In addition,
the characterization of their receptors by cloning and
transfection studies (see Chapter 3) and the successful
synthesis of both peptidic and nonpeptidic antagonists for
these receptors fuel this attraction. New peptides continue to
be discovered, often through attempts to characterize the as
yet undiscovered natural ligands for cloned but orphaned
receptors. The new student entering this field might assume
that the peptide story must be nearly all told. Alas, that
conclusion would be erroneous. As in most of the fields
discussed in this book, better data beget better and more
complex questions.

Almost all peptides made by neurons will affect specific target

cells of the central nervous system (CNS) and peripheral
nervous system. neuronal, glial, smooth muscle, glandular,
and vascular. The census of peptides already exceeds several
dozen, with no obvious end in sight. However, the actual
molar concentration of any given peptide in the brain is
maximally two to three orders of magnitude lower than that
of the monoamines, acetylcholine (ACh), and amino acids;
and their receptors respond at correspondingly lower
concentrations. The specific signaling advantages of peptides
are not yet obvious. Furthermore, with very few exceptions,
no peptides have been linked selectively to a given functional
system in the brain or correlated in any exclusive manner
with any pathological states. Thus, the challenge of devising
new drugs for neurological or psychiatric disorders based on
these peptides is considerably more difficult than just
designing potent congeners that can tweak their receptors on
or off. It requires that we have some reason to do that.
Nevertheless, there are strong implications that the biological
properties evoked by peptide-mediated signals will influence
future pharmacological development.

SOME BASIC QUESTIONS

What Is Different about Peptides?

Peptide-secreting neurons differ in their biology from neurons
using amino acids and monoamines in ways other than the
molecular structure of their transmitter. Amino acid or
monoamine transmitters are formed from diet-derived amino
acids by one or two intracellular enzymatic steps. The end
product of these enzymatic actions is the active transmitter
molecule, which is then stored in the nerve terminal until
release. After release, the transmitter (or choline in the case
of ACh) can be reaccumulated back into the nerve terminal by
the energy-dependent, active reuptake property, thus
conserving the requirement for de novo synthesis. Peptide-
secreting cells employ a somewhat more formidable
approach: synthesis directed by mRNA can take place only on
ribosomes and, thus, only in the perikaryon or dendrites of a
neuron. Furthermore, all known neuropeptides arise from
larger prohormone forms (inactive precursors). The
biologically active, secreted form is cleaved from these
prohormones by the actions of selective proteolytic enzymes.

Thus, the process for peptides starts with ribosomal synthesis

of the prohormone, which is then packaged into vesicles in
the smooth endoplasmic reticulum and transported from the
perikaryon to the nerve terminals for eventual release. These
steps in the molecular processing of secretory products are
described in detail in Chapter 3.

Insofar as the present data reveal, peptide-releasing neurons

share certain basic properties with all other chemically
characterized neurons. Transmitter release is Ca2+-
dependent. Postsynaptically, all fully characterized receptors
for neuropeptides are G protein-coupled, as described in
Chapter 4. The effects of neuropeptides thus are mediated
indirectly by regulating ion channel properties through second
messengers such as Ca2+, cyclic nucleotides, or inositol
triphosphates (see Chapter 5). Furthermore, like the
monoamines, peptides identical to those extracted from the
nervous system are often made by nonneural secretory cells
in endocrine glands and the cells lining hollow viscera like the
intestinal tract. Some observers of the peptide scene
emphasize characteristics of neuropeptide action that are
extrapolations from the actions of endocrine peptide
hormones. Thus, they point to targets that are distant from
release sites, release into portal or other circulatory systems,
are of higher potency, and have a longer time course of
effects. Although endocrine hormones (peptides or amines)
can produce effects in the nanomolar range or below, we do
not yet know the time course or thresholds for the same
neuropeptides to act on their target cells in the brain or
ganglia. In amphibian autonomic ganglia, at least one peptide
does act in such a local endocrine, or paracrine, fashion,
producing synaptic potentials that last for hundreds of
milliseconds. These effects are observed with an endogenous
peptide that is very similar to hypothalamic luteinizing
hormone-releasing hormone and produced on a nonsynaptic
neuronal target by diffusing at least 10-100 um from
preganglionic synaptic release sites, without apparently acting
at all on the immediately adjacent postsynaptic neuron.
Whether these particular paracrine properties are ways
around the simple amphibian autonomic anatomy or a
prototype for all central and peripheral peptide actions
remains debatable.

How Are New Peptides Found?

Before considering the interesting relationships among the
many known peptides, we can anticipate future events by
examining the recurrent pattern of events that seems to
underlie every wave of peptide discovery and
characterization. From this historical overview, the student
can see that new peptides do not just happen into existence.
The primary necessity for the discovery of a new peptide has
classically been the recognition of a biological function
controlled by an unknown factor. This biological action is then
employed as the bioassay for the subsequent steps of the
discovery process. This approach was used extensively for the
characterization of gastrointestinal peptide hormones in the
1940s and 1950s. However, in their pursuit of the
hypothalamic hormones that regulate secretion from the
anterior pituitary, the process was perfected and greatly
increased in power by Roger Guillemin and Andrew Schally
and their teams during the 1960s and 1970s. Typically, a
bioassay system was used to detect enrichment of an active
factor when extracts of a tissue thought to make the factor
were put through a series of chemical manipulations (i.e.,
differential extractions, molecular sieves, etc.). For some
factors, the purification of nanograms of active peptide
required hundreds of thousands of brains. The peptidic nature
of the purified material was inferred by the loss of biological
activity after treatment with peptidases (carboxypeptidases or
endopeptidases). The cleaved peptide fragments were then
sequenced, and the entire factor was eventually
reconstructed. With modern technology, many of those
historical steps can be rapidly set aside.

When a sequence has been achieved and matched against the

amino acid composition of the pure factor, it is necessary to
synthesize the peptide and determine if the synthetic
replicate matches the actions and potencies of the purified
natural factor. Only when the replicate's properties match
those of the factor can the natural peptide be declared
"identified." At this point, the factor is given a name that
reflects its actions in the original bioassay.

However, this is far from the end of the discovery process.

When the chemical structure of a peptide has been identified
and synthetically replicated, two new opportunities arise. The
synthetic material can be tested for physiological and
pharmacological properties. At the same time, the synthetic
material can be used to generate antibodies. The antibodies
are then employed for quantitative measurements (i.e.,
radioimmunoassay) and to determine the qualitative
distribution of the peptide by immunocytochemistry. Those
data then frequently reveal that the peptide has a much
broader distribution and, therefore, ostensibly, a much more
general series of actions than was originally presumed.
Autoradiographic mapping of potential response sites (first by
ligand binding for peptides and more recently by in situ
hybridization for the mRNAs of their receptors) is generally
the last step in the molecular and cellular characterization.
Curiously, these receptor maps reveal a flagrant disregard for
the cellular maps of peptides themselves. This mismatch
strengthens in some minds the concept of paracrine diffusion
for peptide actions at a distance. Empowered by the assays
and knowledge of the chemical neuroanatomy of the peptide-
containing circuitry, the peptide or an analog of it is
subsequently proposed to be the cause or cure of a major
mental illness. In many of the more recent reports, the
peptide isolated and found to be active is, in fact, not the only
active form contained in the tissue. In many cases, better
extraction methods confirmed by radioimmunoassay reveal
larger molecules with full or even greater potency. Although
the general rule is that active peptides are synthesized from
larger precursor hormones, when prohormone processing
gives rise to active peptides larger than the form first found,
the larger forms cannot really be viewed as "precursors" since
they have potent activity in their own right.

As more and more peptides have been identified within the

nervous system and from other sources, their internal amino
acid sequence similarities allow us to consider them as
members of one or another family grouping of peptides.
Family groupings are useful because they indicate that certain
sequences of amino acids have had considerable evolutionary
conservation, presumably because they provide unambiguous
signals between secreting and responding cells.

Two aspects of peptide family groupings merit appreciation.

(1) Some peptide families employ separate propeptide genes
or mRNAs yet share one or more short peptide sequences
that are expressed by different cell groups in the same
species. (2) In other peptide families, the peptides expressed
by homologous cell groups share generally identical peptide
sequences, with minor to moderate variations. The first
category, exemplified by the opioid peptide and the tachykinin
peptide families, implies that the natural ligand's receptors
may diverge into subtypes that can discriminate the fine
differences in peptide sequences. Although some individual
peptides have already earned their own receptor subtypes
(vasopressin, somatostatin), in other cases, such as the
tachykinins and the opioid peptides, the multiple receptor
subtypes predicted the later definition of multiple natural
ligands. In cases in which the propeptide form gives rise to
multiple different agonists, it is possible, but by no means yet
broadly supported, that more than one peptide messenger is
released from the possible array. Some dreamy-eyed pundits
have theorized that different branches of a neuron might have
the option of changing the mix of cleavage products available
for release, perhaps under steroid-driven conditions.
The latest additions to the list of neuroactive peptides have
been discovered by several strategies. In some cases, the
discovery hinged on specific common general features, such
as a C-terminally amidated sequence (seen in the
secretin/vasoactive intestinal peptide [VIP] and pancreatic
polypeptide families among others). In other cases, discovery
was based on the detection of potential cleavage fragments
when the propeptide sequence was deduced by molecular
cloning of the mRNA or gene (e.g., the -melanocyte-
stimulating hormone [MSH] found within proopiomelanocortin
mRNA or calcitonin gene-related peptide). Most recently,
cloning of the abundant G protein-coupled receptors from the
brain has identified new peptides that can activate the once-
orphaned receptors (e.g., orphanin FQ). By refinements of
differential display hybridization and searching for region-
specific brain peptides by mRNA detection, still other peptides
have been identified (cortistatin and hypocretin) Although
these historical discoveries are individually entertaining
stories, what the peptides do and how to make useful drugs
based on these actions remain elusive.

Transmitters or Not?
In Chapter 2, we considered the kinds of evidence needed to
demonstrate that a substance found in nerve cells is the
transmitter that those neurons secrete at their synapses (in
brief: [1] synthetic capacity, [2] presence, [3] activity-
dependent release, and [4] mimicry of the effects of the
presynaptic neuron). One other criterion is especially
pertinent to the objectives of this book. Drugs that can
simulate or antagonize the effects of presynaptic neuronal
activation must have an identical influence on the effects of
the substance applied exogenously to the target cell.

In the case of peptides, the opportunities for supplying these

data would at first glance seem better than they are with the
many categories of small molecules discussed in previous
chapters. Once the peptide has been isolated and its structure
determined, immunocytochemical localization of the peptide
usually follows promptly, as does the demonstration that, with
the right sensitivity assay, the peptide can be released from
brain slices by a Ca2+-dependent, depolarization process. The
development of a suitable synthetic ligand or receptor mRNA
in situ probe helps define receptor distribution, possible
actions on presumptive synaptic targets, their mechanisms of
regulation, and the discrimination of receptor subtypes. In
several peptide families, synthetic agonists and antagonists
have documented behavior-altering actions of the synthetic
peptide, establishing, by inference, a potential role of the
endogenous form in behavior and possibly in disease states.
Recent advances here include short-term inactivation of
peptide effects by intracerebral injection of antisense
oligonucleotides and longer-term regulation of peptide
synthesis in transgenic animals by overexpression or gene
knockouts of specific peptides or their receptors. These
molecular tools have accelerated the early partial fulfillment
of transmitter criteria for peptides generally. However, despite
intense efforts, there are very few peptide-containing circuits
within the CNS for which the complete array of transmitter
criteria have been satisfied. Perhaps for the present, it would
serve our knowledge best to refer to the role played by
neuropeptides as being a modulatory one (see Chapter 5).

They Are Not Alone

A once-predominant principle in neurotransmitter lore held
that a given neuron operated by one and only one transmitter.
According to this concept (attributed to Henry Dale, one of
the pioneer transmitter discoverers), one aspect of a neuron's
phenotype, as classic as its size, shape, location, circuitry,
and synaptic function, was its neurochemical designation as
"GABAergic," "cholinergic," "adrenergic," or whatever. All
card-carrying pharmacologists knew that autonomic neurons
came in only two color-coded categories: adrenergic and
cholinergic. Then, a curious finding began to repeat itself:
neuroactive peptides started showing up in autonomic
neurons, where there was no need for additional transmitters.
Initially viewed by many as an oddity, the idea of coexisting
peptides in central as well as peripheral neurons that are
already occupied by an amino acid or monoamine or even
another neuropeptide has now gained wide acceptance (see
Table 11-1). Actually, what Dale formulated as a principle was
that a given neuron would release the same neurotransmitter
from all of its many synapses. There being at the time very
few other viable chemicals to consider as transmitters, this
was scarcely heretical. However, Dale's principle was
subsequently restated by Eccles, to hold that a given neuron
used only one transmitter, which is almost, but not quite, the
same thing and established the straw man that the
neuropeptides blew away.

The notion of coexistence of other transmitters with peptides

leads one to ask whether there is any such thing as a truly
"peptidergic" neuron or, rather, only neurons in which a
peptide or two is there to expand the armamentarium of
messenger molecules available. Observations on cultured cells
transfected to express peptide and amine receptors suggest
that, in some cases, G protein-coupled heteromeric forms
could allow for dual convergent effects.

Such a revolutionary concept then forces us to ask a simple

question. If the peptide is not the sole signaling molecule but,
rather, a minor, fractional percent of the messenger-signaling
capacity of the nerve terminal, how might peptides affect
signals transmitted by the coexisting amino acid or
monoamine? Except for some recent work with opioid peptide
circuits in the hippocampus (see Opioid Peptides below), "We
don't know" remains the easy answer. However, in the
autonomic nervous system, where Tomas Hokfelt and his
peptide liberation team have found at least one peptide in
every sympathetic, parasympathetic, or enteric neuron they
have studied, some solid peptide-monoamine interactions are
emerging. In the salivary gland, low-frequency stimulation of
the parasympathetic nerves releases ACh, but VIP is also
released when the cholinergic nerves are stimulated at high
frequency. VIP augments parasympathetic control of
salivation by increasing glandular blood flow. Neuropeptide Y
(NPY), found in many sympathetic neurons, is also released at
higher frequencies of activation and sensitizes smooth muscle
target cells to their adrenergic signals. Other peptide actions,
although not necessarily between coexisting amino acid or
monoamine transmitters, are worth noting. For example, VIP
greatly accentuates the cAMP response of cortical neurons to
low doses of norepinephrine, and galanin can change the
release of ACh in the ventral hippocampus.

One way to recast these players in the proper perspective

might be to consider that a peptide usually embellishes what
the primary transmitter for a neuronal connection seeks to
accomplish. Such an effect may be to strengthen or prolong
the primary transmitter's actions, especially when the nerve is
called on to fire at higher than normal frequencies. The
peptide may provide part of the intraneuronal signal to alter
the rate of production of the primary transmitter. There may
still be places in which the peptide itself can fulfill all of the
effects of a primary transmitter. Of course, accepting
coexistences leads to new logistical problems for peptide-
using neurons. How do such neurons coordinate synthesis of
the peptide with their small-molecule transmitters? (Recall
that the latter can be made within the synapse and are often
reaccumulated after release but that peptides can be made
only with ribosomes and not in nerve terminals.) How do such
neurons decide when the secreted transmission should
include the peptide?

Other important concepts for pharmacology also emerge: for

example, the effects of peptide-directed drugs will depend not
only on the location of the receptors and their actions but also
on the context of its coexisting signals. Thus, the
pharmacology of NPY or neurotensin may be most
appropriately understood as a part of the total picture of
central noradrenergic pharmacology, given the degree to
which these two peptides may participate in that
monoaminergic neuron's transmission. Given the relatively
modest number of response mechanisms on which peptide
messengers may operate through G protein-coupled
receptors, a great enrichment of signaling possibilities
becomes attainable through the interplay between frequency-
dependent and diffusion-dependent release and response
sites.

In terms of the development of drugs active on the nervous

system, research on peptides has given us a humbling view of
the riches that may await mining in the inner depths of the
brain. The rest of this chapter reviews some emerging trends
in neuropeptide research, followed by pharmacologically
relevant data for the grand peptide families, a few individual
noteworthy peptides, and a newly recognized large family of
neuroactive and "glia-active" peptides, the cytokines, which
may figure prominently in pharmacological attacks on
neurodegenerative disorders.

EMERGING TRENDS IN NEUROPEPTIDE

RESEARCH
Observers of the active peptide parade may find it useful to
monitor certain recurring themes in several of the brain and
autonomic peptide systems. They illustrate potential
principles that could apply to peptide-signaling systems
generally.

1. Peptides are dynamically regulated. With current

technology, the dynamic synthesis and catabolism of
neuropeptides has been estimated by measuring
simultaneously mRNA, the precursor propeptide, and the
signaling form. These studies have helped clarify the effects
on peptide synthesis of drug treatments or other
experimental perturbations that alter the levels of the
coexisting amines in a given class of neurons. Furthermore,
examination of the genes for some neuropeptides has
revealed regulatory elements that make gene transcription
sensitive to gonadal hormones (oxytocin, vasopressin, and
proenkephalin) or to intracellular second messengers (VIP,
proenkephalin, neurotensin, corticotropin-releasing hormone,
and prosomatostatin), providing direct proof of independence
from their cotransmitters.

2. Postrelease cleavage products may provide different

signals. Based on the absence of any specific data, it had
been assumed that, despite the enormous logistical prelude to
constructing a propeptide and then cleaving and storing the
active form, its secretion was the end of the line. In several
peptide systems, however, pharmacological evidence suggests
that catabolic forms of the secreted peptide (i.e., fragments of
the secreted form) may have real signal value, both on the
neuron that secreted the full peptide and on the postsynaptic
neuron. The data suggest that such shorter forms may
participate in fine-tuning the levels of interaction between the
peptides and the effects of other coexisting transmitters.

3. Nonpeptidic agonist and antagonist drugs have arrived. The

third emerging trend is less an example of peptide physiology
than an expression of appreciation for the achievements of
medicinal chemists in synthesizing rigid nonpeptidic molecules
that can act as agonists and (experimentally more important)
as antagonists of the peptide's receptors. While it was clear
that nature could do this in the form of the basic morphine
molecule by making an agonist for opioid peptide receptors,
extension of this principle to other peptides was slow.
However, new molecules are coming forth, and as nonpeptidic
antagonists are created for peptide systems, the ability to
determine their functional roles will almost certainly be
improved.

THE GRAND PEPTIDE FAMILIES

Vasopressin and Oxytocin

These two highly similar nonapeptides (see Fig. 11-1) with
internal 1,6-disulfide bridges are the original mammalian
peptide "neurohormones." They are synthesized in the
neuronal perikarya, found in the large neurons of the
supraoptic and paraventricular nuclei, and stored in the axons
of these neurons in the neurohypophysis, from which they are
released into the bloodstream. Each peptide is synthesized as
part of a larger propeptide (see Chapter 3), with which it is
stored and released and from which it is cleaved as part of
the release process. In the kidney, vasopressin (also known
as antidiuretic hormone) facilitates distal tubular water
reabsorption, while oxytocin stimulates epididymal and
uterine muscle contraction.

The oxytocin and vasopressin peptides are expressed within

the classic magnocellular neurons of the paraventricular and
supraoptic nuclei, whose axons form the neurohypophysis,
where the peptides are secreted into the bloodstream for their
peripheral targets. The same peptides are also expressed
within the parvocellular neurons of these two nuclei for
secretion into the pituitary portal circulation, where
vasopressin can act synergistically with corticotropin-releasing
hormone (see below) to release ACTH. In addition,
vasopressin is expressed within a subset of neurons of the
suprachiasmatic nucleus; outside the hypothalamus,
vasopressin is also expressed within some neurons of the bed
nucleus of the stria terminalis (which project to the lateral
septum, medial amygdala, and periaqueductal gray, among
others), the medial amygdala (which project to the ventral
hippocampus), the septum, and reputedly within neurons of
the locus ceruleus. While neurons in these target regions are
definitely responsive to the peptides infused locally,
documentation of synaptic actions is still lacking. An
interesting but mysterious aspect of the biology of these
neurons is their capacity to transport the mRNAs for their
peptides into their axons. This transport is more easily
detected during periods of functional load (postpartum or salt
load), when the mRNA is also transported back again to the
perikaryon, where the messages might either be translated or
in some manner regulate translation.

The behavioral actions of vasopressin are quite impressive,

although the mechanisms accounting for them remain
unclear. These effects are described in somewhat greater
detail in Chapter 13.

Analogs of vasopressin and oxytocin have now been

developed with selective agonist or antagonist properties;
these molecules, often shorter and more stable than the
native peptide, should make useful pharmacological probes.
Based on the selective actions of synthetic vasopressin
analogs, two subtypes of peripheral receptor have been
characterized: V1 receptors, which mediate responses on
arteriolar walls, and V2 receptors, which mediate the effects
on the renal tubules. A V1 receptor agonist has been reported
to maintain long-term potentiation in neurons of the lateral
septal nucleus (see Chapter 13). The effects of vasopressin
analogs on either memory or learning behaviors or on
secretory responses mediated through the median eminence
do not adhere to either receptor class and may indicate that
more subtypes will be characterized. At least four arginine
vasopressin receptors have been cloned. The dynamic
regulation of oxytocin and vasopressin expression is sensitive
to gonadal steroids, thus providing a gender-specific neuronal
expression pattern that is turned on at puberty and that can
be further regulated through reproductive function. Novice
neuropeptide researchers have an opportunity left behind by
the marauding peptide pioneers: what is the function of
oxytocin in the male brain?

Figure 11-1. Molecular sequence of oxytocin and

vasopressin, in which amino acid names are
symbolized by the standard single-letter symbols
and * indicates an amidated C terminus. (Key to
amino acid single-letter symbols: A, Ala; R, Arg;
N, Asn; D, Asp; C, Cys; Q, Gln; E, Glu; G, Gly; H,
His; I, Ile; L, Leu; K, Lys; M, Met; F, Phe; P, Pro;
S, Ser; T, Thr; W, Trp; Y, Tyr; V, Val.)

The Tachykinin Peptides

In 1931, von Euler and Gaddum discovered an unexpected
pharmacologically active substance in extracts of brain and
intestine, which they later named substance P because the
activity was present in the dried acetone powder of the
extract. Although studied intermittently through bioassays,
substance P remained somewhat shaded in obscurity until
almost 40 years later, when Leeman and co-workers purified a
sialogogic peptide from hypothalamic extracts while looking
for the still elusive corticotropin-releasing factor. This
sialogogic peptide turned out to have an amino acid content
and pharmacological profile identical to substance P, and,
when finally purified, sequenced, and synthesized, it was
identified as substance P, an undecapeptide (Fig. 11-2).
Knowledge of the structure and availability of large amounts
of the synthetic material permitted the development of
radioimmunoassays and immunocytochemical tests, which
were then used to map the brain and assay its substance P
content. Substance P is present in small neuronal systems in
many parts of the CNS, and upon subcellular fractionation, it
is found in the vesicular layer. It is especially rich in neurons
projecting into the substantia gelatinosa of the spinal cord
from the dorsal root ganglia and has even been proposed to
be the transmitter for primary afferent sensory fibers. While it
is very potent as a depolarizing substance in direct tests of
spinal cord excitability (about 200 times stronger on a molar
basis than GLU), its long duration of action prompted caution
in accepting substance P as the primary sensory transmitter.
Other peptides have also been identified in dorsal root
ganglion cells, and GLU released from cultures of these cells
also produces a prompt, powerful excitatory action on spinal
dorsal horn neurons. If GLU and substance P were
cotransmitters for some sensory fibers, substance P could be
viewed as prolonging and intensifying the transmission into
the cord.

Radioimmunoassays and immunocytochemistry also show

brain regions other than the spinal cord to be rich in
substance P, especially the substantia nigra, caudate
putamen, amygdala, hypothalamus, and cerebral cortex.
Tests with iontophoretic application in these regions generally
indicate excitatory actions, again with a long duration. The
dynamic regulation of substance P within the rat striatum has
been informative. Substance P-containing neurons of the
striatum project to the dopamine neurons of the substantia
nigra in reciprocal circuitry. Dopamine antagonist or 6-(OHDA)
treatment of the substantia nigra leads to a substantial drop
in substance P content; analysis of the peptide, propeptide,
and mRNA indicates that this drop is due to decreased gene
expression and decreased release. After treatment with
methamphetamine, substance P synthesis is accelerated.
After axotomy, the levels of substance P and several of its
coexisting neuropeptides plunge in dorsal root ganglion
neurons but rise in axotomized sympathetic ganglioneurons.
In the human neurological disease Huntington's chorea,
characterized by profound movement disorders and
psychological changes, substance P levels in the substantia
nigra are considerably reduced while alternative
neuropeptides are expressed.

Tachykinin research has been one of the windfall areas

benefiting from the application of molecular biological
methods. The cloning of the precursor forms of the three
main tachykinin peptides in the mammalian CNS and their
first receptor subtypes has helped to clarify their relationships
to each other and to other neurotransmitters. There are two
mammalian tachykinin genes. The neurokinin A (NKA) gene
(on human chromosome 7) can produce three forms of mRNA
through alternative splicing of its many exons. The least
prevalent form encodes only substance P, and the other two
forms encode both substance P and NKA (the term the
tachykinin nomenclature committee prefers for the peptide
previously called substance K). In general, the tissue and
cellular distributions of substance P and NKA are similar. A
second tachykinin gene, located on human chromosome 12,
encodes the neurokinin B (NKB) precursor, previously called
neuromedin K. Receptors for all three peptides have been
cloned, and all are G protein-coupled receptors (see Chapter
6). The receptors are called NK1, NK2, and NK3. Substance P
is more potent than NKA or NKB at the NK1 site but much
less potent than NKA or NKB at the NK2 site. At the NK3 site,
NKB is slightly more potent than NKA or Substance P. Inositol
phosphate breakdown is a major transducing pathway for at
least the NK1 and NK2 receptors. After activation, either
physiologically or pharmacologically, NK1 receptors are
internalized into the smooth endoplasmic reticulum of
receptor-bearing dendrites (see Fig. 11-3). Highly selective
nonpeptidic agonists and antagonists have helped to define
the potential behavioral actions of substance P. However,
drugs acting at the human NK3 receptor do not work well on
rat NK3, eliminating the most frequently used animal model.
Based on the rich expression of NK1 receptors on 5-HT, NE
and DA-containing neurons, an NK1 antagonist originally
developed for its antiemetic properties was clinically tested as
an antidepressant and found in initial clinical trials to be
equipotent with a serotonin-selective reuptake inhibitor. The
field is now waiting for replication of this exciting finding.
However, NK1 antagonists have had no effect on pain, even
though knockout of the receptors or the dorsal horn neurons
expressing them does reduce chronic inflammatory pain.
However, dorsal horn afferent fibers contain multiple other
peptides (see Fig. 11-4).

Figure 11-2. Structural homologies between the

peptides of the tachykinin family, presented
according to the schema of Figure 11-1. pE,
pyroglutamate.

Figure 11-3. The vasoactive intestinal peptide

(VIP)-related peptide family represented by their
single-letter amino acid symbols. The sequences
of PHI-27, PHM-27, growth hormone-releasing
hormone 1-24, glucagon, and secretin that match
those of VIP are indicated in bold letters. For an
interesting exercise, the reader may wish to
construct a complementary table in which matches
to glucagon are highlighted.

Figure 11-4. Electron micrographs of large

dense-core vesicles (LDCVs) in calcitonin gene-
related peptide (CGRP)-immunoreactive primary
afferent terminals in lamina II of the dorsal horn
of rat lumbar spinal cord after triple-immunogold
staining for galanin (GAL)-, CGRP-, and substance
P (SP)-like immunoreactivity (LIs) (A); for GAL-,
cholecystokinin (CCK)-, and CGRP-LIs (B); or for
GAL-, CGRP-, and neuropeptide Y (NPY)-LIs (C).
The size of gold particles for each peptide is
indicated in the micrographs. The different
examples of colocalization of neuropeptides in
 examples of colocalization of neuropeptides in
LDCVs are shown. (A) Curved arrow indicates 3-
labeled LDCVs. Arrowhead, CGRP alone. (B)
Curved arrow indicates 3-labeled LDCVs.
Arrowhead, CGRP alone; open arrow, CGRP plus
growth hormone. (C) Curved arrow indicates 3-
labeled LDCVs. Arrowhead, GAL and CGRP; double
arrowheads, CGRP alone; open arrow, GAL + NPY;
thick arrows, CGRP + NPY; thin arrow points to a
synapse. Bars indicate 100 nM. (From X. Zhang
and T. Hokfelt.)

VIP-Related Peptides
VIP, a peptide composed of 29 amino acids, was originally
isolated from porcine intestine, using the C-terminal
amidation strategy of Mutt. It was originally named for its
ability to alter enteric blood flow. Establishment of its
sequence revealed marked structural similarities to glucagon,
secretin, and another peptide of gastric origin that inhibits
muscular contraction, gastric inhibitory peptide. These
structural similarities constitute the VIP-related peptide family
(see Fig. 11-5), named for the member most enriched in
brain. The development of synthetic VIP for preparation of
immunoassays and cellular localization revealed that VIP
exists independently of its cousins in the gut and pancreas
and that it was prominent in many regions of the CNS and the
autonomic nervous system. In parasympathetic nerves to the
cat salivary gland, VIP coexists with ACh and is apparently
released with ACh as part of an integrated command to
activate secretion and to increase blood flow through the
gland. In the CNS, VIP-reactive neurons are among the most
numerous of the chemically defined cells of the neocortex,
and they exhibit there a very narrow radial orientation,
suggesting that they innervate cellular targets located wholly
within a single cortical column assembly. In peripheral
structures, VIP-reactive nerves innervate the gut, especially
sphincter regions, as well as the lung and possibly even the
thyroid gland.

Additional members were added to this important family of

peptides when Guillemin and colleagues isolated the long-
sought growth hormone-releasing hormone (GHRH) peptide
and recognized the homologous shared amino acid
sequences. Two later members of the family are termed PHI-
27 and PHM-27, to reflect their size, 27 amino acids long, and
the one-letter initials of their N-terminal (H = histidine) and
C-terminal (isoleucine or methionine) peptides (where P
stands for peptide, not proline). PHI-27 was initially isolated
by Mutt and associates from gut extracts while searching for
other C-terminally amidated peptides, a property all active
members of this family share. PHM-27 was identified from the
deduced sequence of the pro-VIP mRNA in cloning
experiments. Growth hormone-releasing hormone has a
relatively limited neuronal distribution concentrated in the
hypothalamus and median eminence. No cellular maps for PHI
or PHM in the brain have yet been reported.

The next addition to the VIP-related peptide family is or

pituitary adenylate cyclase-activating peptide (PACAP), which
has a 1000-fold greater capacity to activate adenylate cyclase
in pituitary cultures relative to VIP. PACAP is a C-terminally
amidated, 27-residue peptide with a very high conservation of
the VIP sequence (see Fig. 11-5), which is also expressed in
the brain; a 38-residue, C-terminally extended form has also
been isolated from the brain and may be the predominant
form. Two forms of PACAP receptor have been cloned, and all
receptors for peptides in this family show strong structural
similarity.

To reverse a trend, molecular cloning of the mRNA for the

precursor of secretin, another member of this peptide
superfamily, indicates that this specific peptide is not actually
found in the brain, as had been proposed based on
immunocytochemistry with polyclonal antibodies some years
ago. Nevertheless, fueled in part by reports that systemic
secretin can improve the behavioral problems of autistic
children, there has been renewed interest in the biological
actions of secretin in the brain. Most recently, a novel
member of this family, found to be enriched in hypothalamus
by differential expression patterns, was identified twice in
rapid order by different strategies. In the first, sequencing of
hypothalamus-enriched mRNA revealed a pair of peptides
(named hypocretin-1 and -2) with sequence similarity to
different domains of secretin than seen with the previously
identified members. Then, independently, a group screening
for natural ligands for orphan G-protein receptors identified
the same peptide and demonstrated a potent appetitive
action, for which it was named orexin A and B. The
hypothalamic neurons expressing the gene for
hypocretin/orexin are held to be engaged in both appetite and
blood pressure regulation. This same peptide gene was
associated with individuals exhibiting inheritable narcolepsy.

Figure 11-5. The pancreatic polypeptide family

represented by their single-letter amino acid
symbols. The sequences of peptide YY (PYY),
avian pancreatic polypeptide (APP), and human
pancreatic polypeptide (HPP) that match those of
NPY are indicated in bold letters.

Pancreatic Polypeptide-Related Peptides

The pancreatic polypeptides were recognized in extracts of
pancreatic islets in the mid-1970s, and those from pigeon,
pig, and human pancreas were found to be highly homologous
36-residue peptides with amidated C termini (see Fig. 11-6).
Antisera against these peptides recognized cells and fibers in
the CNS and autonomic nervous system, but these
immunocytochemical observations were rightly qualified as
"pancreatic polypeptide-like immunoreactivity" because when
the same sera were used in their highly dilute form for
radioimmunoassay, only scant extractable material was
detected. When Mutt and colleagues applied their C-terminal
amidated peptide isolation strategy to these extracts, two
more members of this family were identified. The first one,
found in gut, was given the name PYY (a neuropeptide with
tyrosine [single-letter amino acid symbol Y] residues at both
the N and C termini). However, because the one found in
brain had the same length and the same N- and C-terminal
tyrosines, it was named NPY. In rapid order, this peptide's
distribution was described in rodent and human brains,
showing NPY to be one of the most extensive central peptide
systems. Particularly high amounts were found in the
hypothalamus, limbic system, and neocortex. In many places,
including the autonomic nervous system, NPY coexists with
either NE or epinephrine. NPY increases the sensitivity of
sympathetically innervated smooth muscle to NE and is one of
the most potent natural vasoconstrictors known. Upon
sequencing of the gene in many species, from humans to fish,
the structure of the NPY mRNA is one of the most highly
conserved genes ever reported, surpassing even the
conservation seen in insulin. Within the C terminus of the pre-
NPY peptide, a second potential cleavage product, known as
the C-terminal peptide of NPY, (or CPON) is almost as highly
conserved. Since CPON will be formed to liberate NPY, it is
frustrating that there are still no reports of its actions more
than a decade after its description.

NPY, especially in the autonomic nervous system, documents

the signal value of shortened versions of the secreted peptide.
Three types of receptor have been detected
pharmacologically. The postsynaptic Y1 requires the full
NPY(1-36) for action and enhances responses to coreleased
NE. The presynaptic Y2 reacts as well to N-terminally
truncated forms of NPY and reduces NE release. Central
injection of NPY suggests active roles in anxiety reduction and
appetite stimulation, partially confirmed by the anxiogenic
effects of intraventricularly injected antisense RNA
oligonucleotide for the Y1 receptor but not the one for Y2.
During fasting, hypothalamic levels of NPY rise. Nevertheless,
mice in whom the NPY gene is knocked out continue to eat
normally, although in genetically obese rodents lacking leptin,
knockout of NPY expression significantly reduces the obesity.

Figure 11-6. Structural relationships among the

prohormone, precursor forms of the three major
branches of the opioid peptides, depicted as a bar
diagram, whose length in amino acid residues is
indicated by the small number at the
corresponding C termini. Locations of the
repeating peptide sequences are indicated. Basic
amino acid sequences that constitute consensus
cleavage sites for processing are indicated as
single or double vertical lines within the bar.
Proopiomelanocortin and proenkephalin have
consensus signal peptides indicated at their N
termini. MSH, melanocyte-stimulating hormone;
ACTH, corticotropin.

Opioid Peptides
An endorphin is any endogenous substance (i.e., one naturally
formed in the living animal) that exhibits the pharmacological
properties of morphine. When this term was first coined in
mid-1975, it was a useful abstraction covering "morphine-like
factors" from brain extracts and spinal fluids that were active
in opiate ligand-displacement assays or opiate-sensitive
smooth muscle assays. Within a year, a highly competitive
effort resulted in the purification, isolation, sequencing, and
synthetic replication of not only one but nearly a half-dozen
peptides that deserved the term endorphin.
The incredible explosion of work on this class of peptides
began with attempts to isolate and characterize the receptor
to opiates as part of a molecular biological approach to the
question of narcotic addiction. When specific binding assays
were developed, substantial evidence was accumulated
independently and almost simultaneously by Sol Snyder's
group at Johns Hopkins, Eric Simon's group at New York
University, and Lars Terrenius' group at Uppsala University
that a high-affinity binding site in synaptic membranes
showed stereoselective opioid recognition properties. A whole
new approach to neurotransmitter identification took shape
when John Hughes, Hans Kosterlitz, and their colleagues in
Aberdeen demonstrated that extracts of brain contain a
substance that can compete in the opiate receptor assays and
show opioid activity in in vitro smooth muscle bioassays.

In late 1975, this endogenous opioid activity was attributed to

two pentapeptides, named enkephalins, which shared a
tetrapeptide sequence, YGGF, varying only in the C-terminal
position: hence, they were called Met5-enkephalin and Leu5-
enkephalin. Perhaps even more dramatic than the
announcement that the brain contained not one but two
opioid peptides was the realization that the entire structure of
Met5-enkephalin is contained within a 91-amino acid pituitary
hormone, -lipotropin ( -LPH), whose isolation and sequence
was reported several years earlier by C. H. Li but whose
function was unknown.

Several groups then reported the isolation, purification,

chemical structures, and synthetic confirmation of three
additional endorphin peptides: -endorphin ( -LPH61-76), -
endorphin ( -LPH61-77), and -endorphin ( -LPH61-91, also
called C fragment). Numerous claims and counterclaims were
parried across the symposium stages for several months as to
one person's peptide being another person's artifact or
precursor.

Subsequent research has greatly clarified the molecular and

genomic relationships between the three major branches of
the opioid peptide family (see Fig. 11-7): the
proopiomelanocortin (POMC)-derived peptides, the
proenkephalin-derived peptides, and the prodynorphin-
derived peptides. The superfamily has been further expanded
with the heptadecapeptide orphanin FQ, also called
nociceptin, identified as an activator of an opioid-like orphan
receptor. The natural peptide shows no affinity for opioid
receptors. Likewise, opioid peptides and ligands do not bind to
the orphanin receptor. Nevertheless, microinjection of
orphanin into the periaqueductal gray blocks the anti-
nociceptive effects of morphine injected into the same site.
While it is known from receptor knockout studies that the m-
opiate receptor is required for the effects of exogenous
morphine, none of the originally described opioid peptides
showed impressive affinity for this receptor. This mystery may
have been partially solved by the identification of two novel
endogenous tetrapeptides that show very little similarity to
the key N-terminal YGGF of all the others. Other structurally
related natural peptides lack opioid receptor activity
altogether, such as the invertebrate cardioacceleratory
peptide FMRF-amide and its mammalian counterparts
neuropeptides FF and AF); nevertheless, neuropeptide
expression is altered under experimental chronic pain and
reduces the response to centrally injected opiates.
Furthermore, some opioid-acting peptides have been found
exogenously in milk and in plant proteins and have been
called exorphins.

Figure 11-7. An ultrastructural basis for

functional interaction between central opioid-
containing neurons and central catecholamine
neurons is provided by this dual label,
autoradiographic image obtained by
 immunocytochemistry. Nerve terminals exhibit
leu-enkephalin immunoreactivity, and one also
exhibits tyrosine hydroxylase immunoreactivity
(small autoradiographic grains, terminal labeled LE
+ TH-T). The latter makes a symmetric synaptic
contact on a dendrite (TH-D) bearing many
autoradiographic grains, indicating extensive
tyrosine hydroxylase content. Scale bar lower
right = 0.5 um. (Modified from Milner et al.,
1989.)

Chemical and Cellular Relationships among

the Opioid Peptides
1. The proopiomelanocortin (POMC) peptides are expressed
independently in the anterior pituitary, the intermediate lobe
of the pituitary, and one main cluster of neurons in the area
of the arcuate nucleus of the hypothalamus. The major
endorphin agonist produced from POMC is the 31-amino acid
C-terminal fragment -endorphin, the most potent of the
natural opioids. N-terminal fragments of -endorphin are
much less potent, and analogs with no N-terminal tyrosine
(so-called des-Tyr versions) lose all opioid activity, although in
some hands des-Tyr peptides are reputed to be active
behaviorally. In the corticotropin-secreting cells of the
anterior pituitary, POMC is processed largely to corticotropin
and to an inactive form of -endorphin; in intermediate lobe
cells and arcuate neurons, the same precursor is processed to
-MSH and active -endorphin. A third MSH-containing
heptapeptide component, discovered during the cloning and
sequencing of the POMC mRNA, suggests that yet another
end product may be possible.

2. The enkephalin pentapeptides Met5-enkephalin and Leu5-

enkephalin are expressed in wholly separate neuronal
systems from the POMC neurons and are much more
pervasively distributed throughout the CNS and peripheral
nervous system, including the adrenal medulla and enteric
nervous system. Cloning and sequencing of the mRNA for the
proenkephalin, starting with mRNA from adrenal medullary
tissue, produced an unexpected dividend in that the precursor
exhibits multiple copies of the two peptides in almost exactly
the 6:1 ratio of Met5- to Leu5-enkephalins that had been
described in regional brain and gut assays. This solved the
mystery of the two similar peptides.

3. The prodynorphin peptides consist of C-terminally extended

forms of Leu5-enkephalin arising from a different gene and
from a different mRNA that encodes for production of four
major peptides: dynorphin A, dynorphin B, and two
neoendorphins, and . These C-terminally extended
peptides act as potent opioid agonists without cleavage down
to the enkephalin pentapeptide form, and upon mapping they
were found to represent a third separated series of rather
generally distributed central and peripheral neurons. The
amino acid sequence of orphanin bears greater resemblance
to dynorphin than to any of the other opioid peptides (see Fig.
11-7).

Each of the separate classes of neurons containing -

endorphin, enkephalin, or dynorphin peptides has distinct
morphological features. The -endorphin-containing neurons
are long projection systems that fall within the general
endocrine-oriented systems of the medial hypothalamus,
diencephalon, and pons. The proenkephalin-derived peptides
and the prodynorphin-derived peptides are generally found in
neurons with modest to short projections, groups of which are
widespread and some of which innervate presumptive
dopaminergic neurons, potentially providing a basis for some
endogenous reward properties (see Fig. 11-8). In some
regions, the enkephalin-derived and dynorphin-derived
peptides show intriguing relationships. For example,
enkephalin-containing neurons project from the entorhinal
cortex to the molecular layer of the dentate gyrus of the
hippocampus, while dynorphin-containing neurons project
from the dentate gyrus to the CA3 pyramidal cells. In the
spinal cord, intrinsic interneurons contain dynorphin peptides,
while descending long axons from the pons and medulla
contain enkephalin peptides. In addition, all of these peptide
systems presumably contain other nonpeptidic transmitters
that have not yet been rigorously identified as well as some
peptides that have.

Figure 11-8. Structure of the prohormone form

of somatostatin, with the location and sequence of
the 28-amino acid residue form at the C terminus
and the relative sequences and locations of the
different forms of somatostatin indicated.

Physiological and Pharmacological Effects

Traditional (i.e., precloning) pharmacology produced at least
three schemes of endorphin-receptor classification based on
(1) the comparative actions of morphine or of morphine-like
drugs with mixed agonist-antagonist actions on the dog spinal
cord (the , , and scheme of Martin and Gilbert), (2) the
relative effects of endogenous and synthetic opioids on
various smooth muscle in vitro assays (the and scheme of
Kosterlitz or the , , and scheme of Herz), or (3) guanosine
triphosphate-modifiable ligand binding. Remarkably, given the
impetus for their discovery, none of the opioid receptors was
affected by chronic exposure to morphine. The best
concordance between receptor subtypes and opioid peptide
systems remains that of the dynorphin-derived peptides
reported by Goldstein and Chavkin, to act on the opioid
receptors in both gut and brain.
The membrane mechanisms transducing the various classes
of opioid receptors are not well understood, but several
categories of opioid action have been characterized at the
molecular level. These include inhibition of adenylate cyclase,
inhibition of the N-type Ca2 conductance that underlies
transmitter release, enhancement of the M current K+
conductance, as well as activation of the inward rectifier and
delayed rectifier K+ conductances. Intracellular recordings of
gut neurons and brain neurons in vitro indicate that these
cells respond by hyperpolarization to acute administration of
opioids mediated by an increased K+ conductance that
secondarily depresses Ca-dependent spike activity. In the
dorsal horn, neither membrane potential nor conductance is
impressively altered but responsiveness to depolarizing
synaptic potentials is reduced.

Iontophoretic tests with enkephalins and -endorphin suggest

that neurons throughout the CNS can be influenced by these
peptides at naloxone-sensitive receptors. In general,
responses tend to be depressant, except in the hippocampus,
where excitatory actions are so profound that hippocampal
seizures can be induced. However, the mechanism of this
excitation is the exception that proves the rule, as it is based
on inhibition of GABA-ergic inhibitory interneurons bearing the
opiate receptors and produces excitation by disinhibition. In
such tests, all endorphins show qualitatively equivalent effects
and similar onset of actions, but -endorphin may be more
potent and longer lasting due to its slower hydrolysis.
Peptides with no N-terminal Tyr are much weaker, but the
potency of endomorphins with only the N-terminal Tyr in
common suggests that there is much more to be learned.

This picture of opioid receptor pharmacology changed

irrevocably within a few short months, beginning in the fall of
1992 when first one (the receptor) and then all three of the
opioid peptide receptor subtypes were cloned. As predicted,
all were similar G protein-coupled receptors with some very
surprising similarities to other peptide receptors outside the
opioid peptide family. Considering all of the potent and
receptor subtype-specific agonists and antagonists and their
assessment in the hippocampus, it was only a matter of time
before the well-defined opioid peptide-containing circuits
there were subjected to analysis of their roles in synaptic
transmission.

The results of these studies (see review by Wagner and

Chavkin, 1994) have paid off in what is arguably the best
evidence yet for central synaptic actions of any neuropeptide.
High-frequency stimulation of the perforant path will activate
dynorphin-containing dentate granule cells. This in turn leads
to a naloxone-sensitive suppression of IPSPs recorded from
CA3 pyramidal cells. This initial claim of synaptic action was
soon attributed to presynaptic opioid suppression of NE
release. A second synaptic effect was then detected as a
blunted excitatory response to perforant pathway stimulation
in the dentate granule cells themselves (the ones containing
the dynorphin). However, this latter effect turned out to be a
highly novel mechanism in which dendritic dynorphin acted as
a retrograde transmitter to reduce presynaptic release of
glutamate. Extended analysis of the ability of naloxone to
block hippocampal long-term potentiation (LTP, see Chapter
12) provided the best evidence for more conventional
synaptic actions. Here, antagonists block the LTP induced in
CA3 neurons by mossy fiber stimulation, and both u- and -
receptor antagonists block the LTP produced by stimulation of
the lateral perforant pathway (a proenkephalin-containing
pathway). Furthermore, LTP produced in dentate granule cells
by high-frequency stimulation in the hilus of the hippocampus
is enhanced with agonists, and this effect is blocked with
antidynorphin antibodies or -selective antagonists.
A minor disappointment in the field of opioid peptide research
has been its failure to deliver on the promise of providing
insight into the nature of drug dependence. Despite many
attempts, little significant change in the concentrations of any
of the opioid peptides or their receptors can be detected with
drug dependence. Studies of morphine-dosed animals show
initial increases in -endorphin mRNA, leading to a C-
terminally shortened form that may be either a very weak
agonist or, more interestingly, an opioid antagonist.
Surprisingly, NMDA antagonists can block opiate tolerance
formation. Chronic naltrexone treatments, which up-regulate
ligand binding sites for the shorter opioid peptides, also
increase their mRNAs by severalfold in the rat striatum and,
to lesser degrees, in other opioid-sensitive brain regions;
however, the concentration of peptides encoded by the
mRNAs is not changed.

Although the behavioral effects of these peptides once

attracted much attention, the area has not maintained that
momentum. The enkephalins produce only transient analgesia
after direct intracerebroventricular injection; in such tests, -
endorphin is 50-100 times more potent than morphine on a
molar basis. The fact that blood-borne peptides do not
penetrate into the brain well or survive the gauntlet of
peptidases to which they are exposed in the process has
spurred the search for more effective analogs, although
surpassing the morphine analogs has been difficult. The sheer
number of whole-animal effects that have been interpreted as
endorphinergic physiology on the basis of naloxone effects
has grown considerably but remains open to more
comprehensive analysis. Among the proposed physiological
properties that may be regulated by one or another of the
endorphin substances are blood pressure, temperature,
feeding, sexual activity, and lymphocyte mitosis, along with
pain perception and memory. All of this seems like a lot to ask
of one peptide family, but then, who are we to ask what a
peptide may, or can, do for us? As we said before, don't blink
or you'll miss the next blazing development.

INDIVIDUAL PEPTIDES WORTH TRACKING

Introduction
We conclude this chapter with brief comments on a few of the
other neuroactive peptides that are under active investigation.
Around those we have selected, bodies of research are
consolidating and the neuropharmacological implications
seem strong.

Somatostatin (Somatotropin Release-

Inhibiting Factor)
In 1971, workers began to look for other goodies in their
hard-won hypothalamic extracts. Vale, Brazeau, and Guillemin
tried their extracts for potency in releasing growth hormone
from long-term cultured anterior pituitary cells and found to
their amazement a factor that inhibited even basal growth
hormone release in minute amounts. They named this factor
somatostatin. Isolation and purification studies eventually
culminated in the characterization of this molecule as a
tetradecapeptide with a disulfide bridge between Cys3 and
Cys14 (Fig. 11-8). Radioimmunoassays,
immunocytochemistry, and whole-animal tests of the
synthetic peptide made it clear that somatostatin was doing
more than just inhibiting growth hormone release.
Somatostatin was found to be widely distributed in the
gastrointestinal tract and pancreatic islets, localized in the
latter tissue to the cells by immunocytochemistry.
Somatostatin in islets can apparently suppress the release of
both glucagon and insulin, and in diabetic humans, who have
no insulin, suppression of glucagon can be an important
element in determining insulin requirements. The mechanisms
of this suppression are not known but may be similar to the
effect of somatostatin on growth hormone release from
pituitary, an action accompanied by suppression of TSH
release.

When somatostatin is injected intracerebroventricularly,

animals show decreased spontaneous motor activity, reduced
sensitivity to barbiturates, loss of slow-wave and REM sleep,
and increased appetite. Radioimmunoassays and
immunocytochemistry of rat brain regions show that
somatostatin is largely concentrated in the mediobasal
hypothalamus, with much smaller amounts present in a few
other brain regions (see Fig. 11-9). Immunocytochemical
studies led to the surprising finding that somatostatin-reactive
cells and fibers are also present in dorsal root ganglia; the
autonomic plexi of the intestine; and the amygdala,
hippocampal formation and neocortex. Avian and amphibian
brains contain considerably more somatostatin than
mammalian brains. Physiological tests on the isolated frog
spinal cord suggest that it may facilitate the transmission of
dorsal root reflexes after a long latent period. Iontophoretic
tests on rat neurons indicate a relatively common depressant
action that is brisk in onset and termination. At the level of
cellular electrophysiology, somatostatin hyperpolarizes
neurons in a potent manner and can dynamically open the so-
called M current channel, which cholinergic muscarinic
receptors close in order to excite hippocampal and other
neurons. The net result of somatostatin's actions in the intact
brain is to enhance responsiveness to ACh. The important
missing datum is what somatostatin does to the responses to
GABA, its frequent but not universal coexistence partner.

An endogenous large somatostatin, extended at the N

terminus by an additional 14 amino acids, also exists in both
brain and gut and shares equipotency for many somatostatin
actions. In rodent and primate neocortex, the N-terminally
extended forms reveal a far more extensive neuronal density,
which at least overlaps with the intrinsic GABAergic cortical
neurons. In early-onset Alzheimer's disease, cortical
somatostatin content is depleted, and somatostatin-28 (SS-
28)-immunoreactive neuritic processes are involved in the
formation of the hallmark Alzheimer's lesion, the plaque.
Cysteamine, a drug developed to treat the metabolic disorder
known as cystinosis, will selectively deplete somatostatin-14
(SS-14) without altering brain levels of SS-28 or SS-281-12,
suggesting that these peptides may be stored in separable
subcellular compartments.

As we have seen with the other peptides, receptor subtypes

are the norm, and here cloning studies have raised the ante
well above what was expected from the physiology: five
different cloned receptors have now been reported. Two of
them align fairly well on distributional maps: SRIF1 was
expressed in dentate granule cells, striatum, locus ceruleus,
and deeper layers of neocortex and showed homologous
desensitization, while SRIF2 was seen more diffusely in
cortex, was more abundant in the CA1 field of the
hippocampus, and did not desensitize. The third cloned form
matches neither, while the fourth form binds SS-28 with
greater affinity than SS-14.

Although somatostatin was an unexpected by-product of the

race to identify the positive regulators of anterior pituitary
release, its pharmacological and diagnostic potentials promise
an important future. Somatostatin deficits in cerebrospinal
fluid have been linked to affective disorder and to Alzheimer's
disease. Long-lasting analogs of somatostatin have already
been clinically applied in the control of pituitary
hypersecretion of growth hormone. Evidence from cell lines
transfected with various G protein-coupled receptor mRNAs
suggests that under some conditions the somatostatin 5
receptor may form functional heterodimers with the DA2
receptor.

Quite surprisingly, in a search for genes expressed in the

cortex and hippocampus, a closely related neuropeptide,
cortistatin, has been identified and is restricted to sparse
GABAergic interneurons. This 14-residue peptide shares 11
amino acids with somatostatin and binds to all five cloned
somatostatin receptors when expressed in transfected cells.
Cortistatin also depresses neuronal activity, antagonizes some
effects of ACh, but can be distinguished from somatostatin.
Cortistatin mRNA accumulates during sleep deprivation.

Figure 11-9. Ultrastructural immunoperoxidase

localization of somatostatin in a nerve terminal
making synaptic contact with an unreactive small
dendritic spine in the rat neocortex. Calibration
bar = 0.5 u. (Unpublished micrographs from
Battenberg and Bloom, The Scripps Research
Institute.)

Cholecystokinin
Among the gut hormones with the longest histories are
gastrin and cholecystokinin (CCK). Subsequently, when their
molecular structures were determined, a striking homology
was revealed between the C termini. After some confusing
interludes, it became recognized that the gastrin-like material
extractable from nervous tissue was caused by antisera cross-
reacting with the C-terminal octapeptide of CCK, whose sole
tyrosine is sulfated. As it turns out, there is more CCK-8 in
the brain than there is in the gut, but the N-terminally
extended forms found in plasma, especially after eating, arise
from the periphery. CCK is among the most promiscuous of
the coexisting peptides, being found with dopamine and
neurotensin in the substantia nigra and ventrotegmental
area; with VIP, NPY, and GABA in thalamocortical and
thalamostriatal connections; and with substance P and 5-HT
in medullary neurons. It is also seen in the dorsal root ganglia
with several other peptides (see Fig. 11-4). Interestingly,
when animals are given morphine, although enkephalins do
not change, brain CCK is depleted while cerebrospinal fluid
CCK rises. Exogenous CCK has opioid antagonist properties.
In iontophoretic tests on spontaneously active hippocampal
pyramidal cells, CCK-8 potently activates firing, a potentially
synaptic action compatible with the CCK contained within the
dentate granule cell mossy fiber synapses to these neurons.

Two (and only two, a rarity in this business) forms of CCK

receptor were characterized pharmacologically for their
responsivity to the sulfated (CCKA, known to be rich in the
pancreas) or unsulfated (CCKB, known to be rich in the
stomach) forms of gastrin and CCK. Both forms have now
been cloned and are also expressed in brain. In fact, the
CCKB receptor is the predominant brain form, while CCKA is
expressed by neurons in the nucleus of the solitary tract, the
area postrema, the posterior hypothalamus, and the nucleus
accumbens. CCK pharmacology was for a time the leader in
de novo development of nonpeptidic antagonists. Antagonists
of the CCKA form have been linked functionally to
enhancement of opiate analgesia and to centrally elicited
postprandial satiety. Antagonists of the CCKB form have been
linked functionally to anxiety disorder, which is supported by
the observation that intravenous CCK tetrapeptide or CCK-8 is
anxiogenic in humans and experimental animals. Low doses
of systemic or iontophoretic benzodiazepines block CCK-
induced cell firing.

Neurotensin
Neurotensin, a tridecapeptide (see Fig. 11-10) and one of the
few that are not C-terminally amidated, became quite
interesting to neuropsychopharmacologists after its
inauspicious origins as a by-product discovery during the
search for corticotropin-releasing factor (CRF). In screening
for what this structurally independent peptide might do,
researchers found that its neuropharmacological properties
after intracisternal administration included potent analgesia,
hypothermia, accentuation of barbiturate and ethanol sleeping
time, and increased release of growth hormone and prolactin.
Mapping studies showed that the largest amounts of
neurotensin are present in the anterior and basal
hypothalamus, the nucleus accumbens and septum, the
midbrain dopamine neurons (along with CCK), and selected
scattered neurons in the spinal cord and brain stem. This led
to the speculation that with its profile of sedation,
hypothermia, and analgesia, neurotensin might be an
"endogenous antipsychotic peptide." However, biochemical
neuropharmacologists really began to pay attention when it
was found that treatment of rats with conventional
antipsychotic drugs increased levels of neurotensin and its
mRNA in both striatum and nucleus accumbens, while
comparable treatment with the so-called atypical
antipsychotic drugs (i.e., those that do not cause the
extrapyramidal symptom complex known as tardive
dyskinesia, see Chapter 13) only increased neurotensin
peptide and mRNA in nucleus accumbens. Furthermore, this
change in gene expression would occur whether D2 dopamine
receptors were blocked or not; D1 agonists produced similar
effects. While it has been reported that cerebrospinal fluid
levels of neurotensin are reduced in unmedicated
schizophrenics and recover somewhat with treatment, the
story remains incomplete as to its role in either pathogenesis
or treatment. The human and rat neurotensin receptors have
been cloned and show similar pharmacology, with inositol
triphosphate hydrolysis as the transducing mechanism.
Cellular tests of neurotensin action suggest that it may
produce excitation, but the mechanisms are not yet defined.
 Figure 11-10. Homologies between ovine and
rodent corticotropin-releasing factor (CRF) and
sauvagine are indicated by bold single-letter
amino acid symbols. Below are indicated the
amino acid sequences of calcitonin gene-related
peptide (CGRP) and the C-terminal octapeptide of
cholecystokinin (CCK), with a sulfated tyrosine at
position 2. pE, pyroglutamate.

Corticotropin-Releasing Factor
The search for the CRF ended in 1981 when Wylie Vale and
colleagues completed the purification and sequence analysis
of the peptide taken from extracts of sheep hypothalamus.
Thus ended a three-decade-long search for a mystery factor
that put the secretion of corticotropin and other anterior
pituitary hormones under neuronal control and allowed the
brain to replace the pituitary as the master gland of the body.
The remaining member of the original four horsemen of the
hypophysiotrophic-hypothalamic control system was lassoed a
year later when Vale and Roger Guillemin independently
reported the isolation, sequencing, and synthetic replication
of growth hormone-releasing hormone. Nonparticipants in this
highly competitive struggle probably had mixed emotions
when these quests ended since along the way the world of
neuropeptide discovery was at least triply blessed first by
substance P, then somatostatin, and eventually neurotensin,
all of which were discovered by teams intent on CRF.

When CRF was in hand, it turned out to be an interesting 41-

amino acid peptide (see Fig. 11-10) structurally related to a
rather obscure frog skin peptide, sauvagine, and to another
peptide, urotensin, obtained from an even more obscure
organ, the urohypophysis of certain fish. The cellular origin of
mammalian CRF was traced to that portion of the
paraventricular nucleus that projects to the median eminence
but not into the posterior pituitary. Later, these and other
CRF-containing circuits were identified as innervating a very
extensive group of neurons in the pons and medulla, cortex,
and amygdala, perhaps representing a stress-related circuitry.
CRF is an extremely potent ACTH secretogogue, but its effects
are significantly augmented by vasopressin (produced by the
adjacent magnocellular paraventricular neurons) as well as by
NE and angiotensin.

In addition to its premier action in the regulation of

corticotropin secretion, synthetic CRF has equally potent
effects on neurons in vitro and in vivo, increasing the
frequency of action potentials in cells that fire in bursts, such
as the hippocampal pyramidal cells. Even modest doses of the
peptide can induce seizure-like activity within the limbic
system. At still smaller doses given intracerebroventricularly,
CRF is a potent activator of spontaneous locomotion and can
produce an anxiogenic response (opposite to the effects of
antianxiety drugs like benzodiazepines and ethanol) in
different behavior tests. Clearly, the complete physiological
effects of CRF at the pituitary and at other neuronal sites
represent a much more comprehensive basis for the
mobilization of bodily systems during stress.

Initial efforts to define CRF receptors focused on ligand

binding autoradiography, which more or less replicated the
tissue patterns of CRF immunoreactivity, with highest binding
in the cortex, amygdala, hippocampus, and pons. CRF is also
expressed in an as yet unexploited circuit crying for functional
analysis, a projection from the inferior olive to cerebellar
cortex containing CRF and glutamate. In the final days of
pregnancy, maternal blood is very high in apparently high
molecular weight forms of CRF believed to arise in placenta
and yet without signs of excessive corticotropin secretion
generally. Efforts to solve this puzzle led to the definition of a
CRF-binding protein, also well expressed in brain, although
with a distribution different from that of CRF fibers.
The Vale group subsequently characterized two CRF receptors
and developed potent CRF antagonist peptides, knockout mice
for CRF, their receptors, and mice overexpressing CRF. In
addition, they characterized a highly related peptide that is
probably the endogenous ligand for CRF2, termed urocortin
(for its effects, which are similar to urotensin in appetite
suppression; see Fig. 11-10).

A READER'S GUIDE TO PEPTIDE POACHING

Until more is learned about the functions, sites, and
mechanisms of action of any one peptide, we will have trouble
formulating even tentative roles for these substances in the
neurotransmitter-neurohormonal regulation of central drug
actions.

Only the stern hand of our stingy editor prevents us having

the space to discuss in any depth other important
neuropeptides. The next three would be galanin (a free ice
cream cone to the first reader who deduces how it got its
name; clue: it has alanine at its C terminus), a
neuroendocrine peptide that has expression and
pharmacological interest for those focused on depression,
pain regulation, and possibly epilepsy; cocaine and
amphetamine-regulated transcript (CART), a widely expressed
neuropeptide of still uncertain amino acid sequence length,
whose 55-102 fragment is potent at suppressing feeding and
may be among those brain systems responsive to leptin; and
the cytokines and chemokines, an increasingly important
family of peptides for neuropharmacologists. Cytokines are a
large and diverse family of polypeptide regulators, produced
widely throughout the body by cells of many embryological
lines. In general, cytokines interact as a network with
synergistic, additive, or opposing actions. Within the immune
system, macrophages and activated T lymphocytes are the
major producers of the macrophage-derived cytokines
interleukin (IL)-1A, IL-1B, and IL-6 and tumor necrosis factor
. These are the cytokines that have received the most
attention for potential regulatory roles in nervous system
inflammation (as with the early dementia of CNS human
immunodeficiency virus infection) and in recovery from
traumatic injury. Had they been discovered in brain by
neuropharmacologists, they might have "growth factor"
names instead of cytokine numbers. Chemokines are
cytokines whose expression and receptors may naturally
control the interstitial migration of lymphocytes and
macrophages into all organs, including the brain.

As the story of neuroactive peptides unfolds (the sheer

number of peptides available for pursuit makes the going
slower owing to the division of the available workforce),
interested students should be alert for answers to the
following questions: Are there general patterns of circuitry?
Are specific peptidases amenable to selective pharmacological
intervention, or do neuronal and glial peptidases read only the
dipeptides, whose bonds they are about to cleave? Do the
peptides act presynaptically, postsynaptically, or at both sites?
How are peptidases specifically activated either to release
neuroactive peptides from precursors or to terminate the
activity of the peptide? Can the peptides modulate the release
or response to transmitters of other neurons or those with
which they coexist in a given neuron? Can the promise of
peptide chemistry deliver useful antagonists to prove the
identity of receptors with nerve pathway stimulation? Does
the presence of receptors to other centrally active drugs
indicate that still more endogenous peptides should be
sought? By now it should have occurred to the reader that
nothing in the neurosciences is simple.

SELECTED REFERENCES

General sources
Asensio, V. C. and I. L. Campbell (1999). Chemokines in the
CNS: plurifunctional mediators in diverse states. Trends
Neurosci. 22. 504-512.

Bloom, F. E. (1990). Peptides: regulators of cell function in

brain and beyond. In FIDIA Research Foundation
Neuroscience Award Lectures. vol. 4. Raven Press, New York,
pp. 229-268.

Branchek, T. A., K. E. Smith, C. Gerald, and M. W. Walker

(2000). Galanin receptor subtypes. Trends. Pharmacol. Sci.
21, 109-116.

Campbell, I. L. (1998). Transgenic mice and cytokine actions

in the brain: bridging the gap between structural and
functional neuropathology. Brain Res. Brain Res. Rev. 26,
327-336.

Civelli, O., R. K. Reinscheid, and H. P. Nothacker (1999).

Orphan receptors, novel neuropeptides and reverse
pharmaceutical research. Brain Res. 848, 63-65.

Elmquist, J. K., C. F. Elias, and C. B. Saper (1999). From

lesions to leptin: Hypothalamic control of food intake and
body weight. Neuron 22, 221-232.

Hokfelt, T., C. Broberger, Z. Q. Xu, V. Sergeyev, R. Ubink, and

M. Diez (2000). Neuropeptides-an overview.
Neuropharmacology 39, 1337-1356.

Hokfelt, T., M.-N. Castel, P. Morino, X. Zhang, and A.

Dagerlind (1994). General overview of neuropeptides. In
Psychopharmacology: The Fourth Generation of Progress (F.
E. Bloom and D. J. Kupfer, eds.). Raven Press, New York, pp.
483-492.
Kask, K., M. Berthold, and T. Bartfai (1997). Galanin
receptors: involvement in feeding, pain, depression and
Alzheimer's disease. Life Sci. 60, 1523-1533.

Poyner, D., H. Cox, M. Bushfield, M. J. Treherne, and M. K.

Demetrokopoulos (2000). Neuropeptides in drug research.
Prog. Drug Res. 54, 121-149.

Oxytocin and vasopressin

Alescio-Lautier, B. and B. Soumireu-Mourat (1998). Role of
vasopressin in learning and memory in the hippocampus.
Prog. Brain Res. 119, 101-119.

Bohus, B. and D. de Wied (1998). The vasopressin deficient

Brattleboro rats: a natural knockout model used in the search
for CNS effects of vasopressin. Prog. Brain Res. 119, 555-
573.

de Wied, D. (1983). The importance of vasopressin memory.

Trends Neurosci. 7, 62-64.

Gash, D. M. and G. J. Thomas (1983). What is the importance

of vasopressin in memory processes? Trends Neurosci. 6,
197-198.

Jirikowski, G. F., P. P. Sanna, D. Macjiewski-Lenoir, and F. E.

Bloom (1992). Reversal of diabetes insipidus in Brattelboro
rats: intrahypothalamic injection of vasopressin mRNA.
Science 255, 996-998.

Koob, G. F., R. Dantzer, F. Rodriguez, F. E. Bloom, and M. Le

Moal (1985). Osmotic stress mimics the effects of vasopressin
on learned behaviour. Nature 315, 750-752.

Neurotensin
Carraway, R. and S. E. Leeman (1975). The amino acid
sequence of a hypothalamic peptide, neurotensin. J. Biol.
Chem. 250, 1907-1912.

Kinkead, B., E. B. Binder, and C. B. Nemeroff (1999). Does

neurotensin mediate the effects of antipsychotic drugs? Biol.
Psychiatry 46, 340-351.

Kobayashi, R. M., M. Brown, and W. Vale (1977). Regional

distribution of neurotensin and somatostatin in rat brain.
Brain Res. 126, 584-590.

Vincent, J. P., J. Mazella, and P. Kitabgi (1999) Neurotensin

and neurotensin receptors. Trends Pharmacol. Sci. 20, 302-
309.

Tachykinin peptides
Maggi, C. A. (2000). The troubled story of tachykinins and
neurokinins. Trends. Pharmacol. Sci. 21, 173-175.

Maggio, J. E. (1988). Tachykinins. Annu. Rev. Neurosci. 11,

13-28.

Nawa, H., H. Kotani, and S. Nakanishi (1984). Tissue specific

generation of two preprotachykinin mRNAs from one gene by
alternative RNA splicing. Nature 312, 729-734.

Nicoll, R. A., C. Schenker, and S. E. Leeman (1980).

Substance P as a transmitter candidate. Annu. Rev. Neurosci.
3, 227-268.

Rupniak, N. M. J. and M. S. Kramer (1999). Discovery of the

antidepressant and anti-emetic efficacy of substance P
receptor (NK1) antagonists. Trends. Pharmacol. Sci. 20, 485-
490.
Vanden Broeck, J., H. Torfs, J. Poels, W. Van Poyer, E.
Swinnen, K. Ferket, and A. De Loof (1999). Tachykinin-like
peptides and their receptors. A review. Ann. N.Y. Acad. Sci.
897, 374-387.

VIP-related peptides
Dicicco-Bloom, E., N. Lu, J. E. Pintar, and J. Zhang (1998).
The PACAP ligand/receptor system regulates cerebral cortical
neurogenesis. Ann. N.Y. Acad. Sci. 865, 274-289.

Guillemin, R., P. Brazeau, P. Bohlen, F. Eseh, N. Ling, and W.

B. Wehrenberg (1982). Growth hormone releasing factor from
a human pancreatic tumor that caused acromegaly. Science
218, 585-587.

Hosoya, M., H. Onda, K. Ogi, Y. Masuda, Y. Miyamoto, T.

Ohtaki, H. Okazaki, A. Arimura, and M. Fujino (1993).
Molecular cloning and functional expression of rat cDNAs
encoding the receptor for pituitary adenylate cyclase
activating polypeptide (PACAP). Biochem. Biophys. Res.
Commun. 194, 133-143.

Kopin, A. S., M. B. Wheeler, and A. B. Leiter (1990). Secretin:

structure of the precursor and tissue distribution of the
mRNA. Proc. Natl. Acad. Sci. U.S.A. 87, 2299-2303.

Lundberg, J. M., B. Hedlund, and T. Bartfai (1982). Vasoactive

intestinal polypeptide enhances muscarinic ligand binding in
cat submandibular gland. Nature 295, 147-149.

Magistretti, P. J., J. R. Cardinaux, J. L. Martin (1998). VIP and

PACAP in the CNS: regulators of glial energy metabolism and
modulators of glutamatergic signaling. Ann. N.Y. Acad. Sci.
816, 213-225.
Matsumoto, Y., M. Tsuda, and M. Fujino (1993). Regional
distribution of pituitary adenylate cyclase activating
polypeptide (PACAP) in the rat central nervous system as
determined by sandwich-enzyme immunoassay. Brain Res.
602, 57-63.

Peyron, C., D. K. Tighe, A. N. van den Pol, L. de Lecea, H. C.

Heller, J. G. Sutcliffe, and T. S. Kilduff (1998). Neurons
containing hypocretin (orexin) project to multiple neuronal
systems. J. Neurosci. 18, 9996-10015.

Rivier, J., J. Spiess, M. Thorner, and W. Vale (1982).

Characterization of a growth hormone releasing factor from a
human pancreatic islet tumor. Nature 300, 276-278.

Tatemoto, K. and V. Mutt (1981). Isolation and

characterization of the intestinal peptide porcine PHI (PHI-
27), a new member of the glucagon-secretin family. Proc.
Natl. Acad. Sci. U.S.A. 78, 6603-6607.

Yung, W. H., P. S. Leung, S. S. Ng, J. Zhang, S. C. Chan, and

B. K. Chow (2001). Secretin facilitates GABA transmission in
the cerebellum. J. Neurosci. 21, 7063-7068.

Pancreatic polypeptide-related peptides

Allen, Y. S., T. E. Adrian, J. M. Allen, K. Tatemoto, T. J. Crow,
S. R. Bloom, and J. M. Polak (1983). Neuropeptide Y
distribution in the rat brain. Science 221, 877-879.

El Majdoubi, M., A. Sahu, S. Ramaswamy, and T. M. Plant

(2000). Neuropeptide Y: a hypothalamic brake restraining
onset of puberty in primates. Proc. Natl. Acad. Sci. U.S.A. 97,
6179-6184.

Gehlert, D. R. (1999). Role of hypothalamic neuropeptide Y in

feeding and obesity. Neuropeptides 33, 329-338.
Hokfelt, T., C. Broberger, X. Zhang, M. Diez, J. Kopp, Z. Xu, M.
Landry, L. Bao, M. Schalling, J. Koistinaho, S. J. DeArmond, S.
Prusiner, and J. Gong (1998). Neuropeptide Y: some
viewpoints on a multifaceted peptide in the normal and
diseased nervous system. Brain Res. Brain Res. Rev. 26, 154-
166.

Mutt, V. (1983). New approaches to the identification and

isolation of hormonal polypeptides. Trends Neurosci. 6, 357-
360.

Wahlestadt, C. and M. Heilig (1994). Neuropeptide Y and

related peptides. In Neuropsychopharmacology: The Fourth
Generation of Progress(F. E. Bloom and D. J. Kupfer, eds.).
Raven Press, New York, pp. 543-551.

Wahlestedt, C., E. M. Pich, G. F. Koob, F. Yee, and M. Heilig

(1993). Modulation of anxiety and neuropeptide Y-Y1
receptors by antisense oligodeoxynucleotides. Science 259,
528-531.

Opioid peptides
Akil, H., C. Owens, H. Gutstein, L. Taylor, E. Curran, and S.
Watson (1998). Endogenous opioids: overview and current
issues. Drug Alcohol Depend. 51, 127-140.

Darland, T., M. M. Heinricher, and D. K. Grandy (1998).

Orphanin FQ/nociceptin: a role in pain and analgesia, but so
much more. Trends. Neurosci. 21, 215-221.

Goldstein, A., S. Tachibana, L. I. Lowney, and L. Hold (1979).

Dynorphin-(1-13), an extraordinarily potent opioid peptide.
Proc. Natl. Acad. Sci. U.S.A.76, 6666-6670.

Hughes, J., T. W. Smith, H. W. Kosterlitz, L. A. Fothergill, G. A.

Morgan, and H. R. Morris (1975). Identification of two related
pentapeptides from the brain with potent opiate agonist
activity. Nature 258, 577.

Nakanishi, S., A. Inoue, T. Kita, A.C.Y. Chang, S. Cohen, and

S. Numa (1979). Nucleotide sequence of cloned cDNA for
bovine corticotropin-lipotropin precursor. Nature 257, 238-
240.

Nicoll, R. A., G. R. Siggins, N. Ling, F. E. Bloom, and R.

Guillemin (1977). Neuronal actions of endorphins and
enkephalins among brain regions: a comparative
microiontophoretic study. Proc. Natl. Acad. Sci. U.S.A. 74,
2584.

Reinscheid, R. K., A. Ardati, F. J. Monsma, Jr., and O. Civelli

(1996). Structure-activity relationship studies on the novel
neuropeptide orphanin FQ. J. Biol. Chem. 271, 14163-14168.

Trujillo, K. A. and H. Akil (1994). Inhibition of opiate tolerance

by noncompetitive N-methyl-D-aspartate receptor
antagonists. Brain Res. 633, 178-188.

Wagner, J. J. and C. Chavkin (1994). Neuropharmacology of

endogenous opioid peptides. In Psychopharmacology: The
Fourth Generation of Progress (F. E. Bloom and D. J. Kupfer,
eds.). Raven Press, New York, pp. 519-529.

Somatostatin
DeLecea, L., J. R. Criado, O. Propsero-Garcia, K. M. Gautvik,
P. Schweitzer, P. E. Danielson, C. L. M. Dunlop, G. R. Siggins,
S. J. Henriksen, and J. G. Sutcliffe (1996). A cortical
neuropeptide with neuronal depressant and sleep-modulating
properties. Nature 381, 242-245.

de Lima, A. D. and J. H. Morrison (1989). An ultrastructural

analysis of somatostatin-immunoreactive neurons and
synapses in the temporal and occipital cortex of the macaque
monkey. J. Comp. Neurol. 283, 212-227.

Effendic, S. and R. Luft (1980). Somatostatin: a classical

hormone, a locally active polypeptide, and a neurotransmitter.
Ann. Clin. Med. 12, 8794.

Moore, S. D., S. G. Madamba, M. Joels, and G. R. Siggins

(1988). Somatostatin augments the M-current in hippocampal
neurons. Science 239, 278-280.

Patel, Y. C. (1999). Somatostatin and its receptor family.

Front. Neuroendocrinol. 20, 157-198.

Selmer, I., M. Schindler, J. P. Allen, P. P. Humphrey, and P. C.

Emson (2000). Advances in understanding neuronal
somatostatin receptors. Regul. Pept. 90, 1-18.

Vezzani, A. and D. Hoyer (1999). Brain somatostatin: a

candidate inhibitory role in seizures and epileptogenesis. Eur.
J. Neurosci. 11, 3767-3776.

Somatostatin
Baile, C. A., C. L. McLaughlin, and F.M.A. Della (1986). Role of
cholecystokinin and opioid peptides in control of food intake.
Physiol. Rev. 66, 172-234.

Chambers, M. S. and S. R. Fletcher (2000). CCK-B

antagonists in the control of anxiety and gastric acid
secretion. Prog. Med. Chem. 37, 45-81.

Fink, H., A. Rex, M. Voits, and J. P. Voigt (1998). Major

biological actions of CCK-a critical evaluation of research
findings. Exp. Brain Res. 123, 77-83.
Lotti, Y. J., R. G. Pendleton, R. J. Gould, H. M. Hanson, R. S.
Chang, and B. V. Clineschmidt (1987). In vivo pharmacology
of L-364,718, a new potent nonpeptide peripheral
cholecystokinin antagonist. J. Pharmacol. Exp. Ther. 241,
103-109.

Noble, F. and B. P. Roques (1999). CCK-B receptor: chemistry,

molecular biology, biochemistry and pharmacology. Prog.
Neurobiol. 58, 349-379.

White, F. J. and R. Wang (1984). Interactions of

cholecystokinin octapeptide and dopamine on nucleus
accumbens neurons. Brain Res. 300, 161-166.

Wiesenfeld-Hallin, Z., G. de Arauja Lucas, P. Alster, X. J. Xu,

and T. Hokfelt (1999). Cholecystokinin/opioid interactions.
Brain Res. 848, 78-89.

Corticotropin-releasing hormone
Behan, D. P., E. B. De Souza, P. J. Lowry, E. Potter, P.
Sawchenko, and W. W. Vale (1995). Corticotropin releasing
factor (CRF) binding protein: a novel regulator of CRF and
related peptides. Front. Neuroendocrinol. 16, 362-382.

Hernandez, J. F., W. Kornreich, C. Rivier, A. Miranda, G.

Yamamoto, J. Andrews, Y. Tache, W. Vale, and J. Rivier
(1993). Synthesis and relative potencies of new constrained
CRF antagonists. J. Med. Chem. 36, 2860-2867.

Koob, G. F. and S. C. Heinrichs (1999). A role for corticotropin

releasing factor and urocortin in behavioral responses to
stressors. Brain. Res. 848, 151-162.

Perrin, M. H. and W. W. Vale (1999). Corticotropin releasing

factor receptors and their ligand family. Ann. N.Y. Acad. Sci.
885, 312-328.
Stenzel-Poore, M. P., S. C. Heinrichs, S. Rivest, G. F. Koob,
and W. W. Vale (1994). Overproduction of corticotropin-
releasing factor in transgenic mice: a genetic model of
anxiogenic behavior. J. Neurosci. 14, 2579-2584.

Vale, W., J. Spiess, J. Rivier, and C. Rivier (1981).

Characterization of a 41-residue ovine hypothalamic peptide
that stimulates secretion of corticotropin and beta-endorphin.
Science 213, 1394-1397.

(+/-) Show / Hide Bibliography

Send Feedback Teton Server (5.0.0) - ©2007 Teton Data Title Updates
Customer Service Systems User
800.901.5494 Send Us Your Comments Responsibilities
Training Center
What's New
12. Cellular Mechanisms in Learning
and Memory
INTRODUCTION
Virtually every neurotransmitter system presented in this
book is associated with some specific behavioral effects
elicited by its agonists, antagonists, and often genetic
manipulation of its receptors. A far more difficult question is
just what role any given transmitter system plays in initiating,
maintaining, or terminating the behavior. In this chapter, we
consider an issue that represents one of the most complex
behavioral achievements of nervous systems, the ability to
learn and remember.

Substantial progress has been achieved over the last two

decades in understanding the molecular and cellular basis of
neuronal interactions, in particular the role of specific
neurotransmitters engaged in the regulation of sleep and
waking, appetite, and dependence on drugs of abuse. The
same sorts of knowledge are being achieved with regard to
the circuitry and molecular mechanisms underlying learning
and remembering. Although we cannot yet provide inquiring
minds with certified maps to today's Holy Grail in drug
development, the cognitive enhancer designed to delay or
reverse the ravages of Alzheimer's disease, there are now
definite leads.

To develop a convincing basis for the neuropharmacology of

learning and memory requires data establishing that a change
in a specific neurotransmitter action at a specific synaptic
location is necessary and sufficient for a behavioral change in
a living organism. Until relatively recently, such data were
either cellular or behavioral but it was not possible to study
the same synaptic connections in awake, behaving
experimental animals. However, that is no longer the case.
Moreover, a wide range of observations combining behavioral,
molecular, and cellular information in invertebrates and
vertebrates has revealed that each species, despite obvious
differences in neuronal circuitry, shares some cellular and
molecular features in its memory storage functions.

FORMULATING EXPERIMENTS IN MEMORY

AND LEARNING
Biological analysis of memory and learning typically requires
two parallel lines of work: (1) a behavioral component in
which the subject is trained, given a training recess of
variable intervals, and then tested for retention of the trained
response and (2) a functional, structural, or biochemical
component intended to define the essential changes, sites, or
molecular sequelae underlying these events.
Neuropharmacological probes offer excellent tools for the
manipulation of molecular events and their behavioral
outcomes.

Scientists working in this field commonly categorize learning

into "non-associative" and "associative" forms. Nonassociative
learning describes the functional changes that ensue when an
organism interacts with a single form of stimulus. For
example, the decreased response that occurs when the same
sensory stimulus is repeated sequentially, termed habituation,
is a nonassociative form of learning, as is the reversal of the
habituated response, termed sensitization, by a different
intensity or quality of sensory stimulus. Associative learning,
as the name implies, is the form of learning in which the
subject associates a previously neutral stimulus with a
response normally generated by another cue, the sort made
famous by Pavlov's classic experiments, which triggered dog
salivation with a bell instead of raw meat.
An exciting prospect for such work has been opened by the
development of experimental test systems in invertebrates,
and more recently in vertebrates, to show that the behavioral
performance of the organism depends on synaptic events in
specific locations and then to define the changes that occur at
these locations as the organism learns and performs. Such
possibilities for molecular and cellular explanations exist, it
should be recognized, only because of the extensive data that
have been obtained over this same interval.

The great neurocytologist Ramon y Cajal held strongly to the

view that learning was a synaptic event. He regarded the
synapse as the primary site of interneuronal communication.
The subsequent advances in biochemistry, anatomy, and
physiology that are emphasized in this text led to the notion
that drugs should regulate the critical sites of synaptic
transmission. Given this premise, the manipulations of
memory processes by specific drugs used as tools for
determining the roles played by specific transmitters and
defined sites should not only illuminate the underlying
process but also lead to eventual therapeutic interventions.

CELLULAR MODELS OF LEARNING IN

INVERTEBRATE NERVOUS SYSTEMS
Studies carried out in the visceral ganglia of invertebrates,
particularly the marine mollusc Aplysia, have provided
information on the molecular mechanisms of the cellular and
synaptic changes that accompany simple forms of behavior
modification, such as habituation and sensitization.
Sensitization is the ability of nonspecific but strongly arousing
sensory stimulation to enhance the effectiveness of synaptic
transmission in other neuronal pathways. When the gill-
withdrawal reflex is studied for this effect, strong electrical
stimulation of the connections between the visceral ganglion
and the head ganglion facilitates the withdrawal. The strong
stimulation also diminishes habituation of the reflex if tested
with identical repetitive sensory stimuli. As judged by analysis
of postsynaptic potentials, the facilitating stimulation
increases the amount of transmitter released by the afferent
limb of the withdrawal reflex (i.e., the sensory nerves). Once
initiated in the Aplysia ganglion, the effects of sensitization
can last from several minutes to hours.

Further pursuit of the presynaptic mechanism of the

enhanced release became possible when it was found that the
sensitizing stimulation also increased the ganglionic content
of cAMP and that exposure of the ganglion both to serotonin
(known to activate cAMP production in the ganglion) and to
cAMP (applied to the whole ganglion or injected intracellularly
into the sensory nerve cell) replicated the effects of the
sensitizing stimulation. In this case, the molecular mediation
sequence of the second-messenger hypothesis (see Chapter
6) suggests that a serotonin-secreting interneuron is
activated by the sensitizing stimulation, leading to increased
presynaptic levels of cAMP and consequent enhancement of
the sensory transmitter release.

Subsequent studies have shown that the short-term memory

for sensitization of the Aplysia gill- and siphon-withdrawal
reflex is distributed across at least four sites of circuit
modification. Each of these invokes a slightly different
circuitry and a different type of synaptic modification
(presynaptic facilitation, presynaptic inhibition, posttetanic
potentiation, and increased tonic firing rate), all of which
result in facilitation. Nevertheless, all four of these
mechanisms seem to be mediated by a common modulatory
transmitter (serotonin), to utilize the same cAMP-mediated
second-messenger system, and to require a phosphorylation
event (affecting either intravesicular proteins or ion channels)
to enhance transmitter release. These short-term changes,
lasting from minutes to hours, do not require protein
synthesis.
Kandel and associates have extended their definition of the
molecular mechanisms involved in the long-term changes in
the plasticity underlying the gill-withdrawal reflex. Having
established that such long-term functional changes are indeed
dependent on gene transcription and protein synthesis (while
the short-term facilitative changes are not), several of the
early and late genes involved specifically in the synaptic
events have been defined. As with the short-term changes,
cAMP generation appears to be a key step in long-term
facilitation, leading to both protein and structural synaptic
changes. In the most recent phases of this work, attention
has been focused on the cAMP response element-binding
protein (CREB), in which it can be shown by gene transfer
experiments on cultured neurons that cumulative responses
to 5-hydroxytryptamine (5-HT) lead to cAMP synthesis,
activation of the cAMP-dependent protein kinase (PKA),
translocation of the activated enzyme to the nucleus, and
phosphorylation of serine-119 on CREB. By reducing the
preparation to pairs of interconnected neurons in culture, it
has been possible to determine that as few as five pulses of
5-HT are sufficient to engage the long-term changes.

The phosphorylated CREB presumably activates the

transcription of immediate early genes, one of which has been
specifically incriminated as a key player. Both 5-HT and cAMP
rapidly induce the Aplysia CCAAT enhancer-binding protein
(ApC/EBP), even in the presence of protein synthesis
inhibitors. Furthermore, immunologically blocking the function
of ApC/EBP blocks long-term facilitation selectively without
affecting short-term facilitation. The switch from short- to
long-term facilitation requires not only activation of CREB1
but also reversal of the tonic repression of of the Aplysia form
of a second CREB, known as ApCREB2, as well as induction of
ApC/EBP. A transcription activator, Aplysia A Factor, which is
stimulated by PKA has been identified. This protein can
dimerize with both ApC/EBP and ApCREB2 and appears to be
necessary for the long-term facilitation that can be induced by
as little as five pulses of serotonin. Present evidence suggests
that this effect is mediated by either activation of CREB1 or
derepression of ApCREB2. Overexpression of ApAF enhances
the long-term facilitation further. Thus, ApAF is a candidate
memory enhancer gene downstream from both CREB1 and
ApCREB2.

These two themes, namely, that the short-term and long-term

changes rely on different consequences of the 5-HT activation
of cAMP synthesis and that the functional consequences of
specific immediate early gene systems are important, also
epitomize recent work in mammalian models of learning and
plasticity.

SOME GENERAL PRINCIPLES

Kandel, Tauc, Gershenfeld, Strumwasser, Alkon, Levitan,
Kaczmarek, and other explorers of invertebrate psychobiology
launched the scientific search for the cellular and molecular
bases of learning. They thought that these smaller nervous
systems could be fathomed, and they set themselves the task
of working out the rules of connectivity by which these
systems adapt to environmental conditions. Their meticulous
descriptive work has given us a logical, consistent, and
compelling account of the cellular changes accompanying
behavioral modification and the molecular basis of these
changes.

The monoamine hypothesis of memory and learning in

mammals advanced by Kety in the late 1960s could now be
viewed as anticipating, in a more primitive mechanistic way,
the serotonin and cAMP-mediated presynaptic facilitation
model derived from the studies of invertebrate sensitization.
Kety's model called for a brain system that could mediate the
ability of arousal to consolidate experiences into adaptive
behavioral mechanisms. To reconcile the two situations, we
must first substitute the mammalian transmitter
norepinephrine (NE) for the role played by 5-HT in the
invertebrate. The central NE neurons and their synapses have
now been well mapped, and electrophysiological recordings in
behaviorally responding rats and monkeys have established
that the NE cells fire when novel sensory events occur in the
external environment. These data are logical and internally
consistent with a role of NE neurons in mediating arousal.
However, there are clearly many more neurotransmitter
systems at work in mammalian mechanisms of learning and
memory, especially when one begins to focus on discrete
synaptic changes within precisely constrained regions of the
brain in vitro or in vivo. We will examine some of these
specific synaptic sites below. Learning and memory events in
mammalian brain function-as in the invertebrate-are probably
too critical to rely on only one transmitter system. In fact, if
we substitute dopamine for NE in the above discussion, we
have two-thirds of the work that won the Nobel Prize for
Physiology or Medicine in 2000.

THE RABBIT NICTITATING MEMBRANE

REFLEX AND ASSOCIATIVE LEARNING
The rabbit's nictitating membrane has provided another
molecular and cellular model for memory and associative
learning. This mammalian neuronal system permits the direct
observation of an animal's performance during tests of
learning and memory and the direct determination of exactly
where in the circuitry the learning events have occurred.

The basic studies, largely done independently by John

Harvey's and Richard Thompson's groups, showed that the
rabbit blink reflex could be associatively conditioned. Rabbits
will normally blink (i.e., cover their cornea with the nictitating
membrane) every time a puff of air is directed at the cornea.
Repeated presentation of a loud tone just before the air puff
conditioned the rabbits to blink in response to the tone.
Eventually, the rabbits blinked every time they heard the
tone, with no air puff being required.

At this point, chemical and electrolytic lesions as well as

micro-stimulations were used to define the pathways that
relay the unconditional (air puff-induced) blink and the
conditional blink triggered by the tone (see Fig. 12-1). The air
puff stimulus travels through the sensory fibers of the
trigeminal nerve to activate motor neurons of the abducens
nucleus, causing the eye to retract into the orbit and then the
nictitating membrane to extend over the eye. The acoustic
stimulus travels through the cochlear nucleus to the pontine
nucleus and seems quite separate at its early stages from the
corneal sensory pathway. However, the researchers also noted
that even before the rabbits learned to associate the sound
with the blinking, the acoustic stimulus, which was rather
loud, could often increase the response amplitude to the
normal air puff stimulus. This suggested that the acoustic
stimulus could also sensitize the unassociated air puff
stimulus.

Based on comparisons of rabbits in which discrete lesions

were made along the known anatomical paths within the two
systems of sensory processing, the critical site for producing
the association seems to be located in the interpositus
nucleus, one of the deep cerebellar nuclei. The efferent limb
of the motor reflex then courses through the red nucleus
directly to the abducens nerve, where it activates the eyeball
retraction reflex (see Fig. 12-1).

Since most of the transmitters operating in these hierarchical

controlling chains have not yet been defined, it would seem
that the model was not quite ready for transmitter-specific
pharmacology to be a useful analytical tool. The only sites
that have been probed so far are the synapses within the
interpositus and red nuclei; here, the -aminobutyric acid
(GABA) antagonists bicuculline and picrotoxin block the
conditional reflex. Chemical lesions of the inferior olivary
nucleus (the source of the climbing fiber projection to
cerebellar Purkinje neurons) completely disrupt the
acquisition and retention of the conditional reflex. Since
glutamate is presumed to be the transmitter of this pathway,
it is implicated as the transmitter for this final step.
Inactivation of the deep cerebellar nuclei by local anesthetics
abolished eyeblink conditioned responses whether the
conditional stimulus was a tone or stimulation in the lateral
reticular nucleus. Inactivation of the lateral pontine nucleus
prevented only the acquisition and retention of tone-evoked
eyeblink conditioned responses. Using multiple-unit
recordings in the lateral pontine nucleus, Thompson and
colleagues demonstrated that when lateral reticular nucleus
stimulation was used as the conditional stimulus, inactivation
of the interpositus nucleus abolished learning-related
neuronal activity, whereas inactivation of the pontine nucleus
had little effect on similar activity in the interpositus nucleus.
They concluded that the learning-induced neuronal activity in
the lateral pontine nucleus was most likely driven by the
cerebellar interpositus nucleus.

Harvey and colleagues have used the associative eye blink

model to show that low doses of amphetamine as well as
lysergic acid diethylamide (LSD) enhance this form of
learning, while a wide range of drugs from pimozide to
atropine and scopolamine, as well as agonists for and
opioid receptors, decrease acquisition at doses that do not
affect the basic reflex. Because these latter drugs were given
systemically, however, it was not clear where (within the
cranial nerve-cerebellar circuitry or elsewhere) they acted to
produce these effects.

In subsequent experiments, forebrain regions traditionally

viewed as participating in memory and learning functions
have also been implicated in the nictitating membrane
associational reflex. For these experiments, rabbits received
"classic conditioning" of the nictitating membrane response;
the tone stimulus and the air puff were separated by a pause
(or "trace") of 500 milliseconds. The short interruption made
the simple learning task a bit more complicated. The extra
time delay allowed the investigators to distinguish nonspecific
arousal, possibly sensitizing the reflex, from more specific
interactions.

Thompson and colleagues found that lesions of the

hippocampus or cingulate/retrosplenial cortex disrupted
acquisition of the interrupted conditioned response but that
neocortical lesions did not. Neither lesion affected acquisition
of the noninterrupted pairing of tone and air puff. When
animals with hippocampal or cingulate/retrosplenial cortex
lesions were switched to a standard delay paradigm in which
the conditional stimulus and the unconditional stimulus were
given at the same time, rabbits acquired the association in
about the same number of trials as naive animals.

In the interrupted protocol, however, macroelectrode

recordings showed substantial increases of multiunit activity
in the CA1 region of the hippocampus, which began during
the tone and persisted through the trace interval, even before
the rabbit showed consistent association. As the rabbit's
performance in responding to the tone stimulus improved, the
hippocampal activity shifted to later in the trace interval.
Although the transmitter has not been identified for this
activation, it is a safe presumption that GLU is involved; in
fact, changes in AMPA subtype receptor binding, but not in
NMDA receptor kinetics, have been reported to accompany
the learning.

These data thus directly relate behavioral demonstration of

associational learning with discrete neuronal circuitry. Such an
experimental system should eventually lend itself to
pharmacological and electrophysiological analyses of the
changes within the critical neurons that alter synaptic
function. Student volunteers are welcome.

Figure 12-1. Schematic diagram of the sensory

and motor circuitry for the rabbit nictitating
membrane model of cellular mechanisms in
associative learning. Pathways marked as U.S.
participate in the elicitation of the unconditional
reflex response (U.R. components) in which a puff
of air to the cornea will always provoke a blink by
closing the nictitating membrane. The auditory
pathways that carry the tone information that can
become associated with the air puff-induced blink
are indicated as the C.S. (conditional stimulus) or
C.R. (conditional response) components. When the
animal has been well trained, the tone C.S. will
evoke the C.R. even without a U.S.

LONG-TERM POTENTIATION: A VERTEBRATE

MODEL OF SYNAPTIC PLASTICITY
Next, we turn to an increasingly popular form of mammalian
forebrain synaptic plasticity, in which electrophysiological
changes at the synaptic level have been probed extensively
by drugs to identify the transmitter systems responsible for
discrete transductive mechanisms. The term long-term
potentiation (LTP) refers to long-lasting enhancement of
synaptic transmission (10 minutes to days, depending
critically on the conditions used to evoke the response and
where it is tested). The effect is measured as increased
amplitudes of excitatory postsynaptic potentials (or the
currents generated by these potentials) in specific circuits
after high-frequency, high-intensity activation of the same
circuits or other discrete paths. Because the phenomenon was
originally described with macroelectrode recordings in the
hippocampus of intact animals, it has been a candidate to link
learning and cellular changes in vivo. However, the most
intense studies have more often involved analysis of the
pertinent connections in vitro in the now classic slice
preparations of the hippocampus or neocortex in normal
animals or transgenic animals with specific genes mutated.

This special form of enhanced synaptic transmission was first

demonstrated by brief, high-frequency stimulation of the
entorhinal cortex, through the perforant path (see Fig. 12-2),
to enhance activation of the granule cells of the hippocampal
dentate gyrus by subsequent single stimulation of the
perforant path. In awake cats, guinea pigs, and rats, this
enhanced transmission could be seen for periods of days to
weeks. The transmitter for this pathway is now thought to be
glutamate (Glu) or possibly aspartate (Asp), although a role
of at least one family of coexisting peptides, such as the
proenkephalin-derived opioid peptides, has been reported
(see Chapter 11).

LTP has also been observed at two other sites within the
hippocampal formation: (1) at the synapses between the
dentate granule cells and the targets of their mossy fiber
synapses, the CA3 pyramidal neurons, and (2) between the
Schaffer collaterals of CA3 pyramidal neurons, as well as
associational fibers from the contralateral hippocampus, and
the CA1 pyramidal neurons (see Fig. 11-3). The transmitter
for these two pathways is thought to be Asp, Glu, or possibly
homocysteate; coexisting peptides (prodynorphin-derived as
well as cholecystokinin) remain candidates for some part of
the action here, at least for the mossy fiber connection to CA3
pyramidal neurons. Although the basic phenomenon of LTP
after a high-frequency, high-intensity period of afferent
pathway activation is similar at all three synaptic sites, the
relative importance of specific transmitters and their
responsible transductive mechanisms differ on a site-by-site
basis.
The essential pharmacological advance came with the
recognition that the NMDA subtype of excitatory amino acid
(EAA) receptor (see Chapter 7) was responsible for the
induction phase of the long-lasting synaptic enhancement.
Antagonists of this form of EAA receptor (e.g.,
aminophosphovaleric acid) prevent LTP induction without
interfering with previously potentiated transmission or normal
low-frequency transmission in the pathway under study. The
critical link established by several groups is that the required
high-frequency, high-intensity stimulation acts via the NMDA
receptor to couple the depolarization with increased Ca2+
entry. This combined response is then thought to depolarize
the small dendritic domains that are the site of convergent
afferents in the pathway under study and to increase their
subsequent transmission. In the perforant path-to-dentate
circuit, drugs that block u-opioid receptors or -adrenergic
receptors also block LTP. The latter antagonists are also very
effective at blocking LTP of CA3 neurons, although not that of
CA1 neurons. These drug effects coincide nicely with the
regional patterns of monoamine innervation of the
hippocampal formation (see Fig. 11-2).

The NMDA receptor has been accepted as a critical step in

initiating LTP, as well as in a related intrahippocampal process,
long-term depression, that has so far received much less
scientific examination. The current evidence suggests that the
NMDA receptor must be in the appropriate state for LTP to be
induced and that a molecular switch triggered by the
metabotropic Glu receptor is required for the proper setting. A
minor glitch in validating this evolving consensus emerged
when disruption of the NMDA receptor NMDAR1 led to early
postnatal death of transgenic mice, with grossly maldeveloped
brains. However, subsequent efforts to knock out only one of
the NMDA subunits (the NR2 subunit family) were more
successful, leading to normal brain cellular structure and
normal growth and development but significantly reduced LTP
in the hippocampal CA1 subfield and reduced (albeit
moderately so) spatial learning. Disrupted LTP induction in the
transgenic mouse mutant fyn, in which one tyrosine kinase
isoform is disrupted (but not three others), and in mice
lacking the protein kinase C isoform indicates that very
specific protein phosphorylation can regulate the thresholds
for LTP induction. Despite the lack of LTP in the latter mice,
however, spatial learning deficits can be seen only under
rather extreme test conditions.

The NMDA-regulated step alone does not seem to be

adequate for the long-term maintenance of the enhanced
transmission. Given the critical role established in the Aplysia
and other invertebrate models for a cAMP-regulated process
in longer-term plasticity, that process has also been explored
in the hippocampal model. Again, transgenic mice were
employed, in this case expressing mutated knockout
constructs of the CREB and isoforms. These mice
responded normally to short-term conditional fear associative
learning but were unable to retain the learned fear for more
than 2 hours (which the wild-type mice could readily do).
Interestingly, LTP induction was also normal in the CREB
knockout mice but returned to baseline in 2 hours as well.
Lest the reader think it is now safe to memorize these
schemata, if not the underlying facts, we must deliver a
cautionary note. The CREB had been previously knocked out
in a couple of alternate ways, one leading to dwarf mice when
the CREB could not be phosphorylated and another with full
CREB knockout leading to no deficits at all. Attempts to
reconcile these observations led to the recognition that CREB
belongs to a family of transcription factor-regulatory proteins,
which can often compensate for each other. Thus, there is an
alternative interpretation of the LTP and memory problems of
the CREB knockout mice (pointed out to me by a clever young
molecular neurobiologist, who prefers anonymity until this
idea is tested): the knockout CREB should have been
compensated for by another of those transcription factors,
and it is perhaps that failure to compensate which causes the
problems. Life gets confusing, doesn't it?

In allowing the world to spin a few times between editions,

we may have spared the reader at least one apparently
abortive chapter in the LTP saga: nitric oxide (NO) found one
of its many overnight stardom roles as a potential retrograde
transmitter leading to long-term increases in presynaptic
transmitter release, based in part on the ability of nitric oxide
synthase (NOS) inhibitors to block LTP, at least at certain
nonphysiological temperatures. However, when LTP was found
to be unchanged in mice in which neuronal NOS had been
knocked out, enzymatic activity was virtually immeasurable,
and LTP could still be blocked by NOS inhibitors, covert side
effects of those blockers pulled NO from center stage. You can
bet this is not over by a long shot. Thus, much remains to be
accomplished with this system, including the demonstration
that such events within single pathways of the hippocampal
formation are necessary and sufficient for the animal or the
hippocampus to "learn" something.

Readers intrigued by these memory model synaptic systems

should consult the selected references for other models in
which the circuitry has been spelled out and awaits the
application of creative pharmacological analysis: the olfactory
memory pathways, long-term depression of hippocampal and
cerebellar units, and the amygdaloid complex.

Figure 11-2. Structural homologies between the

peptides of the tachykinin family, presented
according to the schema of Figure 11-1. pE,
pyroglutamate.

Figure 11-3. The vasoactive intestinal peptide

(VIP)-related peptide family represented by their
single-letter amino acid symbols. The sequences
of PHI-27, PHM-27, growth hormone-releasing
 hormone 1-24, glucagon, and secretin that match
those of VIP are indicated in bold letters. For an
interesting exercise, the reader may wish to
construct a complementary table in which matches
to glucagon are highlighted.

Figure 12-2. The synaptic circuits of the rodent

hippocampal slice preparation in which long-term
potentiation (LTP) can be elicited are indicated in
the upper panel, while the neurotransmitters
associated with these circuits or with the target
cells capable of expressing the LTP response are
shown in the lower panel. Each of the three major
hippocampal circuits (entorhinal cortex to dentate,
dentate to CA3, and CA3 to CA1) shows a
generally similar LTP, but the pharmacology of the
modification varies with the specific location and
pathways. Glutamate (GLU)-mediated synaptic
events figure prominently, as indicated in the text.

ARE THERE ANY NATURAL MEMORY DRUGS?

Back in the third edition, we noted the growing literature on
the ability of natural hormones such as vasopressin and
corticotropin, as well as "endocrinologically inert" fragments
derived from them, either to repair learning deficiencies in
hypophysectomized rats or to delay or accelerate the
extinction of a previously learned performance. Unfortunately,
as the pathways containing these peptides were more clearly
defined in their projections to targets other than the posterior
pituitary and the known barriers to diffusion of these peptides
from the bloodstream into the brain were shown to apply to
all of them, this once-promising area became a source of
contention. However, this body of research remains an
important case study for scholars of the neuropharmacology
of behavior. Extended analyses of the behavioral responses
regulated by systemic peptide hormones have demonstrated
that visceral afferent nerves can influence the brain even if
the peptides are excluded by the blood-brain barrier.
Furthermore, there is a strong possibility that secrets leading
to the development of memory-enhancing drugs may be
discovered among the intracerebral pharmacological effects of
what were once considered to be exclusively endocrine
peptides.

Therefore, despite the murky nature of this particular

research strategy, students should persist. For example,
several studies have shown that low doses of subcutaneously
administered vasopressin can enhance acquisition of a
behavioral response, such as maze running or extinction of
active avoidance in a shuttle box. Since vasopressin was
named for its potent ability to increase blood pressure, it is
not surprising that, while low in dose weight, the behaviorally
active systemic doses of vasopressin also cause systemic
hypertension. Moreover, vasopressin analogues that block the
peripheral vascular receptors will also block the memory
performance effects of the hormone. All behavioral effects of
vasopressin have been blocked by antagonists of the vascular
receptor, and the apparent central effect is likely to be
indirectly mediated by an arousal secondary to the
inappropriate hypertension. However, substituted and
truncated versions of vasopressin lacking any vascular activity
can still alter memory performance. Since the latter
demonstrations of effectiveness have been obtained with
memory tasks driven by aversive stimuli, their interpretation
is still open.

Regardless of the arguments as to how or where the

peripherally injected vasopressin acts, still smaller doses of
vasopressin (1/1000 the systemic dose) have similar effects
on memory performance when given intracerebroventricularly
or intracerebrally. These low-dose (nanogram or picomolar)
central effects are also reduced or prevented by central
administration of the peripheral vascular vasopressin
antagonist. Furthermore, opposite effects (i.e., memory
performance impairment) have been reported in some tasks
after central administration of vasopressin antagonists that
are thought to act at vascular receptors. This suggests that
somewhere within the central vasopressin-containing
pathways is a circuit whose release of vasopressin is critical in
eliciting proper performance and which has a receptor highly
similar to those on blood vessels. Whether this means that
the vascular receptor is also present on the central targets or
that the central targets are actually on the blood side of the
blood-brain barrier is an unanswered question. In fact,
without directly testing a role for vasopressin, Thompson and
colleagues have reported that water deprivation increased the
magnitude of hippocampal LTP (perforant path to dentate)
and, in parallel experiments, facilitated contextual fear
conditioning.

The concept of vasopressin as a simple natural memory

hormone would seem to be seriously challenged by studies of
the Brattleboro rat, whose genetic defect produces a defective
mRNA that cannot be translated into a secretable vasopressin.
Although initially reported as being memory-deficient, the
Brattleboro rat has supporters who regard its memory ability
as normal, especially when salt and water balance are
maintained by oral or systemic loading. Furthermore, humans
with diabetes insipidus (the human disease of vasopressin
deficiency) are not reported to have significant memory
impairment.

Without some rigorous identification of the precise cellular

sites and specific receptor transductive mechanisms on which
either blood-borne or cerebrospinal fluid-borne vasopressin
acts, this line of work faces difficulties. The possibility that
vasopressin-derived subfragments (whose informational value
has been suggested but never defined) can influence neuronal
operations critical to the acquisition or retention of the
memory tasks that have been studied remains ripe for further
study. Future research will probably establish the superficiality
of such interpretations as the following: (1) vasopressin acts
directly on memory processes and (2) vasopressin can be an
aversive hormone that, when given at nonphysiological doses,
arouses the animal, which then learns better. We await
eagerly the answers to this mind-drug-behavior puzzle, but
they may not be found in the next edition, either.

APPROACHING THE NEUROPHARMACOLOGY

OF HUMAN MEMORY
In the selected animal models reviewed here, several key
ingredients were required for the partial resolution of how
selective drugs could modify the process of memory and
learning: well-defined and relatively simple memory tasks,
well-defined neuronal circuits whose transmitters and
transmitter receptors were identified, and drugs that could be
administered in ways that restricted their access to the sites
being observed. An ultimate goal of such studies would be to
develop a specific neuropharmacology to improve human
memory function, either in normal subjects or in those in the
early stages of dementing illnesses such as Alzheimer's
disease or acquired immunodeficiency syndrome. Until
recently, almost none of the essential facts were in place.
Furthermore, the prospects of inferring from noninvasive
techniques which sites in a human's brain participated in
memory-related functions or how systemically administered
drugs with multiple sites of interaction affected these events
seemed dim not all that long ago.

However, dealing with memory processes in humans or even

nonhuman primates has become far more attractive, partly
because of the ability of methods such as positron emission
tomography to reveal very brief shifts in blood flow within the
brain while humans perform complex cognitive processing,
thereby indicating which locations were active with specific
kinds of higher mental activity. These methods, combined
with experimental neurosurgical approaches to the rapidly
unfolding description of cortical circuits in the monkey brain,
have helped clarify a long-studied area of human memory
research, namely, the amnesias that follow discrete toxic or
traumatic lesions of the human brain. These combined
strategies have allowed investigators to return to a point
made earlier in the chapter (namely, that memory functions
in more complex brains are likely to involve multiple
systems), to ask how the primate brain is organized for
memory functions, and to determine which neuronal systems
are involved.

As exciting as this work is, an interesting paradox has been

recognized: while very discrete lesions of the human
hippocampus can lead to profound anterograde amnesia, the
hippocampus is not among the many cortical sites in which
functional activation (i.e., blood flow or glucose utilization)
occurs when normal humans are performing complex
memory-forming or memory-testing tasks. However, selected
regions of the frontal cortex are activated in these tasks.

From this body of work has emerged the hypothesis that

information processing is partly tied to the specific processing
areas of the brain that are engaged during learning and that
these memory events are stored within the same neural
systems that also participate in the processes of perceiving,
analyzing, and processing sensory information. Thus, in the
visual system, the temporal cortical region involved in
analyzing complex visual patterns (like faces) seems also to
be involved in processing the information and in storing it.
These and other examples of complex sensory systems
suggest that some aspects of learning can be localized to
specific subtasks but that many parts of the brain are
involved with the overall process through intricate, parallel
neuronal circuits.
The effort to bolster short-term memory in aged rhesus
monkeys is an indication of future prospects. Memory
capacity was assessed by a moderately complex task that
required the animal to remember information over short time
intervals and to update this information on every trial.
Through impressive structure activity comparisons of a series
of 2-adrenergic agonists, two forms of memory-related
action could be defined: hypotensive, sedating, and memory-
impairing effects at one type of site and memory-enhancing
effects at another 2-receptor site. This same drug was also
found to improve performance of complex behaviors in young
monkeys as well. In low doses, the 2-adrenergic agonist
guanfacine improved memory without inducing hypotension
or sedation, offering hope for the treatment of at least some
memory disorders in humans. Guanfacine has also been
reported to improve the behavioral and learning problems of
children with attention-deficit hyperactivity disorder.

SELECTED REFERENCES
Abel, T., K. C. Martin, D. Bartsch, and E. R. Kandel (1998).
Memory suppressor genes: inhibitory constraints on the
storage of long-term memory. Science 279, 338-341.

Alkon, D. L. (1988). Memory Traces in the Brain. Cambridge

University Press, Cambridge.

Arnsten, A. F. (1993). Catecholamine mechanisms in age-

related cognitive decline. Neurobiol. Aging 14, 639-641.

Bao, S., L. Chen, and R. F. Thompson (2000). Learning- and

cerebellum-dependent neuronal activity in the lateral pontine
nucleus. Behav. Neurosci. 114, 254-261.

Bartsch, D., M. Ghirardi, A. Casadio, M. Giustetto, K. A. Karl,

H. Zhu, and E. R. Kandel (2000). Enhancement of memory-
related long-term facilitation by ApAF, a novel transcription
factor that acts downstream from both CREB1 and CREB2.
Cell 103, 595-608.

Baudry, M. and G. Lynch (1988). Properties and substrates of

mammalian memory systems. In Psychopharmacology: The
Third Generation of Progress (H. Y. Meltzer, ed.). Raven Press,
New York, pp. 449-462.

Bear, M. F. and R. C. Malenka (1994). Synaptic plasticity: LTP

and LTD. Curr. Opin. Neurobiol. 4, 389-399.

Bortolotto, Z. A., Z. I. Bashir, C. H. Davies, and G. L.

Collingridge (1994). A molecular switch activated by
metabotropic glutamate receptors regulates induction of long-
term potentiation. Nature 368, 740-743.

Gallagher, M. and P. C. Holland (1994). The amygdala

complex: multiple roles in associative learning. Proc. Natl.
Acad. Sci. U.S.A. 91, 11771-11776.

Haley, D. A., R. F. Thompson, and J. I. V. Madden (1988).

Pharmacological analysis of the magnocellular red nucleus
during classical conditioning of the rabbit nictitating
membrane response. Brain Res. 454, 131-143.

Harvey, J. A. (1987). Effects of drugs on associative learning.

In Psychopharmacology: The Third Generation of Progress (H.
Y. Meltzer, ed.). Raven Press, New York, pp. 1485-1491.

Ito, M. (1987). Long term depression as memory process in

the cerebellum. In Synaptic Function (G. M. Edelman, W. E.
Gall, and W. M. Cowan, eds.). Wiley-Interscience, New York,
pp. 431-447.

King, D. A., D. J. Krupa, M. R. Foy, and R. F. Thompson

(2001). Mechanisms of neuronal conditioning. Int. Rev.
Neurobiol. 45, 313-337.

Markowitsch, H. J. and E. Tulving (1994). Cognitive processes

and cerebral cortical fundi: findings from positron-emission
tomography studies. Proc. Natl. Acad. Sci. U.S.A. 91, 10507-
10511.

Mayford, M. and E. R. Kandel (1999). Genetic approaches to

memory storage. Trends Genet. 15, 463-470.

Mintz, M., D. G. Lavond, A. A. Zhang, Y. Yun, and R. F.

Thompson (1994). Unilateral inferior olive NMDA lesion leads
to unilateral deficit in acquisition and retention of eyelid
classical conditioning. Behav. Neural Biol. 61, 218-224.

Rampon, C. Y.-P. Tang, J. Goodhouse, E. Shimizu, M. Kyin, and

J. Z. Tsien (2000). Enrichment induces structural changes and
recovery from non-spatial memory deficits in CA1 NMDAR1-
knockout mice. Nat. Neurosci. 3, 238-245.

Richter-Levin, G., M. L. Errington, H. Maegawa, and T. V. Bliss

(1994). Activation of metabotropic glutamate receptors is
necessary for long-term potentiation in the dentate gyrus and
for spatial learning. Neuropharmacology 33, 853-857.

Scahill, L., P. B. Chappell, Y. S. Kim, R. T. Schultz, L.

Katsovich, E. Shepherd, A. F. Arnsten, D. J. Cohen, and J. F.
Leckman (2001). A placebo-controlled study of guanfacine in
the treatment of children with tic disorders and attention
deficit hyperactivity disorder. Am. J. Psychiatry 158, 1067-
1074.

Schacher, S., E. R. Kandel, and P. Montarolo (1993). cAMP and

arachidonic acid stimulate long-term structural and functional
changes produced by neurotransmitters in Aplysia sensory
neurons. Neuron 10, 1079-1088.
Tang, Y. P., E. Shimizu, G. R. Dube, C. Rampon, G. A.
Kerchner, M. Zhuo, G. Liu, and J. Z. Tsien (1999). Genetic
enhancement of learning and memory in mice. Nature 401,
63-69.

Williams, J. H., Y. G. Li, A. Nayak, M. L. Errington, K. P.

Murphy, and T. V. Bliss (1993). The suppression of long-term
potentiation in rat hippocampus by inhibitors of nitric oxide
synthase is temperature and age dependent. Neuron 11, 877-
884.

(+/-) Show / Hide Bibliography

Send Feedback Teton Server (5.0.0) - ©2007 Teton Data Title Updates
Customer Service Systems User
800.901.5494 Send Us Your Comments Responsibilities
Training Center
What's New
13. Treating Neurological and
Psychiatric Diseases
INTRODUCTION
By this stage, the reader will almost certainly have detected a
not-so-subtle underlying theme, namely, that a full
understanding of the molecular mechanisms by which healthy
neurons communicate and are sustained constitutes the
starting point for developing drugs to treat neurological and
psychiatric disorders. Except for a few passing references,
however, we have not said very much about the nature of the
brain's diseases or indicated those for which treatments are
now effective and why. In this final chapter, we provide a brief
introduction to those important neuropharmacological
motivations.

The modern neuropsychiatric therapeutic armamentarium

contains some powerful, selective, and highly effective drugs
that often save lives and restore function. With the few
exceptions noted below, however, most of these treatments
came about by refining drugs that were initially developed for
dubious disease hypotheses because astute clinicians
observed potentially useful central nervous system (CNS)
actions initially considered to be side effects. Many times, the
predecessors for today's drugs told us more about the nature
of the brain's normal operations by revealing drug-sensitive
functions that had not been previously recognized. Moreover,
several of the most effective drugs (e.g., those used to treat
complex diseases of thinking, emotion, movement, or
appetite) work for reasons that are still not very clear. In such
cases, pursuit of molecular explanations of the
pathophysiology of the disease continues to drive research.
The corollary is that the diseases that we understand the
most clearly in terms of causative mechanisms are not yet
very treatable.

One last confession before we begin: New drugs will

unquestionably emerge from the mountains of molecular
information now being compiled in the push to sequence the
entire human genome (see below). Still, the goal remains to
gain far more insightful views of pathogenesis and then to
exploit that knowledge for ways to treat and prevent disease.

WHEN IT'S ALL IN THE GENES

Introduction
In Chapter 3, we went through some of the fundamentals of
molecular genetics in order to define the basic properties of
the neurons and glia and their functions. To understand the
genetic basis of CNS disorders when there is one, we need to
revisit those concepts. The genome is the complete genetic
set of a given individual, established at the moment of
conception when the half-set of genes in a sperm combine
with the half-set in the lucky ovum. In humans, the cells of
the embryo contain 22 pairs of homologous autosomes and
one pair of sex chromosomes that are heterologous (XY) for
males and homologous (XX) for females. Each chromosome
contains many genes, each with its own specific location (its
locus) along the chromosome. Since each chromosome in the
pair has one copy (or allele) of each gene, a person can be
either homozygous (if both genes are the same) or
heterozygous for any given gene. One of the early surprises
from the human genome sequencing effort is that humans
and other mammals may have far fewer than the 50,000-
100,000 genes that had been previously conjectured, perhaps
only twice as many as the fruit fly, the most complex
organism of the previously sequenced genomes. If this
conclusion holds, one interpretation is that in order to create
organisms as complex as humans and other mammals from
relatively few genes probably means that the richness of the
required proteins is based on their modifications, either
during transcription of the gene by much more intense
splicing events or after translation of the mRNA into the
protein.

Small, nonfunctional differences in nucleotide sequences that

arise by mutation allow for some genes to contain
polymorphisms. Because these differences also change the
vulnerability of the DNA to cleavage by specific restriction
endonucleases, a given digestion protocol can produce DNA
fragments of different lengths in the alleles for a given gene.
These varying sequence patterns and their differential
cleavage patterns led to the method of restriction fragment
length polymorphisms (RFLP) analysis to follow the
inheritability of a particular allele by its telltale-sized
fragments across family pedigrees. Often, the polymorphisms
are the result of single nucleotide mutations that may or may
not change the structure of the protein that the gene
encodes. Polymorphisms can also occur within introns, where
they can influence gene expression or splicing. The most
recent molecular trends to emerge from the tracking of
single-nucleotide polymorphisms (SNPs) and disease
frequencies or morbidities is that clusters of SNPs, occurring
within distinct genes, may form the inherited altered
vulnerability for the pathology.

With that background, we can now categorize disease-

producing genetic alterations in individuals in three ways:

1. Single-gene mutations, which, because they are inherited

in recognizable patterns within families, are often referred to
as mendelian mutations. These can be further categorized as
autosomal dominant (when having one disease-causing allele
is all it takes, as with Huntington's disease or
neurofibromatosis) or autosomal recessive (when the disease
may be carried across generations and not be detected unless
the individual inherits the gene mutation from both mother
and father and thereby becomes homozygous). Genomic
information has already had an enormous impact on the
medical sciences by providing absolute diagnostic markers for
the highly inheritable neurodegenerative disorders that have
generally borne the names of dead neurologists (like
Huntington's disease), including the rare forms of some
diseases that have traditionally not been regarded as
inheritable (like Parkinson's and Alzheimer's). Special forms of
recessive disorder arise from the X chromosome. Because the
Y chromosome carries no somatic traits and codes only for
gender-determining factors, a single X-linked disease-
producing allele will cause the disease in males. However,
nonsymptomatic females will pass it on to half of their sons
but none of their daughters. Familial dysautonomia is an
autosomal recessive condition affecting Ashkenazi Jews that
can be diagnosed by a supersensitivity of the iris to
methacholine. Because the gene for phenylalanine
hydroxylase is subject to several different forms of inheritable
mutation, even heterozygous carriers can produce offspring
vulnerable to absence of the functional enzyme, leading to
phenylketonuria and progressive mental retardation unless
dietary phenylalanine is severely restricted well into early
adolescence. Other inheritable metabolic disorders leading to
mental retardation are shown in Table 13-1.

2. Polygenic diseases, in which multiple genetic factors are

responsible for the appearance of the disorder. Many of the
most prevalent brain diseases, such as depression,
schizophrenia, multiple sclerosis, and epilepsy, are complex
multigenic diseases. That is, for an inheritable vulnerability to
eventuate in the expression of the disease, several genetic
events must coexist, perhaps brought on by environmental
influences, such as a particularly intense life event. One
consequence of developing drugs for such complex,
multigenic diseases is that, even with today's molecular
wizardry, scientists cannot create a pertinent transgenic
animal model for complex genetic diseases in which to
evaluate new drugs, as they can with those much more rare
brain diseases caused by mutations in a single gene. On the
other hand, another way to look at complex genetic diseases,
arising from studies of monozygotic twins, is that the
concordance for these diseases is far less than the expected
100%, indicating that factors other than genetic must be able
to modify the vulnerability or resistance to the disease.
Pursuit of that line of reasoning may lead to medications that
enhance disease resistance rather than treat the disease after
its emergence.

3. Physically abnormal chromosomes can arise during

embryogenesis as cells divide, and the structural abnormality
of the chromosomes can be detected by microscopy. Trisomy
21 (a triplication of all or parts of chromosome 21, better
known as Down syndrome), fragmentation and deletions of
chromosome 7 resulting in Williams syndrome, and the fragile
X form of mental retardation are examples of this kind of
genetic problem.

Inheritable Metabolic Errors and Brain

Development
Most of the diseases listed in Table 13-1 are single-gene
defects leading to mental retardation or other forms of
behavioral abnormality. Almost all of them are untreatable, so
the forward-looking emphasis has been on genetic screening.
In Lesch-Nyhan syndrome (an X-linked recessive disorder),
affected males are both physically and mentally retarded and
exhibit a characteristic self-mutilating behavior. Although the
enzyme is expressed in many other cellular systems, it is
basically only the formation and function of the brain in which
the problems arise, for reasons that are not clear. Mice
lacking this enzyme show no behavioral phenotype, a lesson
for all those seeking to make transgenic mouse models of
human disease.

Cretinism, a severe neonatal hypothyroidism, may arise from

loss of thyroxin-producing enzymes and can be treated by
replacement therapy as soon as it is recognized. The same is
true of deficiencies of growth hormone and of gonadotropin-
releasing hormone. The latter hormone is lost in males by a
mutation (Kallmann's syndrome) in a specific neuronal
adhesion molecule required to provide the trail by which the
gonadotropin-releasing hormone neurons migrate from their
birthplace in the olfactory bulb to their intended final location
in the ventral hypothalamus. The porphyrias are a group of
mostly autosomal dominant mutations of enzymes that
synthesize the blood pigment heme; in one form, acute
intermittent porphyria, an otherwise asymptomatic case can
be triggered by infections, dehydration, or other unknown
factors into a variety of incapacitating neurological and
behavioral symptoms, including depression, mania, and
hallucinations. King George III had this disease with severe
manifestations during the American Revolution.

WHEN GENES ARE INFLUENCED BY

ENVIRONMENT

Introduction
In many diseases, several unknown genes clearly play a role
but in a manner that is both environmentally and behaviorally
influenced. The extent to which specific gene products are
expressed in the brain is a matter of the demands placed on
the neuron or glial cell by the incoming information, as we
saw in Chapter 3. The environment, both physical and social,
is the source of much of this external information. Eventually,
the information is converted into synaptic signals, which are
in turn transduced into intracellular second messengers and
ultimately into altered cytoplasmic and nuclear signals that,
through transcriptional regulation, can determine which genes
are turned on or kept off. Thus, the degree to which our
genotype is reflected in our functional form or phenotype
comes under environmental influence. The effects of the
environment are likely to be cumulative and perhaps vary
with the stages of development, and these influences on early
brain function may not be apparent until much later in life
(see Fig. 13-1).

Therefore, the cumulative effects of external events will

continually modify gene expression as the cellular systems of
the brain adapt to the adverse conditions imposed by
environmental challenges. So long as the neurons and their
supporting glia can adapt, the system will appear to be
"healthy." However, when the environmental demands exceed
in magnitude or duration the ability of the brain's adaptations
to maintain normal function, the dysfunctional state that
emerges is recognizable through the signs and symptoms of a
neurological or psychiatric disease. If learned behaviors have
their roots in changes in synaptic efficacy and response to
specific external sensory signals, then early "personality-like"
behaviors could appear to be inheritable when they are
instead learned. For example, Meaney and colleagues
reported that some rat mothers (dams) are much more active
in grooming and tending to their pups than others and that
the female pups grow up to follow their dam's style of
maternal duty. However, when pups are cross-fostered from
high- to low-intrusive dams and vice versa, the daughters still
grow up to exhibit maternal behaviors like the dam who
raised them regardless of their biological dam.

The three examples we will now examine are relatively

common disorders, and many powerful medications have
been developed to provide significant alleviation for some, but
not all, of those affected.

Figure 13-1. A neuron sees the world through

 the information received by its synaptic afferents.
The cumulative results of synaptic activity modify
gene expression by means of the nuclear actions
of metabolic end products (intracellular second
messengers, ions, and immediate early genes
leading to transcriptional regulation) triggered by
synaptic events. Changes in gene expression,
resulting from the influence of external events or
from changes in the internal environment of the
brain, alter the neuron's phenotype (e.g., the rate
of utilization of its own transmitter or receptors)
and, hence, the operation of the circuits in which it
participates.

Depression
The family of psychiatric diseases epitomized by changes in
mood (emotion, affect) is called the affective disorders, most
commonly depression, and they are very common. Between
10% and 20% of the population will have a serious (i.e.,
clinically significant) episode over their lifetime. Depression
and other serious affective manifestations may account for as
many as 50% of those hospitalized for psychiatric reasons.
There are two major forms, with different gender and genetic
characteristics. Bipolar depression is characterized by
recurrent, wide swings in mood from depression to extreme
elation, occurs slightly more often in women than men, shows
strong familial patterns of inheritability, and is primarily found
in young adults aged 18-45. Unipolar or major depression is
characterized by one or more prolonged episodes of
depression (measured in weeks), occurs far more frequently
in the population than the bipolar form, and is nearly three
times as common in women as in men. Depression can also
occur in childhood and early adolescence when the genders
are nearly equal in vulnerability, perhaps an important clue to
etiology. Studies of those with depression epitomize the
possibility for successful biological approaches to the
treatment of psychiatric disorders and exemplify the current
thinking about the study of animal models of human brain
diseases.

Catecholamine Theory of Affective Disorder

Animal studies indicate that catecholamines play a role in the
periphery with regard to stress and emotional behavior. The
catecholamine hypothesis of affective disorder states that, in
general, behavioral depression may be related to a deficiency
of catecholamine (usually norepinephrine) at functionally
important central adrenergic receptors, while mania results
from excess catecholamine. While substantial experimental
work supports this proposal, most of the experiments on
which this hypothesis is based derive from studies on the
brains of "normal" animals.

The original impetus for developing the catecholamine

hypothesis was the finding in patients being treated for
tuberculosis that various monoamine oxidase inhibitors (MAO-
Is) administered with other goals in mind, notably iproniazid,
acted clinically as mood elevators or antidepressants. Shortly
thereafter, this class of compounds also produced marked
increases in brain amine levels. By the same token, reserpine,
a potent tranquilizer and effective antihypertensive, depletes
brain amines (serotonin as well as the catecholamines) and
produces a serious depressed state (clinically
indistinguishable from endogenous depression) in about 20%
of those treated, and even suicide attempts in some people.
The similarity in overt behaviors of "reserpinized" rodents and
humans is one of the foundations of neuropharmacology and
generated biochemical insight into the underlying disease.

Both drugs (MAO-Is and reserpine) alter brain levels of

catecholamines and 5-HT quite equally. However, the fact that
a precursor of catecholamine biosynthesis, L-DOPA, can
reverse most of the reserpine-induced symptomatology in
animals has tended to bias many researchers in favor of the
catecholamine theory. In fact, by no means does the available
evidence for the involvement of norepinephrine rule out the
participation of dopamine, 5-HT, epinephrine, or other
putative transmitters in similar events. The three general
classes of drugs used to treat depressive disorders are the
MAO-Is, the tricyclic antidepressants, and the psychomotor
stimulants such as amphetamine. All of these pharmacological
agents appear to interact with catecholamines in a way that is
consistent with the catecholamine hypothesis. Thus, all four of
the suggested modes of action of amphetamines (partial
agonist, inhibitor of catecholamine reuptake, competitive
inhibitor of MAO, and displacer of presynaptic norepinephrine
and dopamine) would be expected to increase catecholamines
temporarily at their receptors. Administration of long-term or
high-dose amphetamine produces an eventual depletion of
brain norepinephrine and dopamine and inhibition of neuronal
activity in catecholamine neurons. This chronic depletion of
transmitter and the prolonged inactivity of catecholamine
neurons may account for the clinical observation of
amphetamine tolerance or the well-known poststimulation
depression or fatigue observed after chronic administration of
this class of drugs.

Well after they were recognized as effective antidepressants,

studies of norepinephrine conservation revealed that these
drugs shared an important action, the ability to block the
monoamine transporters. All of the tricyclic antidepressants
inhibit both the catecholamine and 5-HT presynaptic plasma
membrane transporters. Other reuptake inhibitors, not
necessarily of the tricyclic structural motif, have been made
selective for one or another monoamine transporter. For
example, mazindol and bupropion are dopamine-selective
reuptake inhibitors, and fluoxetine (Prozac) was the first of
the serotonin-selective reuptake inhibitors (SSRIs). These
drugs are far more selective than was the original tricyclic
imipramine, which is now recognized as a mixed
norepinephrine/5-HT transporter inhibitor. Although they have
proven to be effective in treating subsets of depressed
patients, some depressed patients, especially those with
onset in childhood or adolescence, are quite resistant to such
treatment.

The action of the MAO-Is also supports the catecholamine

hypothesis. All of these agents inhibit an enzyme responsible
for the metabolism of norepinephrine and various other
amines (5-HT, dopamine, tyramine, tryptamine). This
inhibition results in a marked increase in intraneuronal levels
of norepinephrine. According to the conceptual model of the
adrenergic neuron presented above, this intraneuronal
norepinephrine might eventually diffuse out of the neuron and
reach receptor cells, thereby overcoming the presumed
deficiency. A similar mechanism may also explain the
antagonism of reserpine-induced sedation with MAO-Is, since
there will be an initial replenishment of the norepinephrine
deficiency caused by reserpine.

Lithium, one of the main agents used to treat manic episodes

in bipolar disease, has also been studied with regard to its
action on the life cycle of the catecholamines. Interestingly,
pretreatment with lithium blocks the stimulus-induced release
of norepinephrine from rat brain slices. Other investigators
have suggested that lithium may facilitate reuptake of
norepinephrine. If the inhibition of release observed in
stimulated brain slices is due to a facilitated recapture
mechanism, then the mechanism of action of lithium is the
exact opposite of that of the antidepressant drugs, as would
be expected according to the catecholamine hypothesis.
Although a single injection of lithium may antagonize
responses to norepinephrine, chronic treatment of rats leads
to modest supersensitive norepinephrine responses.

On closer scrutiny of the clinical and experimental

pharmacological data cited in support of the catecholamine
theory of affective disorder, however, even the keenest
enthusiasts recognize significant inconsistencies. (1) Cocaine
is a very potent inhibitor of catecholamine reuptake and thus,
like the tricyclic antidepressants, should increase the
availability of norepinephrine at central synapses; but cocaine
does not possess any significant antidepressant activity. (2)
Iprandole, a tricyclic compound without any significant effect
on catecholamine uptake or any influence on central
noradrenergic neurons, is an effective antidepressant.
Furthermore, desensitization of postsynaptic -receptors after
chronic administration of tricyclic antidepressant drugs is not
entirely compatible with the proposed theory of a synaptic
catecholamine deficit. (3) In laboratory studies, the ability of
tricyclic antidepressants to inhibit catecholamine uptake and
the ability of MAO-Is to block MAO and elevate brain
catecholamines are apparent soon after administration.
Clinically, however, antidepressants must be given for several
days (10-14) to produce therapeutic effects. Perhaps some
novel genes must be turned on or off, or enhanced
catecholamine action at recurrent axons onto monoamine cell
bodies may further reduce their firing such that even less
catecholamine is released at synapses. (4) On the surface,
the pharmacological and clinical effects of the tricyclics and
lithium seem to fit nicely with the catecholamine theory of
affective disorder. These agents produce opposite effects on
norepinephrine disposition, and lithium is effective at treating
mania while the tricyclics are useful in treating depression.
However, lithium is also effective in the treatment of bipolar
depressed patients, where it dampens the emotional
oscillations into either mania or depression.

The fact that several questions have been raised about the
pharmacological data used to support the catecholamine
theory of affective disorder has prompted many investigators
to seek direct evidence of the involvement of norepinephrine
systems in affective disorders. The most extensive studies
have involved analysis of urinary excretion patterns of
catecholamine metabolites in patients with affective disorders.
The rationale has been that the urinary excretion of a
particular metabolite, like MHPG, may be a useful reflection of
central catecholaminergic processes. Despite the inherent
problems in urinary catecholamine metabolite measurement,
such as complications because of large contributions from
peripheral sources, some interesting findings have emerged.

Findings from several clinical studies now indicate that (1)

depressed patients as a group excrete less than normal
quantities of MHPG, (2) diagnostic subgroups of depressed
patients are particularly likely to have low urinary MHPG
values, (3) bipolar patients who switch from a depressive to a
euthymic or hypomanic state show a corresponding increase
in urinary MHPG, and (4) pretreatment urinary MHPG values
are predictive of the type of therapeutic response obtained
with catecholamine-directed reuptake inhibitors. Although
provocative and essentially consistent with the norepinephrine
theory of affective disorder, these clinical findings hinge on
the issue of the degree to which urinary MHPG reflects central
norepinephrine metabolism.

For the above reasons, the original hypothesis that

antidepressant drugs increase the availability of monoamines
in the brain has been updated to include the effects of long-
term antidepressant treatment on monoamine receptor
sensitivity. A wide array of effects of long-term treatment with
antidepressants has been reported in several monoamine
systems. The most consistent findings following chronic
administration of most of the clinically effective
antidepressant drugs to experimental animals are (1) a
reduction in the number of adrenoceptors and
downregulation of -adrenoceptor functioning; (2) an increase
in the sensitivity of central adrenoceptors, suggesting up-
regulation in central -adrenoceptor functioning; and (3)
similarly, both behavioral and electrophysiological studies also
point to up-regulation in the sensitivity of central 5-HT
receptors. These changes require time to be accomplished,
and perhaps that is why the delay in antidepressant action is
observed. However, all of the monoaminergic neurons have a
rich recurrent innervation, and enhancement of this
autoinhibition would also be expected from the use of re-
uptake inhibitors. If the monoamine neurons are thus reduced
in their excitability, the actual amount of transmitter released
from distal axons may be initially reduced, even with reuptake
inhibition. Therefore, another part of the lag in response to
treatment may be the requirement for reequilibration of
perikaryal excitability and restoration of baseline synaptic
function before the reuptake inhibition per se can enhance
postsynaptic receptor durations of action. Alternatively,
perhaps new gene products must be made and time is
required to ship them in sufficient amounts to the synapses.

Throughout this analysis of antidepressant drug actions, we

have noted frequent intrusions by 5-HT-related effects. In
fact, 5-HT deficiency deserves attention as an independent
causative factor in some forms of depression. Diminished
plasma levels of the rate-limiting 5-HT precursor L-tryptophan
(L-TRP) have been seen in several series of depressed
patients and linked to either reduced intestinal absorption or
increased hepatic catabolism of absorbed L-TRP stimulated by
pyrrolase; dietary restriction of L-TRP may also precipitate
depression in recently remitted patients. Cerebrospinal fluid
levels of the 5-HT catabolite 5-hydroxyindoleacetic acid (5-
HIAA) have been studied extensively in depressed subjects,
but the best correlations here are with attempts to commit
suicide and with impulsive, violent behavioral patterns. Some
postmortem studies of depressed patients have shown
increased numbers of certain 5-HT receptors, especially in
frontal cortical regions. Positron emission tomographic (PET)
studies of patients with depression show decreased glucose
utilization in the frontal cortex. Other pharmacological
challenges to depressed subjects, based on changes in sleep
electroencephalographic (EEG) patterns and neuroendocrine
secretion patterns, further support the concept of a central 5-
HT deficiency and postsynaptic receptor upregulation. Perhaps
the greatest support for this concept arose with the major,
and to some extent unexpected, antidepressant success of
fluoxetine, sertraline, and paroxetine, the SSRIs. An
interesting polymorphism in the transcriptional regulatory
domain of the 5-HT transporter has been found to be a
potential marker for patients who will more rapidly respond to
SSRIs. In this polymorphism, a block of 40 nucleotides may
be present or absent, but the "normal" form is not clear.
Those with the fragment missing seem to make less of the
transporter, perhaps suggesting why they show greater
sensitivity to its inhibition. However, it is not a good
diagnostic marker for depression.

More direct studies are necessary, however, before it is

possible to conclude whether any or all of the therapeutic
actions of antidepressant drugs are indications that the
emotional disorder was truly caused by functional deficiencies
of one or another monoamine. These studies await the
development of appropriate models for monitoring central
neurotransmitter functioning in humans, especially those at
high risk of developing depression based on their family
pedigrees. An interesting development, yet to be confirmed,
is the potential for other classes of drugs, in particular
antagonists of the neuropeptides substance P and
corticotropin-releasing hormone, to work as antidepressants
in both rodent models and patients with depression

Dopamine Hypothesis of Schizophrenia

Schizophrenia is a chronic disorganization of mental function
that affects thinking (paranoid ideas, inability to maintain a
focused thought, easily distracted, loose associations between
thoughts), feeling (typically referred to as blunted affect and
inappropriate responses to social situations), and movement
(from hyperactivity and excitement to persistent inactivity to
the point of maintaining bizarre postures for long periods of
time). The disease is most commonly recognized in very
young adults, particularly when confronted with severely
stressful life events, although evidence of social withdrawal
and disorganized thinking may have been noted before the
episode. About one-third of those having a single episode
may fully recover. Like bipolar depression, schizophrenia
shows strong familial patterns of inheritance in some cases,
but no mendelian pattern; thus, it is best relegated to the
category of complex genetic disorder. Available medications
can be very beneficial, but for most affected individuals, the
blunted affect, apathy, lack of volition, and social withdrawal
(the "negative" symptoms) may be intractable even though
the medications dampen the hallucinations and aggressive
behaviors (the "positive" symptoms).

In analogy to the catecholamine hypothesis of depression, a

biochemical explanation for schizophrenia arose from
observations that the only consistent feature among the
antipsychotic drugs used to treat the disease was their ability
to antagonize D2 dopamine receptors and that chronic
administration of amphetamine and other (mainly) dopamine-
mediated psychostimulants could produce a psychotic state
loosely resembling some aspects of schizophrenia. This
hypothesis in its simplest form states that schizophrenia may
be related to a relative excess of central dopaminergic
neuronal activity. Attempts to validate this hypothesis in
clinical studies have been intense but inconclusive. While D2
receptors are consistently increased in most postmortem
studies of schizophrenic brains, the influence of prior
antipsychotic treatments confounds the interpretation, and no
direct evidence of increased dopamine synthesis, turnover, or
function has been obtained, nor has noninvasive imaging of
dopamine receptor densities in previously unmedicated young
schizophrenics produced any consistent evidence for an
increase in D2 receptors within the striatum. However, the D4
receptor subtype, functionally akin to the D2 receptors in
inhibiting adenylate cyclase, has greater affinity for the
atypical antipsychotic drug clozapine (these drugs were called
"atypical" precisely because they were not good D2
antagonists) and to be increased specifically in postmortem
schizophrenic brains.

Experiments in animals have generated substantial support

for the idea that antipsychotic drugs are effective blockers of
dopamine receptors, but most of these animal studies
(whether behavioral, biochemical, or electrophysiological)
have been carried out after acute drug administration. This is
a very serious drawback since, as with the antidepressants,
the clinical effects of antipsychotic drugs (both antipsychotic
and neurological) take days, weeks, or even months to
develop. In nonhuman primates, chronic treatment with
antipsychotic drugs causes tolerance to the homovanillic acid
(HVA) increase normally observed in the putamen, caudate,
and olfactory cortex after a challenge dose of an antipsychotic
drug. In cingulate, dorsal frontal, and orbital frontal cortex,
however, increased levels of HVA are maintained throughout
the time course of chronic treatment. Similar observations
have also been made in studies carried out on autopsied
human brain specimens. If patients with the diagnosis of
schizophrenia are chronically treated with antipsychotic drugs,
a significant increase in HVA is found in the cingulate and
frontal cortex but not in the putamen and nucleus accumbens,
suggesting a locus for the therapeutic action of these drugs
and providing the first direct experimental evidence that
antipsychotic drug treatment increases the metabolism of
dopamine in the human brain in a regionally specific manner.
Establishing with certainty that the elevation in dopamine
receptor density is part of the disease process would be very
important for both etiological and diagnostic purposes and
would serve as a possible basis for treatment strategies. The
use of new noninvasive techniques, such as PET and single-
photon emission computed tomography (SPECT), to examine
dopamine receptor distribution and density in schizophrenia
holds promise that this may be accomplished soon.

Unlike depression, where there is no obvious neuropathology,

postmortem microscopic examination of the brains of patients
with schizophrenia has revealed pathological evidence for
abnormalities of neuronal density but without frank
neurodegeneration. Together with other findings (enlarged
ventricles, thin cortex) that are stable over years of
observation, this has led to the interpretation of schizophrenia
as a neurodevelopmental disorder. The negative symptoms of
cognitive disruption and affect have been correlated with
reduced brain metabolic activity in the frontal lobes, and this
may be an adaptive downregulation of dopaminergic activity
in this region of cortex. The atypical antipsychotic drugs have
received increasing attention for three reasons: (1) they can
reduce symptoms in patients resistant to other drugs, (2)
they produce fewer side effects, and (3) their pharmacology
may extend the neurotransmitter base for schizophrenia to 5-
HT as well as dopamine. Other evidence points to disruption
of corticostriatal glutamatergic circuits in the disease. In part,
this is based on the observation that the hallucinatory side
effects of so-called dissociative anesthetics, such as ketamine
(named for their apparent opposing effects on the EEG and
behavior) are also antagonists of the N-methyl-D-aspartate
(NMDA) and glutamate receptors. Animal models of altered
brain dopamine neurochemistry and behaviors have been
proposed in both rodents and primates by acute or chronic
injection of phencyclidine, another NMDA antagonist and
hallucinogen. The area is being intensively explored (see Fig.
13-2).

Figure 13-2. Schematic diagram of the main

neuronal circuits and transmitters dysregulated in
schizophrenia. The frontal neocortical projections
 to the striatum and other subcortical structures
are primarily glutamatergic (GLU) efferents.
Within the cortex, these units are regulated by -
aminobutyric acid (GABA)-containing interneurons
(neuropeptides of those GABA neurons not shown)
and by dopamine (DA) projections from the
ventral tegmental area (VTA) and serotonergic (5-
HT) projections from the raphe (R) nuclei. The
latter monoaminergic neurons may project to
either the cortical interneurons or to the cortical
efferent neurons. The raphe also projects to the
substantia nigra (SN), whose DA axons innervate
the basal ganglia but not the cortex. Current
evidence favors decreased levels of operation in
the nigrostriatal projections but overactivity in the
mesolimbic DA systems. "Atypical" antipsychotic
drugs may work in part through their capacity to
inhibit cortically enriched DA receptors (such as
the D4) as well as 5-HT receptors. (Based on
circuitry described by Maes and Meltzer, 1994; see
Meltzer, 1999.)

Drug Abuse
The continued, compulsive obsession with obtaining,
consuming, and experiencing self-administered drugs is a
major social and medical problem throughout the world.
Specific drugs of abuse have specific patterns of use and
dependence. Seven families of drugs have been recognized to
be obsessively self-administered by humans. In order of
prevalence, they are caffeine (as in coffee or tea), nicotine,
alcohol (grouped with benzodiazepines and barbiturates),
marijuana (and the congeners hashish and
tetrahydrocannabinol), psychostimulants (cocaine and
amphetamines), opiates (morphine, heroin, and other
agonists), and the hallucinogenic drugs [LSD], phencyclidine,
and MDMA [3,4-dioxymethylene methamphetamine],
otherwise known as "ecstasy"). Note that the most widely
abused substances, caffeine, nicotine, and alcohol, are legal
in the United States and that both the federal and state
governments collect substantial "sin taxes" on the latter two,
while attempts to place taxes on caffeine helped spark the
American Revolution. Recreational use of alcohol or nicotine,
however, may serve as the gateway to illicit and powerful
drugs of abuse.

The neurobiological substrate for the acute rewarding effects

of the four major classes of abused drugs (alcohol, nicotine,
opiates, and psychostimulants) has focused on the area of the
ventral forebrain that surrounds the nucleus accumbens. This
nucleus, with neurons of several forms and efferents, is
heavily innervated by pontine monoamine neurons and is
considered to be in a continuum with the nearby neurons of
the amygdaloid complex. Increased release of dopamine
within the region of the nucleus accumbens has been directly
observed by microdialysis in animals trained to self-
administer alcohol and nicotine. In the case of cocaine,
antagonists of the D3 dopamine receptor selectively reduce
the rewarding effects of the drug, as shown by decreases in
self-administering behavior. However, while the key site for
eliciting opiate reinforcement also seems to be within the
nucleus accumbens, the dopamine afferents are not required
for maintenance of opiate self-administration.

Certain forms of alcoholism are thought to arise from

unknown but genetically transmittable factors that increase
vulnerability to alcohol dependence, especially in the male
offspring of male alcoholics, who may also show antisocial
personality disorder. Animal models of genetically
transmittable alcohol preference have been achieved several
times by constant cross-breeding of the few rats that initially
show modest alcohol preference. Later generations develop
an almost universal predilection to consume alcohol for its
pharmacological (as opposed to nutritional or thermal)
properties. In the alcohol-preferring animals developed in the
United States and in Sardinia, there is diminished dopamine
and 5-HT innervation of the nucleus accumbens and increased
GABA innervation of that site. Among other manipulations, 5-
HT-specific reuptake inhibitors will reduce alcohol self-
administration in these rats. Alcohol-accepting rats bred in
Finland, however, do not show these neurochemical
differences.

Alcoholism vulnerability in Asian human populations is actively

suppressed by a prevalent genetic mutation leading to
reduced capacity to oxidize the ethanol catabolite
acetaldehyde. Those carrying this mutation (half of these
populations) react to alcohol as though they were receiving
the drug antabuse, which directly inhibits acetaldehyde
dehydrogenase. After a brief flurry of excitement over the
possibility that certain alleles of certain dopamine receptors
might be associated with alcoholism and other forms of
dependence, the search for the responsible genes has
resumed. While there are major differences in the sensitivity
of certain inbred rat strains to the reinforcing actions of
opiates and psychostimulants, the evidence for such inherited
vulnerability has not been well studied in humans, largely
because these drugs are universally highly reinforcing once
their recreational use has been initiated. Nonetheless, these
high- and low-sensitivity animal models of drug response may
prove useful in determining the genetic basis for vulnerability
to drug dependence in humans.

Efforts to deal with drug abuse have focused on reducing the

supplies of illicit drugs by law-enforcement strategies aimed
at the prevention of drug smuggling. After spending hundreds
of millions of dollars on "supply reduction" only to see the
suppliers able to sell more products at lower prices, agencies
have devised other strategies directed toward demand
reduction. For many years, methadone, a mixed agonist-
antagonist for morphine with reduced euphorigenic
properties, and some of its long-acting congeners have been
known to reduce opiate dependence, increasing job
performance and reducing criminal behavior; but its
widespread adoption was delayed because of the belief in
some quarters that it was merely substituting one drug for
another. Only recently has methadone been approved for the
long-term treatment of opiate abuse. In animal studies, the
opiate antagonist naloxone, also reduces some of the
rewarding effects of alcohol self-administration for reasons
that are not yet clear. Naltrexone, the orally active congener
of naloxone, has been approved for the long-term treatment
of alcoholism. Specially reformulated long-acting naltrexone is
under investigation as a means to improve compliance by
removing the daily decision to take the medication.

WHEN KNOWING THE GENES DOES NOT

HELP

Introduction
Lastly, we consider a group of diseases that fall under the
general heading neurodegenerative disorders, most of which
remain untreatable, although there are animal models that
pose striking opportunities for pharmacological development
in this area. Some of the more creative avenues to restoring
brain function, from transplantation of neurons and neuronal
stem cells to the transfer of autosomal genes, are explored
here.

Alzheimer's Disease
This chronic, progressive, degenerative disorder, which is in
some rare cases familial, is becoming a greater health-care
problem as the population survives to older ages. The nearly
linear incidence of Alzheimer's detection with age makes this
a very prevalent disorder, perhaps being detected in as many
as 50% of those older than age 85, which is no longer a
unique occurrence in our society. Alzheimer's disease is
characterized behaviorally by a severe impairment in
cognitive function, including memory, the ability to recognize
objects, and the ability to orient oneself in time and space.
Neuropathologically, the disease is epitomized by the
appearance of neuritic plaques (amorphous extracellular
deposits of proteinaceous matter) and neurofibrillary tangles
(intracellular whorls of paired helical filaments), generally
restricted to specific cortical regions (hippocampus, frontal
and temporal lobes more than parietal and far more than
occipital lobes) and specific cortical layers (especially
prominent in layers III, V, and VI, where the large neurons
are located). The pathology is also seen within a few of the
subcortical structures (ventral pallidum, locus ceruleus)
projecting to the cortical areas of involvement.

A cholinergic hypothesis of cortical dysfunction has been

implicated in Alzheimer's disease based on the findings that
(1) administration of centrally acting muscarinic-blocking
agents to normal individuals induces a loss of recent memory;
(2) in the cerebral cortex and hippocampus of patients with
Alzheimer's disease, there is a dramatic reduction of ACh,
choline acetyltransferase, and high-affinity choline uptake;
(3) in Alzheimer's disease patients, there is a severe
reduction of neurons in the nucleus basalis of Meynert, the
primary cholinergic input to the cortex; and (4) in some, but
not all, studies, a decrease in muscarinic and nicotinic
receptors has been noted. However, patients with Alzheimer's
disease also have decreased levels of somatostatin,
neuropeptide Y, substance P, and corticotropin-releasing
factor, increased numbers of corticotropin-releasing factor and
NMDA receptors; as well as reduced numbers of locus
ceruleus neurons. More recent detailed comparisons of the
rate at which synapses in general and cholinergic markers
specifically are lost suggest that many of the other
neurotransmitters are affected before the cholinergic
afferents, indicating that the cholinergic decline may be
secondary to the loss of cortical neurons. A decrease in the
neuronal content of the nucleus basalis of Meynert has also
been observed in some patients with Down syndrome
(trisomy 21, the only other condition known to express
neuritic plaques) and with Parkinson's disease.

Thus, the stormy marriage of Alzheimer's disease to a

cholinergic dysfunction may involve some extramarital
relationships. Little success has been achieved to date by
treating patients with choline, lecithin, physostigmine, or the
muscarinic agonist arecoline. Current efforts focus on the
development of drugs that can block acetylcholinesterase
centrally without hepatoxicity and peripheral autonomic side
effects. Future therapy had been thought to focus on specific
M2-muscarinic antagonists, nicotinic agonists, or the
transplantation of fetal cholinergic tissue. However, a
potentially more revealing strategy over the past decade has
been a direct assault on the causes of the disease, made
possible by the isolation, purification, and sequencing of the
proteinaceous matter that has been the hallmark of the
pathology, the amyloid plaques. Once the aggregated forms
had been sequenced, scientists isolated and sequenced the
entire gene and identified its natural product as a large
protein of unknown function, named the amyloid precursor
protein (APP). The structure of normal APP, combined with
studies of relatively rare families with greatly increased
frequencies of Alzheimer's with especially early onset, has led
investigators to focus on errors in APP processing that could
generate the short fragments of the amyloid plaques.
Although no such proteases were previously known,
concentrated efforts to find enzymes capable of such
proteolysis (termed - and -secretases) were soon successful
and will provide important targets for drug development and
perhaps for the known effects of mutations in the proteins
presenilin-1 and -2.
The creation of transgenic mice overexpressing one form of
the -APP gene in cortical neurons and producing plaques
within 18 months offered the possibility of screening for drugs
that may reduce amyloid depositing and perhaps even of
revealing the reasons why the aged brain overproduces this
and other proteins in the first place. In addition, it has been
possible to reverse the extracellular deposits by immunizing
the mice with the fragment of APP that forms the aggregates
(A- 1-42). It remains unclear how other recognized genetic
predispositions (e.g., inheritance of the apoliprotein E-4
allele) also interact with environmental conditions or events
such as head trauma to initiate the process of pathological
change. The recognition that APP may be a receptor for the
light chain of kinesin-1, the motor molecule underlying axonal
transport, suggests a functional role for APP; but as of now,
there is still much to be resolved. Nevertheless, at least in
mice, it appears that combined pathology in the genes of APP
and tau (the major protein of microtubules) is required to get
the full pathological picture of plaques and tangles.

Parkinson's Disease
Parkinson's disease (PD) is a progressive neurodegenerative
disorder of the basal ganglia characterized by tremor,
muscular rigidity, difficulty in initiating motor activity, and loss
of postural reflexes. It is observed in approximately 1% of the
population over age 55. For over 75 years it has been known
that PD is characterized pathologically by loss of pigmented
cells in the substantia nigra, but only since 1960 has the
substantial loss of dopamine in the striatum been
documented. It is now clear that PD can be defined in
biochemical terms as primarily a dopamine-deficiency state
resulting from degeneration or injury to dopamine neurons.
The most striking degenerative loss of dopamine neurons is
observed in the nigrostriatal system. Even in patients with
mild symptoms, a striatal dopamine loss of 70%-80% occurs,
while severely impaired subjects have striatal dopamine
depletions in excess of 90%. Since the dopamine transporter
is heavily expressed in the terminals of dopamine neurons
that are lost in PD, it is not surprising that striatal binding of
agents that label this site (cocaine, nomifensine, and
mazindol) is lost in the parkinsonian striatum. This alteration
corresponds well with the loss of functional dopamine uptake
visualized in vivo by PET, using 18F-L-DOPA uptake or 18F-
nomifensine, or by SPECT, using other presynaptic dopamine
labels. Although striatal dopamine loss represents the primary
neurochemical abnormality in the PD brain, typical
parkinsonism is accompanied by loss of other dopamine
systems and other monoamine neurons. Some degeneration
of dopamine-containing neurons is also apparent in the
mesolimbic, mesocortical, and hypothalamic systems, as is
loss of norepinephrine-containing neurons in the locus
ceruleus and of serotonin neurons. Non-monoamine systems
are also affected, with depletions observed in somatostatin,
neurotensin, substance P, enkephalin, and cholecystokinin-8.
Since many of these nondopamine systems indirectly interact
with mesotelencephalic dopamine systems, changes in some
of them are bound to influence the function of dopamine
neurons in a complex way. Before it was known that
dopamine is severely depleted, for example, treatments with
anticholinergic drugs had been viewed as moderately
effective.

Since the earliest and most substantial neurochemical

abnormality in PD is the loss of dopamine, the modern
strategy for treatment has concentrated on restoring the
dopamine deficit. The theoretical strategies here include
substrate supplementation, direct and indirect dopamine
agonists, metabolic inhibitors (MAO-Is) and uptake inhibitors.
The most successful treatment has been the use of L-DOPA,
recognized by the Nobel Prize for Physiology or Medicine in
2000. Currently, L-DOPA treatment is usually combined with
an inhibitor of DOPA decarboxylase that acts only outside the
brain. This enhances the amount of the absorbed L-DOPA that
can enter the brain and alleviates the gastrointestinal
symptoms that arose from the higher doses required
previously. As the degeneration of dopamine (and other)
neurons progresses, the requirements for L-DOPA increase
and are accompanied by interruptions of its effectiveness that
are poorly understood ("wearing-off" and "on/off" responses).
Direct dopamine agonists have some benefit for patients
whose responsiveness to L-DOPA is greatly reduced or erratic.
So far, the only direct-acting dopamine agonist that has found
extensive use in PD is bromocriptine, primarily a D2 agonist.
However, selective D1 agonists, such as ropinirole and
cabergoline, have been developed which exhibit longer
durations of action without the absorption or blood-brain
barrier permeation problems that have been suggested as
causes for the variations in L-DOPA effectiveness. Other
agents belonging to this class will no doubt prove useful in the
future as supplements or alternatives to L-DOPA.

In view of the behavioral and electrophysiological studies that

suggest that D1 receptor activation is necessary for the
effects of D2 receptor stimulation to be maximally expressed
in normal animals as well as in animals with supersensitive
dopamine receptors, the functional interaction between D1
and D2 receptors could have important implications in PD,
where stimulation of postsynaptic dopamine receptors confers
symptomatic benefit. Knowledge of the optimal ratio of
relative drug activity at D1 and D2 receptors that is required
to elicit effective stimulation of dopamine-mediated function
may provide a basis for the design of new drugs. Also, more
knowledge of the distribution and function of various
dopamine receptor subtypes should facilitate the development
of new agents to treat dopamine-deficiency states. Fetal
neurons and neuronal stem cells are likely to be used in
model systems to treat the earliest symptoms of PD since the
motor effects are an excellent index of functional recovery. As
with Alzheimer's, the recognition of rare familially transmitted
disease has led to the isolation of an abnormal protein,
parkin, in autosomal recessive juvenile parkinsonism and of
-synuclein mutations in adult forms, apparently related to
abnormal accumulations of protein in aging dopamine
neurons, known as Lewy bodies. A special form of dementing
illness in aging that affects frontal lobe function is epitomized
by cortical Lewy bodies whose composition and relation to
nigral neuron Lewy bodies is under intense investigation.

Primate Model of Parkinson's Disease

In 1983, researchers at Stanford University identified a
contaminant in locally produced "synthetic heroin" that
induced a PD-like syndrome in some individuals who self-
administered this street drug. The contaminant identified in
this preparation, 1-methyl-4-phenyl-1, 2,3,6-
tetrahydropyridine (MPTP), exhibits a high degree of
anatomical and species-specific toxicity. MPTP administered
systemically in low doses to nonhuman primates produces
parkinsonian symptoms and destroys nigrostriatal dopamine
neurons while sparing several other brain dopamine systems.
The discovery of the selective neurotoxic properties of MPTP
and the development of a primate model of parkinsonism
stimulated a strong resurgence of inquiry into the causes and
treatment of PD. Still, the mechanism responsible for the
selective dopamine neurotoxic features of MPTP in primates
has not been conclusively established.

MPTP appears to act as a protoxin, and MAO-B-mediated

bioactivation of MPTP to 1-methyl-4-phenylpyridinium (MPP+)
plays a critical role in the ultimate neurotoxic action.
Administration of MAO-B inhibitors, including L-deprenyl,
affords full protection against the neurotoxic action of MPTP.
Since dopamine uptake inhibitors such as mazindol and GBR-
12909 also protect against MPTP-induced neurotoxicity, it has
been suggested that MPTP is oxidized to MPP+ outside the
dopamine neurons, perhaps by MAO in astrocytes. MPP+ is
then transported and concentrated by the dopamine uptake
system in dopamine neurons. Once inside the dopamine
neuron, MPP+ exerts its neurotoxic effect by acting as a
mitochondrial poison through inhibition of respiration at site i
of the electron transport chain. This postulated mechanism for
MPTP's neurotoxicity in dopamine neurons is illustrated in
Figure 13-3.

These animal data on the mechanisms of MPTP killing of

dopamine neurons have been applied in the effort to delay
the progression of early PD: large clinical trials combining
MAO-Is and antioxidants have been interpreted as showing
some effectiveness. Future treatments may also profit from
studies now actively being pursued in the animal model:
surgical transplantation of fetal substantia nigra neurons,
transplantation of autologous (i.e., from the same patient)
adrenal medullary cells or skin fibroblasts genetically
transfected to express tyrosine hydroxylase, and even special
virus constructs to transfer the tyrosine hydroxylase gene into
surviving striatal neurons, with or without some of the
neurotrophic factors which may be required.

Figure 13-3. Hypothesized mechanisms of

neurotoxicity of 1-methyl-4-phenyl-1,2,3,6-
tetrahydropyridine MAO-B, monoamine oxidase
type B; ATP, adenosine triphosphate; MPP+, 1-
methyl-4-phenylpyridinium (MPTP), which
produces parkinsonism in primate species.

FUTURE PROSPECTS
There is no shortage of treatment opportunities for
neuropharmacologists, and we have surveyed only some of
the diseases in which causes and treatments are under
intensive investigation. The interested reader may find
additional targets of opportunity in the literature of other
neurological or psychiatric disorders that follow the dictum of
the moment: that treatment success seems to be inversely
proportional to the molecular understanding of the etiology or
pathophysiology. Along this spectrum from treatable, not
understood to untreatable, almost-understood diseases, we
find the following: The hypertensions: A large collection of
disorders identified by the medical finding, often without
symptoms, of elevated systolic and diastolic blood pressures.
Depending on cause, a variety of medications may restore
blood pressure to normal levels: diuretics, peripheral
sympatholytics, ganglionic blocking agents, central 2-
adrenergic agonists (clonidine) and 1 antagonists (prazosin),
and quite unexpectedly drugs that inhibit angiotensin III,
either by inhibiting its formation through molecules
engineered as angiotensin-converting enzyme inhibitors or by
receptor antagonism.

Obsessive-compulsive disorder and anxiety disorder: Highly

prevalent conditions characterized by chronic, unrealistic
anxiety and compulsive behavior patterns, in some cases
chemically elicitable, exhibiting semispecific cerebral
metabolic alterations. For unknown reasons, these disorders
often respond well to certain medications such as fluoxetine
and other SSRIs; inbred animal model systems may provide
clues someday.

The epilepsies: Convulsive disorders arising from trauma,

infections, tumors, or unknown causes. Postencephalitic or
posttraumatic scarring and acute fever are known causes, but
many others are suspected. Several drugs are partially
effective for reasons that are poorly understood, and the only
useful animal models are those that simulate the disease
condition (electroshock or drug treatments). Again, studies of
the rare familial forms of epilepsy have led to isolation of
mutations in ion channels, which is not surprising in view of
the problems with excitability.

Multiple sclerosis: Chronic or recurrent demyelination leads to

progressive, focal neurodegenerative loss of sensory and
motor function, often accompanied by depression and
cognitive deterioration. Drugs that slow down the immune
system (interferon-b) or reduce cell division in the
macrophage-monocyte lineage (2-Cl-deoxyadenosine) have
shown some potential.

Human immunodeficiency virus (HIV)-associated cognitive

and motor dysfunction: HIV, known not to infect neurons,
nevertheless produces progressive neurodegeneration,
perhaps mediated by either inflammatory cytokines or
neurotoxins of either central (e.g., NO or Glu) or peripheral
(e.g., quinolinic acid) origin. No current treatments are
effective. The high-activity antiretroviral treatment cocktails
that have been successful in many HIV-infected patients at
great financial cost seem not to penetrate the blood-brain
barrier, leaving the infected brain as a sanctuary for the virus.
Primate models of HIV have been found to reflect the slow
inflammatory changes of the early illness and may be a model
for CNS treatment development.

Amyotrophic lateral sclerosis: An untreatable progressive

neurodegenerative disorder restricted to the motor neurons of
the spinal cord, corticospinal tract, and medulla and linked to
a mutation in the enzyme Cu-Zn superoxide dismutase
(SOD). In a mutant mouse model, this enzyme has been
overexpressed, and motorneurons do die, leading advocates
of the free radical toxicity to load up their vitamin E stores to
ward off the ravages of the degenerative process; but that
may not be an explanation for the human disease at all.
Either eliminating natural SOD or overexpressing the natural
SOD has no effect on the toxicity seen when the human
disease's mutated form is expressed, suggesting that there is
something toxic about the molecule itself; and pathological
aggregates have been reported.

Before leaving student readers to find their own path forward

into this daunting field of major untreated diseases, let us
make one final observation. The present-day armamentarium
of drugs useful in the treatment of CNS disorders arose, for
the most part, without good insight into the nature of the
diseases, by making fortuitously clever structural
modifications of drugs often developed initially for the wrong
reasons and by refining animal models that only very loosely
simulated the human condition or predicted the results of
structure-activity modifications in the leading compound.
Future treatments may be accomplished most directly when
one or more disease-causing genes are understood and then
replaced by gene transfer technology. Several promising
animal models have been reported in which cells transfected
with gene constructs are stereotaxically inserted into the
brain or a cerebral ventricle; these cells then express the
desired gene product (e.g., tyrosine hydroxylase or a
neurotrophic growth factor) and provide enduring
replacement at least in experimental PD or Alzheimer's
disease models. Much work is being done to develop viruses
as packaging vectors to deliver next-generation gene
constructs to specific target neurons, without the need for
neurosurgical manipulation to transfer the genes, although
this remains an unrealized goal.

It is not surprising that drug development in this field is a

multibillion dollar annual investment since untreatable chronic
dysfunctions of the brain represent one of the largest pockets
of cost in health care. Investors in this process are firmly
convinced that the "tools of modern drug discovery," as
exemplified by the angiotensin-converting enzyme inhibitors
and the angiotensin III antagonists, will successfully guide
logical drug development in the future. These tools are (1)
the genetic understanding of diseases and their
environmentally sensitive regulatory factors, (2) animal
models that closely simulate the pathophysiology, (3) cellular
models in which to express cloned transmitter receptors, and
(4) detailed three-dimensional, atomic-level understanding of
the target molecules where the drugs are intended to act.

A recent example may illustrate how a previously lethal

neurodegenerative disease can go from untreatable to
potentially treatable through such a sequence of insights.
Huntington's disease, an autosomal dominant disease, was
the first inheritable brain disease to have its mutation mapped
and its sequence identified; the normal sequence was named
"huntingtin." One of the many surprises of this mutated
protein was that it represented the tip of the iceberg of a
series of related genetic disorders in which a trinucleotide
repeat, CAG (coding for the amino acid glutamine), expands
from the normal 10-35 to more than 41 and is a near certain
predictor of the disease having been inherited. (The
spinocerebellar ataxias share a similar trinucleotide expansion
but much greater genetic and clinical variation.) The
expansion of CAGs and glutamines was shown to lead to a
gain of function in protein-protein binding, and the most
recent hot trail was considered to be death of the basal
ganglion neurons, which indicates Huntington's disease signs
and postmortem pathology through some consequence of the
enhanced binding, or failure to dissociate from a protein
partner. However, a totally different picture, loss of a
beneficial effect of the normal protein, has emerged from
recognition that normal huntingtin can upregulate in vitro the
expression of brain-derived neurotrophic factor by cortical
neurons known to project to the striatum, a function that is
lost by the mutations to huntingtin. If brain-derived
neurotrophic factor could be delivered to striatal neurons in
adequate amounts, perhaps they need not die in the future.

Given these tools, the imaginative drug design team will then
rely on the capacity of synthetic chemistry to compute the
potential steric requirements of receptor pockets for
manipulation and to create molecular dynamic simulations of
putative drug candidates. In the next-to-final step before
selecting the color of the tablets, the ideal candidate will then
be synthesized within the computer work station from chiral
molecular building blocks in the archives and then matched
against the world's database of previously made molecules for
initial validation of activity. After that, it is off to the patent
office, the Food and Drug Administration, and a few million
dollars' worth of clinical trials. Piece of cake? Maybe there are
better ways, but no one knows them yet. Students, start your
engines.

SELECTED REFERENCES
Bard, F., C. Cannon, R. Barbour, R. L. Burke, D. Games, H.
Grajeda, T. Guido, K. Hu, J. Huang, K. Johnson-Wood, K.
Khan, D. Kholodenko, M. Lee, L. I. Ieberburg, R. Motter, M.
Nguyen, F. Soriano, N. Vasquez, K. Weiss, B. Welch, P.
Seubert, D. Schenk, and T. Yednock (2000). Peripherally
administered antibodies against amyloid beta-peptide enter
the central nervous system and reduce pathology in a mouse
model of Alzheimer disease. Nat. Med. 6, 916-919.

Bruijn, L. I., M. K. Houseweart, S. Kato, K. L. Anderson, S. D.

Anderson, E. Ohama, A. G. Reaume, R. W. Scott, and D. W.
Cleveland (1998). Aggregation and motor neuron toxicity of
an ALS-linked SOD1 mutant independent from wild-type
SOD1. Science 281, 1851-1854.

Davis, K. L., R. C. Mohs, D. B. Marin, D. P. Purohit, D. P. Perl,

M. Lantz, G. Austin, V. Haroutunian, and A. D. Korczyn
(1999). Neuropeptide abnormalities in patients with early
Alzheimer disease. Arch. Gen. Psychiatry 56, 981-987.

Esler, W. P. and M. S. Wolfe (2001). A portrait of Alzheimer

secretases-new features and familiar faces. Science 293,
1449-1454.

Fox, H. S., M. R. Weed, S. Huitron-Resendiz, J. Baig, T. F.

Horn, P. J. Dailey, N. Bischofberger, and S. J. Henriksen
(2000). Antiviral treatment normalizes neurophysiological but
not movement abnormalities in simian immunodeficiency
virus-infected monkeys. J. Clin. Invest. 106, 37-45.

Francis, D., J. Diorio, D. Liu, and M. J. Meaney (1999).

Nongenomic transmission across generations of maternal
behavior and stress responses in the rat. Science 286, 1155-
1158.

Jentsch, J. D., D. E. Redmond, Jr., J. D. Elsworth, J. R. Taylor,

K. D. Youngren, and R. H. Roth (1997). Enduring cognitive
deficits and cortical dopamine dysfunction in monkeys after
long-term administration of phencyclidine. Science 277, 953-
955.

Jones, J. E., and R. B. Johnston, Jr., eds. (2001). Multiple

Sclerosis: Current Status and Strategies for the Future.
National Academy Press, Washington, D.C.

Kamal, A., G. B. Stokin, Z. Yang, C.-H Xia, and L. S. B.

Goldstein (2000). Axonal transport of amyloid precursor
protein is mediated by direct binding to the kinesin light chain
subunit of kinesin-1. Neuron 28, 449-459.

Koob, G. F. and M. LeMoal (2001). Drug addiction,

dysregulation of reward, and allostasis.
Neuropsychopharmacology 24, 97-129.

Kramer, M. S., N. Cutler, J. Feighner, R. Shrivastava, J.

Carman, J. J. Sramek, S. A. Reines, G. Liu, D. Snavely, E.
Wyatt-Knowles, J. J. Hale, S. G. Mills, M. MacCoss, C. J.
Swain, T. Harrison, R. G. Hill, F. Hefti, E. M. Scolnick, M. A.
Cascieri, G. G. Chicchi, S. Sadowski, A. R. Williams, L.
Hewson, D. E. J. Carlson, R. J. Hargreaves, and N. M. J.
Rupniak (1998). Distinct mechanism for antidepressant
activity by blockade of central substance P receptors. Science
281, 1640-1645.

Kupfer, D. J. and E. Frank (2001). The interaction of drug-

and psychotherapy in the long-term treatment of depression.
J. Affect. Disord. 62, 131-137.

Lewis, J., D. W. Dickson, W.-L. Lin, L. Chisholm, A. Corral, G.

Jones, S.-H. Yen, N. Sahara, L. Skipper, D. Yager, C. Eckman,
J. Hardy, M. Hutton, and E. McGowan (2001). Enhanced
neurofibrillary degeneration in transgenic mice expressing
mutant tau and APP. Science 293, 1487-1491.

Marenco, S., and D. R. Weinberger (2000). The

neurodevelopmental hypothesis of schizophrenia: following a
trail of evidence from cradle to grave. Dev. Psychopathol. 12,
501-527.

Meltzer, H. Y. (1999). Treatment of schizophrenia and

spectrum disorders: pharmacotherapy, psychosocial
treatments, and neurotransmitter interactions. Biol.
Psychiatry 46, 1321-1327.

Nestler, E. J. and G. K. Aghajanian (1997). Molecular and

cellular basis of addiction. Science 278, 58-63.

Phillips, M. I., E. A. Speakman, and B. Kimura (1993). Levels

of angiotensin and molecular biology of the tissue renin
angiotensin systems. Regul. Pept. 43, 1-20.

Plaitakis, A. and P. Shashidharan (1994). Amyotrophic lateral

sclerosis, glutamate, and oxidative stress. In
Psychopharmacology: The Fourth Generation of Progress (F.
E. Bloom and D. J. Kupfer, eds.). Raven Press, New York, pp.
1531-1544.

Price, D. L. S. S. Sisodia, and D. R. Borchelt (1998). Genetic

neurodegenerative diseases: the human illness and transgenic
models. Science 282, 1079-1083.

Sipe, J. C., J. S. Romine, J. A. Koziol, R. McMillan, J. Zyroff,

and E. Beutler (1994). Cladribine in treatment of chronic
progressive multiple sclerosis [see comments]. Lancet 344, 9-
13.

Tallman, J. F. and S. G. Dahl (1994). New drug design in

psychopharmacology: the impact of molecular biology. In
Psychopharmacology: The Fourth Generation of Progress (F.
E. Bloom and D. J. Kupfer, eds.). Raven Press, New York, pp.
1861-1874.

Tiraboschi, P., L. A. Hansen, M. Alford, E. Masliah, L. J. Thal,

and J. Corey-Bloom (2000). The decline in synapses and
cholinergic activity is asynchronous in Alzheimer's disease.
Neurology 55, 1278-1283.

Weinberger, D. R. (1994). Neurodevelopmental perspectives

on schizophrenia. In Psychopharmacology: The Fourth
Generation of Progress (F. E. Bloom and D. J. Kupfer, eds.).
Raven Press, New York, pp. 1171-1184.

Wirdefeldt, K., N. Bogdanovic, L. Westerberg, H. Payami, M.

Schalling, and G. Murdoch (2001). Expression of alpha-
synuclein in the human brain: relation to Lewy body disease.
Brain Res. Mol. Brain Res. 92, 58-65.

(+/-) Show / Hide Bibliography

Send Feedback Teton Server (5.0.0) - ©2007 Teton Data Title Updates
Customer Service Systems User
800.901.5494 Send Us Your Comments Responsibilities
Training Center
What's New