- Summary
-
- The hypothesis that brain noradrenergic
systems have a broad biological function which
is related to human fear or anxiety is reviewed.
Data from studies of the function of the nucleus
locus coeruleus in non-human primates are
presented in the context of recent anatomical,
physiological, pharmacological, and animal
behavioral experiments. Implications are
suggested for the treatment of anxiety, drug
addictions, pain, and psychosomatic
diseases.
-
- The idea that fear, or the related emotion,
anxiety, might be associated with the
noradrenergic-sympathetic nervous system has a
long history. Darwin's 1872 picture of fear
merging into terror (1) provided the basis for
this association, and his observations still
provide a useful beginning and descriptive
definition for a review of such emotions.
-
- The identification of adrenalin (2) which
stimulated many of the physiological changes
described by Darwin
supported the involvement of the sympathetic
nervous system, but the fact that adrenalin
infusion did not reliably induce the subjective
emotions, in addition to other data, led Cannon
to postulate essential central nervous system
mechanisms in pain, fear, and rage (3). There is
still no concensus as to what these central
neural substrates of emotions might be, except
that they involve the "limbic system" (4). The
physiological changes associated with fear
(5-7), in particular, are believed to result
from interactions of the sympathetic and the
parasympathetic nervous systems (8), which are
thought to be non-specific components of arousal
responses to "stresses" (9) and a number of
emotions.
-
- In addition to evidence implicating
adrenergic systems, studies of anxietyreducing
compounds also suggest the involvement of a
number of neurotransmitters (10). Recently, the
discovery of specific opiate receptors (11) and
benzodiazepine receptors (12) has suggested
endogenous aoxiolytic neurotransmitters which
might provide evidence regarding the central
mechanisms of anxiety. It is the purpose of this
review to suggest that these and other
anxiety-related neurotransmitters may interact
with brain norepinephrine systems which mediate
a biological function that at various levels
might be described as surprise, alerting,
astonishment, alarm, anxiety, fear, panic, or
terror. Our hypothesis is based on studies of
brain noradrenergic function in the stump-tailed
monkey (Mscaca arctoides or spscioss) utilizing
and combining electrical stimulation or lesions
of the major noradrenergic nucleus locus
coeruleus (LC) with the administration of
pharmacologic agents which have relatively
specific interactions with noradrenergic
function. We will first describe the rationale
for these studies, review the results briefly,
and then review anatomical, pharmacological, and
physiological evidence pointing to a connection
of this system with human anxiety or fear.
-
- Studies of the nucleus locus coerulsus in
monkeys
-
- Why study the locus coeruleus The dark blue
(coeruleus) streak in the dorsolateral tegmentum
of the pons is thought to have the highest
density of norepioaphrine-containing neurons in
the brain (13). It contains nearly half of the
norepinephrine (NE) neurons (14) and produces
over 70% of the total NE found in macaque brain
(15). In M. arctoides the nucleus is compact and
almost entirely composed of NE containing cells
(16). Therefore, it is a convenient entry point
for studying and altering NE function. The LC
provides the principal NE innervation to the
cerebral and cerebellsr cortices, the limbic
system, and brain stem and spinal cord regions
(17-19). Afferent projections to the LC come
from almost as many areas including the
reticular formation, adjacent central gray, some
forebrain areas, other catecholamine nuclei in
the brain stem (20), sensory brain stem nuclei,
and pain-sensitive neurons in the dorsal horns
of the spinal cord (21-23). Physiological and
biochemical studies confirm the feasibility of
discretely increasing or decreasing brain NE
function by stimulating or lesioning this
nucleus.
-
- However, there are reminders that these
methods have their limitations. Besides the
obvious problem of lesioning or stimulating
adjacent structures or fibers of passage,
manipulations may alter other systems
secondarily and may last for a limited period of
time (29-31). It is also merely an inference
that the presumed deficit after lesions will
reveal the physiologic functions of the intact
system throughout its range of activity. Some LC
stimulation-associated effects are not blocked
by 6-hydrozydopemine (6-OHM) destruction of the
dorsal NE bundle (DB) (32) or of the LC itself
(33). And, even NE-neuron mediated effects will
alter other transmitter systems in a living
-brain, leading to the confusion of primary and
secondary effects. It is clear, therefore, that
multiple approaches combining neurophysiologic
and pharmacologic techniques are necessary to
relate functional effects with specificity to
the LC or to NE function end that this
specificity is always elusive and potentially
illusory. We have attempted, therefore, to look
broadly for behavioral correlates of locus
coeruleus activity in a higher old world primate
species, by combining neurophysiologic and
phermacologic techniques to achieve some degree
of specificity.
-
- Effects of increases or decreases of LC
function in monkeys: Unilateral LC or DB aimed
electrodes have been correctly placed (15) in 16
monkeys for studies of locus coeruleus
activation. Low intensity stimulation of these
electrodes produces alerting, manifested by
widening of the palpebral fissures and increased
body movements which are most obvious in a
drowsy animal. With increasing intensity of
stimulation, chewing mouth and tongue movements
and teeth grinding appear along with grasping of
the chair, scratching, selfmouthing,
yawning, hair-pulling, hand wringing,
escape-struggling, and spasmodic single body
jerks. None of these behaviors are stimulus
bound nor leterelized to the stimulated side.
Vocalizations and facial grimaces, usually
accompanying painful stimuli, were not seen. The
behaviors are not unusual and might easily have
been ignored since they occur frequently in
chair-restrained monkeys at times unrelated to
electrical stimulation. Similarities were noted
originally between these behaviors and those
following direct threatening confrontations by
humans, leading to the suggestion that these
behaviors were related to increased fear or
anxiety (34). Some of these behaviors, in
particular, opening and closing of the mouth,
scratching, and yawning have been
previously noted in field studies of this
species to be associated with situations of
conflict, uncertainty, or impending aggression
(35). Bilateral LC lesions decreased the naturel
occurrence of some of these behaviors in a
social group situation (36), and all were
decreased in response to similar direct
threatening confrontation by humane (37). These
experiments led to a preliminary grouping of the
behaviors just described which were increased
either by LC stimulation or by human threats,
which we have labelled as Group I behaviors.
Other behaviors which appeared to be independent
of the changes in LC function are defined as
Group II behaviors. These include: head and body
turning, facial grimaces, vocalization,
lipsmacking, moving the hands, manipulating
objects or the chair, self-grooming, and
threatening facial gestures. Group III behaviors
appeared to be inversely related to LC
activation: freezing without observable motion
for 5 seconda, eyes partly closed for 5 seconda,
or eyes fully closed for 5 seconds. The effects
of low intensity electrical field stimulation of
LC on these behaviors by 3 monkeys are compared
in Fig. la. Group I behaviors are more
strikingly affected than Group II or III.
-
- Pharmacological agents with known single
neuronal unit effects on LC neurons produced
behavioral effects which were consistent with
those of electrical stimulation or lesions. The
alpha-2 adrenergic antagonist piperoxane is an
example of the effects of an agent which
activates the LC, probably by antagonizing NE or
E-sensitive auto-receptors (38-42) without
activating adjacent non-noradrenergic neurons.
The quantitative and qualitative effects are
similar to those produced by electrical
stimulation of the LC (Fig. lb).
-
- Other agents studied so far which have Group
I behavioral effects consistent with their
single unit effects are yohimbine (increase)
(43), morphine (decrease) (44,45), enkephalin
(decrease) (45), and clonidine (decrease) (46).
D,Lpropranolol, which blocks NE systems at
beta-adrenergic receptors (18,40), also
decreases the saine group of behaviors and
partially blocks the effects of electrical
stimulation of the LC. Diazepam, which decreases
NE turnover (47), decreases the saine group of
behaviors in monkeys (Fig. le) in non-sedative
doses comparable to those used clinically in
humans. Diazepam also diminishes the effect of
LC stimulation on the same behaviors (48). Other
agents which decrease LC activity and have not
yet been studied in monkeys are dopamine,
glycine, GABA, and serotonin (40,44,45,49,50).
Acetylcholine, glutamate, and substance p
(49-51) excite LC neurons and would be predicted
to increase the same behaviors.
-
- Fear stimuli increase the same behaviors
increased LC activation Since the original
classification of behaviors was derived
empirically from studies of LC stimulation, w
have attempted to determine whether these
behaviors were altered by a variety of stimuli
which are associated with fear or anxiety, using
increases in these behaviors as a kind of
"operational" definition of anxiety. We
previously noted the almost identical responses
to human threat gestures and to LC stimulation.
These monkeys also show increases in the same
behaviors in response to threats by
conspecifica, suggesting that they are not
artefacts of conditioning or of chair-restraint.
Natural fear-inducing stimuli, however, have the
disadvantage that they are difficult to control
in intensity and quality. We have, therefore,
also looked at a Pavlovian conditioning method
which has been supposed to induce fear (52-55).
This procedure (56,57) was used to study
behavioral effects of conditioning with a light
stimulus paired with an unconditioned electrical
leg shock (Fig. 2). After 5 days of 2 hours/day
of training, there were clear-cut increases in
Group I and Group II behaviors during the second
condition (W), compared with the
baseline-control level when no shocks ever
occurred (B). There was a trend toward a further
increase in Group I behaviors during the third
condition (R) always preceding shock by two
minutes, end a significant decrease occurred in
both Groups I and II behaviors during the shock
(S) when this period was adjusted to match the
durations of (W) and (R). Group III behaviors
decreased progressively during each period after
baseline (B). The qualitative and quantitative
increases in Group I behaviors were comparable
to the effects of LC activation by field
stimulation or by piperoxane (Figs. la,
lb).
-
- These results appear to support the
association of an empirically determined group
of behaviors (Group I) both with alterations in
LC activity and with a situation supposed to
induce fear. The same group of behaviors are
affected by pharmacologic agents known to
increase or diminish LC function, and to
increase after natural threats, chair restraint,
and a stimulus previously paired with a noxious
electrical shock. It is our epistemological
belief that this hypothesis can only be tested
in humans, based on pharmacologic agents with
identifiable effects on these emotions or tested
by the accurate prediction of effects in humans
based on pharmacologic actions on NE systems in
experimental animals. Although our studies in
this area are only beginning and many obvions
experiments yet undone, the hypothesis seams
already to have had some predictive velue in
humans, and is consistent with an extensive
animal experimental literature. In the remainder
of our discussion we will attempt to examine
this literature and further outline a postulated
NE system function in its context.
-
- Pharmacological physiological and
behavioral evidence relevant to a locus
coeruleus-anxiety hypothesis
-
- Effects in humans The NE (and LC) neuronal
"activators" piperoxane and yohimbine have been
reported to cause anxiety in humans (58-60) as
have a number of agents which interact somewhat
more inconsistently with NE systems,
particularly with beta-adrenergic receptors
(61). Electrical stimulation in the region of
the LC in humans produces feelings of fear and
imminent death (62). In the other direction,
heroin, morphine, and other opiates have
powerful anti-fear and anti-anxiety effects (63)
after acute administration, and both D, L
propranolol (64) and diazepam (65) have clearly
demonstrated anxiolytic properties. All of these
have anti-NE effects. In addition ethanol and
barbiturates are widely used by humans, at least
in part, for their antianxiety effects; and both
also decrease NE function (66,67). Several major
classes of compounds which decrease anxiety in
humans, therefore, have known mechanisms of
interaction with the LC and presumably other
brain NE nuclei. Based on the powerful effects
of clonidine in suppressing LC activity in the
rat (46) and LC stimulation effects in the
monkey, we predicted that clonidine should have
anxiolytic actions in humans (48).
-
- Speculative functions of known structures In
addition to the human pharmacologic evidence,
there are confirming physiologic data based on
anatomical projections of the LC which provide
one-synapse innervation of neuroanatomical
structures associated with specific physiologic
correlates of anxiety or fear (68). Briefly,
these include HYPOTHALAMIC, MEDULLARY, and
SPINAL SYMPATHETIC AREAS (tachycardia,
tmchypnea, hypertension (69,70), piloerection,
gastrointestinal hypermotility, urination,
defecation), the CEREBELLUM (tremor), the
RETICULAR FORMATION (arousal-sleep) (71), OTHER
HYPOTHALAMIC NUCLEI (neuroendocrine changes,
such as increased ACTH secretion [72],
and effects on appetite [73]) and LIMBIC
and CORTICAL AREAS (which might subserve such
functions as conscious awareness of the affect
and or alterations in memory mechanisms (74,75).
Interactions with PARASYMPATHETIC SYSTEMS (dry
mouth, gastric acid secretion) and NUCLEI
CONTAINING OTHER NEUROTRANSMITTERS provide the
neuroanatomic basis for widespread effects on
other systems. Similarly, the LC receives
numerous afferents which might activate this
system in response to stimuli or conditions
which might be expected to elicit fear or
anxiety. SPINAL CORD AFFERENTS come from large
neurons in the dorsal horns, which are known to
be sensitive to noxious stimuli (21) and which
give rise to fibers which do not terminate in
specific sensory areas of the thalamus (22)
providing a direct pathway for activation by
painful stimuli (41). Afferents from SOLITARY
TRACT NUCLEI and from other NE and epinephrine
containing cell groups in the MEDULLA (23) might
provide enteroceptive information from
peripheral sympathetic systems which are
activated by the LC, consistent with clinical
data suggesting that the perception of
physiological manifestations of anxiety may
increase it. The LIMBIC SYSTEM AFFERENTS
described (23) may also activate the LC if
cortical areas identify non-painful stimuli that
were previously learned to be potentially
noxious, or decrease LC activity if novel
non-noxious stimuli are correctly identified
(76).
-
- These possible functions, associated with
clearly-described neuroanstomy, are outlined to
illustrate that the LC is uniquely situated to
subserve almost all of the known physiological
correlates of anxiety or fear, as well as to
have direct pathways for activation under
circumstances where activation would be
predicted. Where the predicted physiological
pathways have been studied, the effect is
consistent with known correlates of anxiety or
fear. These correlates of LC function also
suggest that this nucleus may be necessary but
not sufficient to subserve the emotions of fear
or anxiety which also require other brain areas
and the autonomic nervous system for their
complete expression. In addition, the
differences between the effects of lesions end
pharmacological or electrical activation of the
LC at various levels in monkeys suggest that,
although anxiety or fear may be a part of LC
function, its normal function is much wider and
more complex - a broadly functioning "alarm
system" rather than a neural "substrate" for
anxiety or fear alone.
-
- Evidence from animal "models of anxiety or
fear: Many animal models have been proposed
based on intuitive "operational" definitions of
fear or anxiety, or on suppositions about the
effects of fear on performance. One example of a
classical animal model of fear may serve to
illustrate some of the problems with these
models. Since pain reliably induces both fear
and avoidance (52), it has been suggested that
fear is the prime motivator of avoidance (77).
One might predict, therefore, that animals with
drug-induced decreases in fear would not avoid
aversively conditioned stimuli (53,54). However,
phenothiazine compounds which decrease avoidance
or escape in conditioned avoidance (e. g.
"one-way active" or "two way" avoidance tests)
and therefore would be predicted to have
anti-fear or anti-anxiety properties (78) do not
effectively reduce human anxiety (79) ; while
the anxiolytic benzodiazepines facilitate rather
than diminish performance on the same tests
(80). The problem is complicated and confounded
by drug-induced alterations in sensation,
general alertness, and motor capacity on
particular tests, making it uncertain whether
any effect is due to fear-reduction. But, in
retrospect, one might question the initial
implicit assumption that fear is the ONLY
motivator of avoidance. Would an animal or human
without fear endure avoidable pain? Or under
some circumstances might the physiology of fear
even interfere with, rather than facilitate,
avoidance? Bilateral LC lesioned monkeys
threatened with an electrical shock device do
not show increases in Group I behaviors, nor
increase their heart rates (81), scream,
struggle, urinate, or defecate as do normale,
but will still avoid shock if given the
opportunity (37).
-
- Another prediction from the notion that fear
is the motivator of avoidance would be that LC
electrical stimulation would be avoided, or at
least not selfactivated, if the LC were
associated with fear. Studies in rodents and
monkeys suggest that, contrary to this
prediction, animals will self-stimulate
electrodes implanted close enough to activate LC
neurons (82), with one report indicating that
only placements which activated LC neurons
supported self stimulation (83). Most
investigators agree that the rates and the
pattern of stimulation are different from those
seen in other brain areas such as the
hypothalamus. Some evidence indicates that
projections from the LC to other
neurotransmitter systems or adjacent neuronal
systems are responsible for the aspects of
stimulation that are reinforcing, since neither
destruction of the dorsal NE bundle or of the LC
extinguishes self-stimulation (32,84) but DA
blockers do (85). These studies suggest that,
although the LC brain NE system may not be a
necessary neural substrate for "reward" (82), it
is at least not sufficiently aversive at the
intensities studied to neutralize the
"rewarding" aspects of self-stimulation in the
region. Again, one might also question the
assumption that fear is always aversive. Many
humans engage repetitively in activities which
are fear provoking, such as riding roller
coasters, climbing high mountains, and jumping
from airplanes with parachutes. Whether the fear
generated in these activities becomes dysphoric
and aversive seems to depend on the proximity of
actual danger and, to some extent, on the
intensity of the emotion. Whether animals or
humans would prefer a reduction or an increase
in the function of a brain "alarm system" seems
also to depend on the current level of activity
and the situation, with both extremes being
avoided.
-
- Effects of NE depletion in rats: NE deficits
produced by the neurotoxin, 6-OHM, fail to have
certain effects on a variety of behavioral tests
(31,86-96) which have been interpreted as
showing that ME systems are uninvolved in
anxiety or fear (87-92). It is not clear what
empirical or pharmacological support there is
for many of these interpretations. With a few
exceptions (92) the results reported with 6-OBDA
are consistent with the effects of barbiturates
(BTs), benzodiszepines (8Es), and ethanol (ET)
on the same procedure. (JO). BTs, ET, or BZs
have inconsistent effects on rewarded behavior
and the conditioned suppression of behavior
(80). They do not generally affect escape
behavior from unconditioned aversive stimuli or
one-way active avoidance (80). All of these
drugs improve performance on two-way active
avoidance presumably due to reduced freezing
responses to footshock (80), as reported for
6-OHDA dorsal bundle treated rats by Mason and
Fibiger, who also reported reduced freezing by
NE depleted rats (87). The increases in
aggressive behaviors after 6-ORnA treatment
reported by Crow et al (91) and File et al (92)
have also been reported with low doses of BT5,
SZs, and ET (80) and are similar to the effects
of LC lesions in monkeys. NE depletion in rats
is reported by some (24,86,94) but not all
investigators (95) to produce decreased running
speeds for food reward in a "straight alley,"
consistent with an effect of 8Es (80). One might
also interpret a number of the findings with DE
lesioned rats as being consistent with reduction
of an alarm system related to fear or anxiety.
Increased exploratory activity (89), increased
contact time with a novel object (89), and
decreased disturbance of a licking behavior by a
tone (91) might all be consistent with
diminished function of an alarm system. LC
lesioned (96) or 6-ORDA treated rats (31) which
spent more time in the center of an open-field
test ambulating (31) were described, somewhat
anthropomorphically, as "inattentive and
overconfident, walking all around the enclosure
while making few observing responses" (31). The
investigator also noted that these behaviors
after 6-ORnA changed with time after treatment,
with some significant recoveries by 40 days
(31). This time period from 12 to 40 days after
treatment, characterized by a variety of changes
in NE turnover and alterations in NE-receptors,
is precisely the period studied so extensively
in rats after 6-ORDA. This produces uncertainty
as to the functional consequences of lesions,
which led us to shift our emphasis to
stimulation and pharmacologic studies which
could produce reversible effects at varying
intensities.
-
- Pharmacologic studies of NE function in
rodents: Nearly all testing of a NEfear
hypothesis in rats has been done with
electrolytically or 6-ORnA lesioned animals. Our
data suggest that the LC function is not
monotonically related to anxiety or fear; and
although fear-related effects can be interpreted
in the deficit syndrome, these effects are more
clearly identifiable during increased function.
With the exception of the self-stimulation
studies mentioned, we know of no other studies
of behavioral effects of LC stimulation. Other
animal "models" would also he more interpretable
if opposite effects could he demonstrated with
agents which increase fear or anxiety in humans
from those effects obtained with various classes
of human anxiolytics. Davis et al (97-99) have
found effects of increased as well as decreased
neuronal function of NE systems using amplitude
measurements of a startle reflex that is
augmented by the presentation of a loud tone in
the presence of a light previously paired with a
shock, known as "potentiated startle" (100).
This response is also related to shock-Intensity
in a non-monotonic fashion (101), and is reduced
by the anxiolytics (cf 61) sodium amytal (100),
diazepam (98), flurazepaa (98), morphine (99),
and propranolol (97). The response is also
decreased by clonidios, which decreases NE
neuronal activity, and increased by piperoxane
and yohimbine (97) which increase it. Since all
of the pharmacologic effects on "potentiated
startle" are consistent with those seen in our
monkey studies with agents also affecting human
anxiety, a specific study of the effects of
electrolytic or 6-ORnA lesions of LC on this
measurement would be of interest. Locus
coeruleus lesions were reported by Geyer, et al,
to decrease the magnitude of unconditioned
startle responses to air puff stimuli (102),
suggesting that results might be similar to the
effects of anxiolytic drugs.
-
- An operant procedure in which on-going lever
pressing for food reward is inhibited by
electrical shocks has been studied extensively
because the anxiolytic drugs disinhibit the
response suppression (103,104), while a second
unpunished lever, which is rewarded at a lower
density, serves to control for non-specific
effects. This Geller-Seifter "conflict" test"
(103) has successfully identified a number of
new anxiolytic compounds, and is sensitive to
the effects of BTa, ET in some tests, BZs (cf
80), and recently to clonidine (105). This test
is sensitive to irrelevant changes in motor,
appetitive, and sensory functions; and perhaps
for these reasons, it fails to clearly identify
the anxiolytic properties of propranolol
(106,107), morphine (108) and haloperidol (109).
Nonetheless, the effects on this procedure of
agents which increase LC function, such as
piperoxane, yohimbine, or substance P would also
be of interest as would the effects of LC or DB
lesions.
-
- An empirical correlation has been reported
between the effects of several classes of
anxiolytics and a 7.7 Hz hippocampal theta
rhythm (80), which provided the basis for Gray
to identify the NE systems as involved in
anxiolytic drug effects (110) since similar
effects were also produced by NE synthesis
inhibitors or lesions of the dorsal NE bundle
with 6-OHDA. Because of the large noradrenergic
innervation of the hippocampus from the LC
(17,18), this phenomenon may be one effect of
many LC projections that are relevant to anxiety
mechanisms. A large thoroughly reviewed
literature on the behavioral pharmacology of the
catecholamines also suggests that NE depletion
by synthesis inhibitors produces similar changes
in behavioral tests to those produced by the
anxiolytic drugs (111), but the agents used are
less specific than those discussed above.
-
- Does the LC or central NE function increase
duriq fear? In the rat the LC system reponds
with increased single unit firing to novel
stimuli (41,76), although it rapidly decreases
its response to non-noxious stimuli even in
anesthetized animals, while noxious stimuli
produce the same or incremented responses (41).
In awake rats (76) and squirrel monkeys (112)
this noveltydetecting function is maintained.
These responses are consistent with the "alarm
function" which has been suggested, but noxious
stimuli or stimuli conditioned to noxious
stimuli have yet to be studied. The biochemical
corollary of increased NE neuronal function
during noxious stimuli or anxiety provoking
situations has been supported by considerable
data, which shows that stress decreases static
measures of brain NE (cf 113), most likely the
result of increased activity in NE neurons
(113,27). Recent studies have demonstrated
increases in brain MG during conditioned fear in
rodents (114). In depressed humans, urinary
HIIPC (115) and CSF MHPG (116) correlate
positively with anxiety ratings.
-
- Implications of the thesis of an LC-alarm
function for treatment of medical
problems
-
- Anxiety: The most obvious clinical
implications relate to mechanisms and treatments
to produce or relieve anxiety. Pharmacologic
agents which increase brain NE function should
produce fear or anxiety as the limited evidence
so far suggests for piperoxane, yohimbine and
beta-adrenergic agonists (61). Agents which
decrease net NE function should have anxiolytic
properties (61). Clonidine, and similar chemical
analogs and alpha-2 adrenergic agonists, should
have potent anxiolytic properties (48,117).
Patients with the phobic anxiety syndrome, who
do not usually respond well to benzodiszepines
(79), should be improved by clonidine. New
agents might be utilized which deliberately
affect several modulatory receptors with similar
actions on NE neurons, thereby improving
effectiveness and perhaps decreasing undesirable
effects. Or agents might be deliberately
synthesized to have more specific effects for
certain NE receptors.
-
- Abstinence syndromes The recent clinical
demonstration of suppression of signs and
symptoms of opiate abstinence by clonidine
(118,119) was based on interactions of
clonidine, piperoxane, and morphine with LC
stimulation in the monkey (120). The opiate
abstinence syndrome resembles locus coeruleus
stimulation in many respects, and reproduces
many of the signs ad symptoms of anxiety or
fear. Since clonidine was known to suppress
these signs and behaviors associated with LC
stimulation (48), and since a detailed neuronal
circuitry (11,121,122) supports opiate (or
endorphin) - NE interactions (119), clonidine
was tested for effects directly against the
opioid abstinence syndrome. Since the initial
clinical study, the predicted increases in NE
activity during opiate abstinence have been
confirmed: increased neuronal activity in LC in
rats (123), and increased MIII'S in monkey
brain, CSF, and plasma and in human plasma
(124). Clonidine might also be useful in other
withdrawal syndromes in which anxiety-like
changes are prominent-- those due to chronic
benzodiazepines, barbiturates, ethanol, and
nicotine.
-
- Analgesia: The neuronal circuitry and recent
specific data on analgesic effects of LC lesions
(125) or clonidine (126-128) suggest that some
"analgesic" effects of opioids, which suppress
NE neuronal activity, may be mediated by
decreased NE activity. Combinations of receptor
agents having similar effects at different
receptors might increase analgesia, as has been
shown for the combination of amphetamine and
morphine which was discovered empirically
(129,130) with no understanding of its probable
action via alpha-2 and opiate receptors.
-
- Stress-anxiety exacerbated diseases A number
of conditions which anxiety and "stress" factors
are thought to exacerbate, such as burette's
syndrome (131), "benign essential tremor,"
tardive dyskinesia, and functional bowel
diseases such as duodenal ulcer, ulcerative
colitis, irritable colon syndrome, and chronic
nonspecific diarrhea (132), might benefit from
treatment with agents designed to interact with
specific NE receptors, similar to the effects of
clonidine. Longterm blockade of aspects of LC
activation may also have beneficial effects on
arteriosclerotic cardiovascular disease and
essential hypertension. The risk factors of
"stress," type "A" personality, and chronic
anxiety are well recognized in the etiology of
hypertension, myocardial infarction, and
cerebral thrombosis and infarction, as are the
benefits of "stress" reduction. Many detailed
anatomic and physiologic connections from LC
activation (recently reviewed, cf 61) can be
made to the pathophyaiclogy of arteriosclerosis,
including increases in blood pressure, increased
ACTH, increased circulating NE, gluconeogenoeia,
impaired carbohydrate tolerance,
hyper-beta-lipidemia, increased formation of
triglycerides, effects on betalipoprotein
release, activation of alpha-adrenoreceptora on
platelets which may lead to increased
thrombogenicity, as a basis for the known
increased tendency for the blood to clot. The
predicted benefits of blockade of these chronic
and usually unnecessary effects of activation of
an LC alarm system might be considerable, and
provide an additional rationale for the clinical
use ni clonidine, proprannlnl, or
anti-nnradrenergic agents besides their effects
on hypertension or or cardiac arrhythmiae.
-
- Conclusions
-
- We have summarized experiments in monkeys
suggesting that a group of behaviors is affected
almost identically by increases in LC activity
produced either by electrical field stimulation
or by alpha-2 NE antagonists. The same behaviors
are increased by a variety of stimuli which
might induce fear. Bilateral LC lesions reduce
expression of these behaviors similar to the
effects of clonidine, morphine, diazepam, and D,
L propranolol. These agents are thought to
reduce LC activity via alpha-2 adrenergic,
opiate, or GABA aomatodendritic receptors or by
post synaptic beta-adrenergic blockade
respectively. Since the alpha-2 antagonists
which increase LC activity have been reported to
induce fear-anxiety syndromes in humane, and the
agents which reduce net LC fonction all have
anxiolytic activity in humane, we suggest that
brain NE systems, such as the LC, are involved
in the production of fear or anxiety, and that
they are a major mechanism of action for
auxiolytic drugs. Anatomical and functional
evidence especially supports the unique location
of the nucleus LC for being involved in the
physiological manifestations of anxiety or fear.
The same evidence, however, also suggests that
the LC is not sufficient for the manifestation
of these emotions which require numerous other
systems in the central nervous system.
-
- These studies also suggest that anxiety or
fear is only a part of the function of the brain
NE systems. LC lesions and LC stimulation at
many levels reveal a wide range of functions.
The system's absence or diminished function,
perhaps anhedonic, eight produce certain
liabilities- the failure to withdraw promptly in
the face of danger or the inability to inhibit
certain responses or to learn from certain kinds
of new experience (95,88). At low levels, it
might be a novelty detector, a stimulus
enhancer, or an attention focuser. In a middle
or normal range of function, such a system would
generally be "cautionary," and would be
associated with improved chances for survival.
At higher levels the system continues to respond
to all novel stimuli, differentially amplifies
and distinguishes noxious from non-noxious
stimuli, and helps to prepare for a survival
struggle; but its non-monotonic and modulatory
behavioral effects may range from freezing to
flight. An "alarm system" might be a more
adequate description of this function. The
system has sensory nerve, spinal neuronal, and
forebrain efferents which can modulate its
function directly or in response to interpreted
information, and has efferent connections to the
anatomic structures known to mediate the "fight"
or "flight" response. The term "alarm system"
also conveys an adaptive nature, the deficit
which would be present if the system were
missing, and the unpleasantness of its ringing
loudly or too long. In normal individuels, the
emotions associated with increasing activation,
therefore, might be called attention,
carefulness, interest, surprise, alerting,
astonishment, alarm, anxiety, fear, panic, or
terror. Which was identified would depend on s
number of situational, semantic, intensity, or
personality factors. Clinical or "morbid"
anxiety might be the result of alterations in
the operation of this system. Pharmacologic
agents aimed more specifically at this system
might improve the treatment of anxiety, drug
abstinence syndromes, pain, and some
psychosomatic diseases.
-
|
