. 44
( 51 .)


administration of a CRH antagonist reduces freezing and anxious behavior on sev-
eral tests or symptoms of fear (Koob et al., 1993), and attenuates fear-potentiated
startle (Swerdlow et al., 1989). An increase (or sensitization) in CRH in the brain
occurs after abuse, maternal deprivation, and exposure to other stressful situations
in macaques (Habib et al., 1999). A severe social stressor in macaques is the addi-
tion of an unknown intruder in an adjacent cage. This results in behaviors associ-
ated with anxiety and fear such as body tremors, grimacing and teeth gnashing, and
CRH expression is increased in this situation (Habib et al., 2000). Administration of a
CRH antagonist reduces the fear and anxiety displays, increases exploratory behav-
iors, and reduces the production of CRH during the stressful situation (Habib et al.,
2000). Importantly, lesions of the CeA, and not the PVN, disrupt CRH-potentiated
conditioned fear responses (Liang et al., 1992). That is, only lesions of the
246 J. Schulkin et al.


CSF CRH (pg/ml)



Pre- Post-
novelty novelty

Figure 8.7 Levels of CRH in the CSF of macaques in response to a familiar (pre-novelty) and
unfamiliar (post-novelty) object. Adapted from Habib et al. (2000)

amygdala and not of the hypothalamus disrupt the behavioral response, suggesting
that CRH induced or facilitated fearful behaviors are generated through extra-
hypothalamic brain regions independently of the role of CRH in the HPA axis
(Figure 8.7).
High levels of systemic glucocorticoids are associated with fear (or the percep-
tion of adverse events) in a number of species (e.g. Mason, 1975; Breier, 1989;
Jones et al., 1992), and may be essential for the formation for some forms of fear
conditioning (Pugh et al., 1997). In one set of experiments rats (adrenal intact)
were pretreated with corticosterone to investigate whether it facilitated condi-
tioned fear-induced freezing (Coordimas et al., 1994). All rats received condition-
ing trials in which the unconditioned stimulus (footshock) was presented
concurrently with the conditioned stimulus (auditory tone). Several days after the
trials the rats were treated with corticosterone. The same treatment of cortico-
sterone that increased CRH gene expression in the CeA and bed nucleus of the stria
terminalis also facilitated conditioned fear-induced freezing in rats (Coordimas
et al., 1994).
In a subsequent study (Thompson et al., 2004), contextual fear conditioning was
investigated in groups of rats that were chronically treated with corticosterone or
given a vehicle treatment. CRH expression was differentially regulated in the CeA
and the parvocellular region of the PVN. One week after the completion of the con-
ditioning and the last corticosterone injection, the rats were tested for the retention
of conditioned fear. The corticosterone treated rats displayed more fear condition-
ing than the vehicle treated rats. The data suggest that repeated high levels of cor-
ticosterone could facilitate the retention of contextual fear conditioning, perhaps
by the induction of CRH gene expression in critical regions of the brain such as the
247 Glucocorticoid facilitation of CRH in the placenta and the brain

As noted above, CRH facilitates startle responses. This response does not depend
on the adrenal glands because centrally delivered CRH facilitates startle responses
in the absence of the adrenal glands (Lee et al., 1994). In that study, Lee et al.
demonstrated that high chronic plasma levels of corticosterone in adrenal intact
rats facilitated CRH-induced startle responses (Lee et al., 1994). Perhaps what
occurs normally is that the glucocorticoids, by increasing CRH gene expression,
increase the likelihood that something will be perceived as a threat, which results in
a startle response. Thus, a dose of CRH, given intraventricularly, did not produce a
startle response, but when the adrenal intact rats were maintained at high levels of
corticosterone for several days prior to the CRH injection, the same dose did pro-
duce a startle response.
Implants of corticosterone directly into the amygdala of rats increased CRH
expression in the CeA and reduced their open ¬eld exploratory behavior (Shepard
et al., 2000). Typically, rats initially are hesitant to explore new environments, and
the induction of CRH in the CeA following corticosterone delivery to the amyg-
dala exacerbated this characteristic. In addition, corticosterone implants directly
into the CeA increased levels of CRH expression in the parvocellular region of the
PVN of the hypothalamus (Shepard et al., 2003).
An important study further demonstrated that the CRH response in the amyg-
dala of sheep to a natural (dog) and unnatural (footshock) stressor is regulated by
glucocorticoids (Cook, 2002). Following acute exposure to a dog for 6 min, both
venous and amygdala levels of cortisol increased after 10“30 min. Amygdala CRH
had a large increase during exposure to the dog and a second peak 10“30 min later
corresponding to the increase in cortisol. Similar dual peaks of CRH release also
were found with footshock. Administration of a glucocorticoid receptor antago-
nist blocked the second CRH peak in the amygdala without affecting the ¬rst peak.
These data indicate that the initial response of CRH in the amygdala to an acute
fearful stimulus is independent of cortisol, but the second delayed peak is cortisol
dependent. In addition, and most interesting, the initial CRH response to a stres-
sor following repeated inescapable exposure to the dog came under the control of
cortisol. Sheep were given 7 days of repeated exposure to the dog, either with the
ability to escape or not to escape from the dog. On the eighth day, the sheep were
given a footshock. While venous and amygdala cortisol levels in response to the
footshock were identical in escape and non-escape groups, both peaks of CRH
release in the amygdala were higher in the repeated non-escape group compared to
the escape group and became regulated by cortisol (Figure 8.8).
We interpret these ¬ndings to indicate that during normal acute danger, CRH in
the amygdala increases rapidly to participate in mounting fear responses. This
response is similar to effects of exogenously applied CRH and is not under the con-
trol of glucocorticoids. However, with repeated stress, glucocorticoids sensitize the
248 J. Schulkin et al.


Amygdala CRH (% of baseline)
Amygdala CRH (% of baseline)




Non-escape group

Non-escape group
Escape group
Escape group 100

0 5 10 15 20 25 30 35 40 45 50 55 60
0 5 10 15 20 25 30 35 40 45 50 55 60
Time (min) Time (min)

(a) (b)
Forelimb shock Forelimb shock

Figure 8.8 Facilitated CRH response in the amygdala of sheep to a stressor (footshock) following
inescapable exposure to a dog is blocked by a glucocorticoid receptor antagonist. (a)
The CRH (collected by microdialysis) in the amygdala of sheep exposed to a footshock is
greater following inescapable experience with dog, and (b) Mifepristone, a glucocorticoid
receptor antagonist, blocks the effects of inescapable exposure to a dog. Adapted from
Cook (2002)

amygdala CRH cells so they release exaggerated amounts of CRH to the adverse
event. The psychological stressor of inescapable, repeated danger produces an up-
regulation of the CRH amygdala system. Taken together with other experimental
data, these results demonstrate that high levels of glucocorticoids increase CRH
mRNA expression in the CeA (Swanson and Simmons, 1989; Makino et al., 1994a, b;
Watts and Sanchez-Watts, 1995; Thompson et al., 2004).
The CRH receptors within the amygdala are largely found in the lateral but not
in the central region of the amygdala; the central nucleus produces the peptide, the
lateral region contains the receptors (e.g. Makino et al., 1995; Behan et al., 1996). It
should be noted that the basal lateral region of the amygdala is essential for most
forms of fear (Le Doux, 2000).
CRH in the central nucleus is produced under diverse conditions; CRH receptor
antagonists decrease the behavioral effects of CRH production in the CeA
(Roozendaal et al., 2002). The basal lateral region is importantly involved in memory
consolidation of aversive events. Infusion of glucocorticoids into this region of the
amygdala facilitates the memory of aversive events. CRH type I (see below) receptor
blocker infusions into the basal lateral region reduce the expression of the aversive
memory and CRH gene expression in the CeA (Roozendaal et al., 2002). Thus, the
effect of cortisol on memory consolidation may perhaps affect CRH gene expression
in the CeA, or elsewhere in the brain (e.g. lateral bed nucleus of the stria terminalis).
Part 3 Glucocorticoids and postnatal

There appear to be both pre- and postnatal critical periods in development, and
these critical periods differ among the species. In some species (e.g. rats), the sen-
sitivity to glucocorticoids and the regulation of the HPA axis varies with age
(Levine, 1975; 2000; Levine et al., 2000). Alteration of corticosterone levels during
critical stages of postnatal development has effects on behavior. For example, rats
deprived of corticosterone between 10 and 14 days post partum do not express the
normal fear of unfamiliar objects; infusion of corticosterone either systemically
or centrally restores or facilitates the behavioral responses (Takahashi and Kim,
1994). Perhaps this occurs via the induction of CRH gene expression in the brain.
However, excessive CRH injections in neonatal rats resulted in compromised brain
function and vulnerability to diverse forms of behavioral dysfunction (Brunson
et al., 2001).
Indeed, early life events have long-term consequences for both brain and behav-
ior and alter CRH expression in the brain (Meaney et al., 1993; Levine, 2000). For
example, adult rats, deprived of maternal closeness for 3 h a day for a 2-week period
as pups, were found to have higher levels of CRH mRNA expression in the PVN,
CeA and the lateral bed nucleus of the stria terminalis as adults than those sepa-
rated for only 15 min a day (Plotsky, 1996; Levine, 2000). These maternal-deprived
rats were also more likely to develop helpless behavior in uncontrollable aversive con-
texts suggesting that these rats were excessively stressed or fearful. Interestingly, their
systemic levels of corticosterone as adults were not different from normal rats, but the
central state of exaggerated fear induced by the early experience was long-lasting.
Infant monkeys reared by mothers experiencing unpredictable foraging con-
ditions had higher CRH in cerebrospinal ¬‚uid (CSF) in adulthood than infant
monkeys reared by mothers that had either a predictable overabundance or a pre-
dictable scarcity of food. The studies show that unpredictability in early life, and
not just chronic hardship, is associated with persistently higher CRH levels in the
CSF in adulthood, up to 5 years later (Coplan et al., 2001). Perhaps the induction of
CRH gene expression by cortisol partially explains why this occurs.
The lateral bed nucleus of the stria terminalis, a region of the brain rich in CRH
cell bodies (Swanson and Simmons, 1989; Makino et al., 1994a, b; Watts and
Sanchez-Watts, 1995), has been linked to general anxiety (Davis et al., 1997).
Infusing CRH in this region potentiates anxious arousal (Davis et al., 1997).
Importantly, glucocorticoids are known to facilitate increases in CRH gene expres-
sion in the lateral region of the bed nucleus of the stria terminalis (Makino et al.,
250 J. Schulkin et al.

1994a, b; Watts and Sanchez-Watts, 1995). The bed nucleus might be considered
the primary central ganglia of the PVN; it massively projects and is known to regu-
late CRH PVN release (Herman and Cullinan, 1997). Moreover, the central nucleus
and other regions of the amygdala have access to the PVN largely through the
amygdala innervation of the bed nucleus of the stria terminalis (Herman et al.,
2003). Therefore, perhaps the induction of CRH during these environmental
events contributes to the sense of unease, the exaggerated sense of arousal, alertness,
and uncertainty in the animal.

Temperamental shyness, cortisol, and CRH
Kagan and his colleagues have been instrumental in describing the origins and
developmental course of temperamental shyness in children over the last two
decades. More speci¬cally, Kagan™s group has been interested in variations in nor-
mal children™s reactions to novelty. Kagan™s group has noted that a subset of nor-
mally developing infants and children (5“10%) exhibit extreme fear and wariness
to the presentation of novel social and nonsocial stimuli, and this subset can be
described as behaviorally inhibited. Kagan et al. (1987; 1988) speculated that indi-
vidual differences in infant reactivity to novelty may be linked to sensitivity in fore-
brain circuits involved in the processing and regulation of emotion, and they argued
that children who become easily distressed and subdued during the presentation of
novel stimuli may have a lower threshold for arousal in forebrain areas, particularly
the CeA. This hypothesis is based largely on ¬ndings from studies of animals in
which the amygdala plays an important role in the regulation and maintenance of
conditioned fear, as noted above.
Conceptually, shyness in humans might re¬‚ect a preoccupation with the self
in response to real or imagined social encounters (Kagan et al., 1988; Schmidt
and Schulkin, 1999). Although a large percentage (90%) of the population has
reported experiencing shyness at some point in their lives, a smaller percentage
(5“10%) of individuals are characterized by temperamental or dispositional shy-
ness. Temperamental shyness is an early emerging form of shyness that is linked to
early infant reactions to novelty, associated with a number of distinct psychophys-
iological responses at rest and in response to social stress, remains modestly pre-
served through the young adult years, and is predictive of social and emotional
Preschool-aged children with temperamental shyness generally have increased
levels of cortisol (Kagan et al., 1988; Gunnar et al., 1989; Schmidt et al., 1997;
Figure 8.9(a)). They are more fearful in response to novel social events. It was sug-
gested some time ago that this exaggerated fear might re¬‚ect a ˜hyperactive amyg-
dala™ (Kagan et al., 1988; see also Rosen and Schulkin, 1998). Children at 21 months
of age were assessed as having an inhibited or uninhibited temperament, and
251 Glucocorticoid facilitation of CRH in the placenta and the brain

salivary cortisol (ug/dl)

Average morning



. 44
( 51 .)