Exogenous prenatal corticosterone exposure mimics the effects of prenatal stress on adult brain stress response systems and fear extinction behavior

Article (PDF Available)inPsychoneuroendocrinology 38(11) · August 2013with30 Reads
DOI: 10.1016/j.psyneuen.2013.07.003 · Source: PubMed
Abstract
Exposure to early-life stress is a risk factor for the development of cognitive and emotional disorders later in life. We previously demonstrated that prenatal stress (PNS) in rats results in long-term, stable changes in central stress-response systems and impairs the ability to extinguish conditioned fear responding, a component of post-traumatic stress disorder (PTSD). Maternal corticosterone (CORT), released during prenatal stress, is a possible mediator of these effects. The purpose of the present study was to investigate whether fetal exposure to CORT at levels induced by PNS is sufficient to alter the development of adult stress neurobiology and fear extinction behavior. Pregnant dams were subject to either PNS (60min immobilization/day from ED 14-21) or a daily injection of CORT (10mg/kg), which approximated both fetal and maternal plasma CORT levels elicited during PNS. Control dams were given injections of oil vehicle. Male offspring were allowed to grow to adulthood undisturbed, at which point they were sacrificed and the medial prefrontal cortex (mPFC), hippocampus, hypothalamus, and a section of the rostral pons containing the locus coeruleus (LC) were dissected. PNS and prenatal CORT treatment decreased glucocorticoid receptor protein levels in the mPFC, hippocampus, and hypothalamus when compared to control offspring. Both treatments also decreased tyrosine hydroxylase levels in the LC. Finally, the effect of prenatal CORT exposure on fear extinction behavior was examined following chronic stress. Prenatal CORT impaired both acquisition and recall of cue-conditioned fear extinction. This effect was additive to the impairment induced by previous chronic stress. Thus, these data suggest that fetal exposure to high levels of maternal CORT is responsible for many of the lasting neurobiological consequences of PNS as they relate to the processes underlying extinction of learned fear. The data further suggest that adverse prenatal environments constitute a risk factor for PTSD-like symptomatology, especially when combined with chronic stressors later in life.

Figures

Exogenous
prenatal
corticosterone
exposure
mimics
the
effects
of
prenatal
stress
on
adult
brain
stress
response
systems
and
fear
extinction
behavior
Brian
C.
Bingham
a,1
,
C.S.
Sheela
Rani
a,1
,
Alan
Frazer
a,b,1
,
Randy
Strong
a,b,1
,
David
A.
Morilak
a,1,
*
a
Department
of
Pharmacology
and
Center
for
Biomedical
Neuroscience,
University
of
Texas
Health
Science
Center
at
San
Antonio,
7703
Floyd
Curl
Drive,
San
Antonio,
TX
78229,
United
States
b
Research
Service,
South
Texas
Veterans
Health
Care
Network,
Audie
L.
Murphy
Division,
7400
Merton
Minter
Drive,
San
Antonio,
TX
78229,
United
States
Received
8
May
2013;
received
in
revised
form
22
June
2013;
accepted
10
July
2013
Psychoneuroendocrinology
(2013)
38,
2746—2757
KEYWORDS
Corticosterone;
Fear
conditioning;
Fear
extinction;
Glucocorticoids;
Post-traumatic
stress;
Disorder;
Prenatal
stress;
Stress
vulnerability;
Tyrosine
hydroxylase
Summary
Exposure
to
early-life
stress
is
a
risk
factor
for
the
development
of
cognitive
and
emotional
disorders
later
in
life.
We
previously
demonstrated
that
prenatal
stress
(PNS)
in
rats
results
in
long-term,
stable
changes
in
central
stress-response
systems
and
impairs
the
ability
to
extinguish
conditioned
fear
responding,
a
component
of
post-traumatic
stress
disorder
(PTSD).
Maternal
corticosterone
(CORT),
released
during
prenatal
stress,
is
a
possible
mediator
of
these
effects.
The
purpose
of
the
present
study
was
to
investigate
whether
fetal
exposure
to
CORT
at
levels
induced
by
PNS
is
sufficient
to
alter
the
development
of
adult
stress
neurobiology
and
fear
extinction
behavior.
Pregnant
dams
were
subject
to
either
PNS
(60
min
immobilization/day
from
ED
14—21)
or
a
daily
injection
of
CORT
(10
mg/kg),
which
approximated
both
fetal
and
maternal
plasma
CORT
levels
elicited
during
PNS.
Control
dams
were
given
injections
of
oil
vehicle.
Male
offspring
were
allowed
to
grow
to
adulthood
undisturbed,
at
which
point
they
were
sacrificed
and
the
medial
prefrontal
cortex
(mPFC),
hippocampus,
hypothalamus,
and
a
section
of
the
rostral
pons
containing
the
locus
coeruleus
(LC)
were
dissected.
PNS
and
prenatal
CORT
treatment
decreased
glucocorticoid
receptor
protein
levels
in
the
mPFC,
hippocampus,
and
hypothalamus
when
compared
to
control
offspring.
Both
treatments
also
decreased
tyrosine
hydroxylase
levels
in
the
LC.
Finally,
the
effect
of
prenatal
CORTexposure
on
fear
extinction
behavior
was
examined
following
chronic
stress.
Prenatal
CORT
impaired
both
acquisition
and
recall
of
cue-conditioned
fear
extinction.
This
effect
was
additive
to
the
impairment
induced
by
previous
chronic
stress.
Thus,
these
data
suggest
that
fetal
exposure
to
high
levels
of
maternal
CORT
is
responsible
for
*
Corresponding
author.
Tel.:
+1
210
567
4174,
fax:
+1
210
567
4300.
E-mail
address:
morilak@uthscsa.edu
(D.A.
Morilak).
1
For
the
STRONG
STAR
Consortium.
Available
online
at
www.sciencedirect.com
jou
rn
a
l
home
pag
e
:
ww
w.
el
sev
ier.
com/
loca
te
/psyn
eu
en
0306-4530/$
see
front
matter
#
2013
Elsevier
Ltd.
All
rights
reserved.
http://dx.doi.org/10.1016/j.psyneuen.2013.07.003
1.
Introduction
Post-traumatic
stress
disorder
(PTSD)
is
a
disabling
affective
disorder
that
occurs
as
a
consequence
of
a
physically
or
emotionally
traumatic
experience.
It
is
characterized
by
intru-
sive
memories,
a
state
of
hyper-vigilance,
and
an
inability
to
inhibit
fear
responses
to
trauma-associated
cues.
An
estimated
9
million
people
in
the
United
States
suffer
from
PTSD,
yet
this
is
only
a
portion
of
those
who
experience
trauma
(Kessler
et
al.,
2005).
Therefore,
other
factors
are
likely
to
confer
vulnerability
to
developing
PTSD
subsequent
to
traumatic
stress,
including
experiential,
environmental,
or
biological
predispositions.
Clinical
studies
suggest
that
early
life
stres-
sors,
such
as
childhood
exposure
to
trauma,
low
socioeconomic
status,
and
familial
instability
increase
susceptibility
to
PTSD
later
in
life
(Breslau
et
al.,
1999;
Koenen
et
al.,
2007).
Prenatal
stress
(PNS)
is
an
adverse
early
life
event
that
has
been
associated
with
increased
risk
for
anxiety,
ADHD,
schizophre-
nia,
developmental
delays,
and
hypothalamic—pituitary—
adrenal
(HPA)
axis
dysregulation
in
humans;
however,
very
little
is
known
about
its
role
as
a
potential
risk
factor
for
PTSD
(Davis
and
Sandman,
2012;
Talge
et
al.,
2007).
Animal
studies
suggest
that
PNS
programs
the
adult
stress
system
to
create
a
stress-reactive
phenotype,
showing
fear-
and
depression-like
behaviors
that
resemble
aspects
of
PTSD
(Weinstock,
2008).
PNS
has
been
shown
to
permanently
pro-
gram
the
brain
corticosteroid
and
brain
monoamine
systems,
both
of
which
are
implicated
in
the
formation
and
extinction
of
fear
memories.
PNS
can
reduce
glucocorticoid
receptor
(GR)
and/or
mineralocorticoid
receptor
(MR)
expression
in
adult
offspring
(Brunton
and
Russell,
2010;
Green
et
al.,
2011;
Harris
and
Seckl,
2011;
Weinstock,
2008).
PNS
also
alters
catecho-
lamine
release
in
brain
areas
associated
with
behavioral
and
cognitive
components
of
the
stress
response.
It
has
been
shown
to
decrease
basal
and
stress-induced
norepinephrine
release
in
the
prefrontal
cortex
(PFC)
and
locus
coeruleus
(LC)
as
well
as
dopamine
in
the
LC
(Carboni
et
al.,
2010;
Takahashi
et
al.,
1992).
In
addition
to
these
biochemical
effects,
PNS
causes
enduring
behavioral
changes,
including
anxiety-like
behavior
on
the
elevated
plus
maze,
an
increase
in
freezing
behavior
following
footshock,
and
compromised
performance
in
cogni-
tive
tasks
like
the
Morris
water
maze
(Brunton
and
Russell,
2010;
Kofman,
2002;
Salomon
et
al.,
2011;
Takahashi
et
al.,
1992;
Weinstock,
2008).
This
altered
physiological
and
behavioral
response
to
stress
may
create
a
state
of
vulnerability
to
chronic
stressors
later
in
life
and
thus
increase
the
risk
for
PTSD.
Indeed,
after
experiencing
a
traumatic
stress,
many
individuals
continue
to
endure
a
secondary,
persistent
state
of
chronic
stress
that
is
produced
by
intrusive
memories,
nightmares,
and
increased
physiological
stress
responses
to
cues
associated
with
the
initial
trauma.
In
individuals
who
may
be
impaired
in
their
ability
to
cope
with
stress,
this
persistent
chronic
stress
may
facilitate
the
transition
from
acute
stress
disorder
to
chronic
PTSD
(Davidson
and
Baum,
1986;
Wessa
and
Flor,
2007).
Likewise,
chronic
stress
may
also
impair
the
ability
to
extin-
guish
trauma-associated
fear
memories.
Preclinical
studies
in
rodents
have
demonstrated
that
chronic
stress
in
adulthood
facilitates
conditioned
fear
behavior
and
impairs
the
retention
of
subsequent
extinction
of
conditioned
fear
(Farrell
et
al.,
2010;
Garcia
et
al.,
2008).
We
recently
showed
that
both
PNS
and
adult
chronic
stress
independently
impaired
acquisition
of
fear-extinction.
These
effects
appeared
to
be
additive,
such
that
rats
receiving
both
PNS
and
adult
chronic
stress
were
consistently
the
most
impaired
in
their
ability
to
extinguish
fearful
associations,
a
hallmark
trait
of
PTSD
in
humans
(Green
et
al.,
2011).
Fetal
exposure
to
maternal
glucocorticoids
represents
one
potential
mechanism
whereby
PNS
may
program
the
adult
stress
response
in
utero.
During
PNS,
glucocorticoids
are
released
by
the
dam,
and,
at
high
concentrati on,
can
cross
the
placental
barrier
to
exert
direct
effects
on
gene
transcrip-
tion
in
the
fetus
(Harris
and
Seckl,
2011;
Takahashi
et
al.,
1998).
Fetal
exposure
to
high
levels
of
glucocorticoids
results
in
long-term
impairments
in
cognitive
and
emotional
regulation
(Alexander
et
al.,
2012).
Both
the
direct
administration
of
glucocorticoids
and
the
inhibition
of
placental
barrier
enzymes
mimic
some
of
the
effects
of
PNS
(Welberg
et
al.,
2000).
Likewise,
maternal
adrenalectomy
is
able
to
prevent
some
of
the
lasting
effects
of
PNS
(Barbazanges
et
al.,
1996;
Salomon
et
al.,
2011).
However,
it
is
unknown
if
prenatal
glucocorticoid
exposure
mimics
the
effects
of
PNS
on
the
formation
and
extinction
of
fear
memories
in
the
adult
offspring.
It
is
also
unknown
whether
a
history
of
prenatal
glucocorticoid
exposure
interacts
with
later
stress
to
further
impair
fear
extinction,
i.e.,
creating
vulnerability
for
a
PTSD-like
phenotype.
To
address
these
questions,
we
compared
the
effects
of
prenatal
corti-
costerone
(CORT)
administration
in
the
absence
of
maternal
stress
to
those
of
PNS.
We
first
determined
a
dose
of
exogenous
CORT,
delivered
to
the
mother,
that
best
mimics
both
fetal
and
maternal
circulating
CORT
levels
induced
by
PNS.
To
determine
if
CORT
treatment
mimics
the
neurobiological
consequences
of
PNS,
we
then
measured
the
mRNA
and
protein
expression
of
the
GR,
corticotrophin
releasing
factor
(CRF),
brain-derived
neu-
rotrophic
factor
(BDNF),
and
tyrosine-hydroxylase
(TH)
in
the
brains
of
the
adult
male
offspring.
Finally,
we
measured
the
effect
of
prenatal
CORT
exposure
on
fear
conditioning
and
extinction
behavior
in
the
adult
offspring,
with
and
without
exposure
to
chronic
stress.
We
hypothesized
that
prenatal
CORT
exposure
would
mimic
the
neurobiological
effects
of
PNS
and
create
an
additive
detrimental
effect
on
fear
con-
ditioning
and
extinction
behavior
when
combined
with
chronic
stress
later
in
life.
2.
Methods
2.1.
Animals
Timed-pregnant
female
Sprague-Dawley
rats
(Harlan,
Indianapolis)
arrived
on
embryonic
day
(ED)
6
and
were
many
of
the
lasting
neurobiological
consequences
of
PNS
as
they
relate
to
the
processes
underlying
extinction
of
learned
fear.
The
data
further
suggest
that
adverse
prenatal
environments
constitute
a
risk
factor
for
PTSD-like
symptomatology,
especially
when
combined
with
chronic
stressors
later
in
life.
#
2013
Elsevier
Ltd.
All
rights
reserved.
Fetal
corticosterone
and
adult
stress
neurobiology
2747
single-housed
in
standard
Plexiglas
cages
(25
cm
45
cm
15
cm)
on
a
12/12
h
light—dark
cycle
(lights
on
at
7:00
h)
with
food
and
water
available
ad
libitum.
On
postnatal
day
(PD)
5,
litters
were
culled
to
eight
pups
each,
maximizing
the
number
of
males.
Upon
weaning
(PD
21),
male
pups
were
housed
2—3
per
cage
with
littermates
until
PD
45,
at
which
time
they
were
single-housed.
In
total,
154
adult
male
off-
spring
(from
30
litters:
11
stressed,
9
CORT-treated,
10
control)
were
used
in
these
experiments.
An
additional
32
females
were
sacrificed
at
E16
or
E20
to
provide
maternal
and
fetal
CORT
levels.
For
the
social
defeat
procedure,
12
adult
male
Long-Evans
rats
(Harlan),
weighing
at
least
400
g,
were
used
as
defeaters.
Each
resident
male
was
housed
in
a
large
cage
(80
cm
55
cm
40
cm)
with
an
ovariectomized
female,
in
a
separate
room
from
the
experimental
colony.
All
experi-
ments
were
conducted
during
the
light
phase.
All
procedures
were
conducted
according
to
NIH
guidelines
for
the
care
and
use
of
laboratory
animals
and
were
reviewed
and
approved
by
the
Institutional
Animal
Care
and
Use
Committee
of
The
University
of
Texas
Health
Science
Center
at
San
Antonio.
All
efforts
were
made
to
minimize
animal
pain,
suffering,
or
discomfort,
and
to
minimize
the
number
of
rats
used.
2.2.
Prenatal
stress
From
ED14
to
ED21,
stressed
females
were
injected
daily
with
either
sesame
oil
vehicle
(1.4
ml/kg,
sc.
Sigma—Aldrich)
or
saline
vehicle
(4.5
ml/kg,
sc.),
depending
on
the
experiment.
They
were
then
immobilized
for
60
min.
They
were
held
gently
but
firmly
on
a
flat
rack
while
the
limbs,
shoulders,
and
hips
were
taped
to
the
rack.
The
midsection
was
not
taped
to
avoid
putting
physical
pressure
on
the
fetuses.
The
animals
were
unable
to
move,
but
respiration
was
unhin-
dered
and
they
were
not
enclosed
to
avoid
hyperthermia.
Unstressed
females
were
injected
with
either
CORT
(10
mg/
kg,
sc.
Sigma—Aldrich),
saline,
or
oil
vehicle
and
returned
to
their
home
cages.
This
dose
was
established
in
pilot
studies
to
approximate
stress
CORT
levels.
2.3.
Measurement
of
fetal
corticosterone
levels
Dams
from
each
treatment
group
were
sacrificed
by
rapid
decapitation
immediately
after
1
h
immobilization,
or
1
h
fol-
lowing
CORTor
vehicle
injection
on
day
ED16
or
ED20.
Maternal
trunk
blood
was
collected
in
chilled
15
ml
conical
tubes
con-
taining
100
ml
of
0.5
M
EDTA.
Immediately
thereafter
(<1
min),
pups
were
removed
by
Cesarean
section
and
rapidly
decapi-
tated.
For
the
E20
fetuses,
approximately
70
ml
of
trunk
blood
was
collected
from
each
fetus
(n
=
1—3/litter)
with
a
hepar-
inized
capillary
tube,
and
deposited
into
ice-cold
eppendorf
vials.
Because
of
the
small
fetal
blood
volume
at
E16,
each
sample
(n
=
1—2/litter)
represents
blood
pooled
from
3
sibling
fetuses,
approximately
70
ml
total.
After
collection,
blood
samples
were
centrifuged
at
4
8C
(3000
g
for
15
min)
and
the
plasma
removed
and
stored
at
20
8C.
Plasma
CORT
levels
were
determined
via
radioimmunoassay
as
previously
described
(Roth
et
al.,
2012).
2.4.
Chronic
plus
acute
prolonged
stress
(CAPS)
To
investigate
the
interaction
between
prenatal
CORT
expo-
sure
and
behavioral
susceptibility
to
chronic
stress
later
in
life,
offspring
in
each
treatment
condition
were
subjected
to
chronic
plus
acute
prolonged
stress
(CAPS)
treatment
from
PD
46—48
to
PD
60—62.
The
chronic
component
of
the
CAPS
procedure
entailed
14
days
of
chronic
intermittent
cold
stress.
The
rats
were
transported
in
their
home
cage,
with
food,
bedding,
and
water,
into
a
cold
room
(4
8C,
6
h/day)
for
14
consecutive
days.
The
acute
component
on
day
15
con-
sisted
of
3
acute
stressors
administered
sequentially
in
a
single
1-h
session:
social
defeat
(20
min),
immobilization
(30
min),
and
forced
swim
(10
min).
For
social
defeat,
the
ovariectomized
Long-Evans
female
was
removed
from
the
resident
cage,
and
the
test
rat
was
placed
into
the
resident
cage.
After
the
resident
Long-Evans
male
rat
attacked
and
defeated
the
test
rat,
defined
by
the
test
rat
assuming
a
submissive
posture
for
at
least
4
s,
the
test
rat
was
placed
under
a
wire
mesh
cage
for
20
min,
protecting
it
from
further
physical
contact
but
allowing
continued
sensory
interaction.
Immobilization
involved
taping
the
torso,
head,
and
limbs
gently
but
firmly
in
a
prone
position
on
a
flat
platform,
allowing
no
movement
for
30
min.
For
swim
stress,
the
rat
was
placed
in
a
cylindrical
tank
(30
cm
diameter
60
cm
height)
filled
to
a
depth
of
30
cm
with
water
at
approximately
23
8C.
2.5.
Fear
conditioning
and
extinction
Fear
conditioning
and
extinction
were
performed
as
pre-
viously
described
with
minor
modification
(Green
et
al.,
2011).
One
day
after
the
termination
of
CAPS
treatment
(or
the
comparable
time
point
for
controls),
rats
were
habi-
tuated
to
two
contexts
for
15
min
each.
Twenty-four
hours
after
habituation,
the
rats
received
cued
fear
conditioning
in
Context
A,
a
shock
chamber
with
metal
walls
and
a
grid
floor.
Each
rat
was
placed
into
the
chamber
and,
after
a
5
min
acclimation
period,
experienced
four
pairings
of
a
tone
(10
kHz,
75
dB,
20
s)
co-terminating
with
a
shock
(0.8
mA,
0.5
s,
average
inter-trial
interval
=
120
s).
Extinction
training
occurred
3
days
later.
The
rats
were
placed
in
Context
B,
a
similar
chamber
but
with
smooth
vinyl
floors
and
walls
to
avoid
contextual
freezing.
They
were
exposed
to
16
trials
of
the
tone
alone,
with
an
average
inter-trial
interval
of
2
min.
On
the
following
day,
the
rats
were
returned
to
Context
B,
and
the
retention
of
extinction
was
tested
by
presenting
them
with
16
additional
tones.
Behavior
during
each
stage
was
video-recorded
and
freezing
behavior
during
each
tone
presentation
was
analyzed
off-line
using
the
FreezeFrame
and
FreezeView
software
(Coulbourn
Instruments
#ACT-100).
Freezing
was
defined
as
behavior
below
a
motion
index
threshold
of
10
lasting
at
least
1
s.
2.6.
Tissue
collection
Rats
were
sacrificed
on
PD
65—67
by
rapid
decapitation.
The
medial
prefrontal
cortex
(mPFC),
hippocampus,
hypothala-
mus,
and
the
pontine
area
containing
the
locus
coeruleus
(LC)
were
quickly
dissected
using
a
brain
matrix
on
ice,
as
described
previously
(Green
et
al.,
2011).
The
brain
samples
were
frozen
on
dry
ice
and
stored
at
80
8C
until
use.
TH
mRNA
and
protein
were
analyzed
in
the
LC
samples;
mRNA
and
protein
for
GR
in
mPFC,
hippocampus,
and
hypothalamus
samples;
mRNA
and
protein
for
CRH
in
hypothalamus
2748
B.C.
Bingham
et
al.
samples;
and
mRNA
and
protein
for
BDNF
in
hippocampus
samples.
2.7.
mRNA
analyses
Total
RNA
was
isolated
from
each
brain
region
and
converted
to
cDNA
as
described
previously
(Green
et
al.,
2011).
Real-
time
PCR
was
performed
using
the
following
Taqman
gene
expression
assays
(Applied
Biosystems/Life
Technologies,
Carlsbad,
CA):
rat
TH,
Rn00562500_m1;
rat
GR
(NR3C1),
Rn00561369_m1;
rat
CRH,
Rn01462137_m1;
and
rat
BDNF,
Rn01484924_m1.
All
assays
consisted
of
intron-spanning
pri-
mers
and
FAM-labeled
probes.
Results
were
normalized
using
the
eukaryotic
18S
rRNA
endogenous
control
assay
(4319413E)
labeled
with
VIC
dye.
Assays
were
performed
in
triplicate
after
validating
with
the
ABI
Prism
7900
HT
instrument
and
following
the
MIQE
guidelines.
Real
time
PCR
data
were
analyzed
by
the
2
DDCt
method.
Relative
expression
of
the
gene
of
interest
in
treatment
groups
was
expressed
as
percent
of
control.
2.8.
Protein
analyses
TH
protein
in
the
LC
region
was
analyzed
by
Western
blot
as
described
previously
(Green
et
al.,
2011).
GR
protein
was
analyzed
using
an
ELISA
kit
(TransAM
GR
kit,
Active
Motif,
Carlsbad,
CA)
and
protein
levels
were
computed
from
A450
values.
CRH
protein
levels
in
hypothalamus
or
hippocampus
were
assayed
using
an
extraction-free
Enzyme
Immunoassay
kit
(Phoenix
Pharmaceuticals,
Burlingame,
CA).
Similarly,
BDNF
protein
levels
in
the
hippocampus
were
assayed
using
a
rat
BDNF
ELISA
kit
(Syd
Labs
Inc.,
Malden,
MA)
following
the
manufacturer’s
protocol.
All
results
were
expressed
as
a
percent
of
the
oil-treated
non-stressed
control
mean.
2.9.
Data
analysis
and
statistics
Maternal
and
fetal
CORT
measures
were
analyzed
by
ANOVA
at
each
age,
with
Newman—Keuls
Multiple
Comparison
post-
tests
where
significant
effects
were
revealed.
All
adult
neu-
rochemical
measures
were
analyzed
by
ANOVA
with
Dunnett’s
post-test
used
for
comparison
to
the
vehicle
control
group.
For
the
fear
conditioning,
extinction,
and
retention
data,
group
differences
in
percent
freezing
were
first
analyzed
for
each
session
by
a
three-way
ANOVA
(prenatal
CORT
adult
stress
tone)
with
repeated
measures
over
tone.
In
addi-
tion,
total
freezing
during
extinction,
represented
by
the
mean
area
under
the
extinction
curve,
was
analyzed
by
2-way
ANOVA
with
Bonferroni
post-tests
for
pairwise
comparisons.
Subsequently,
to
better
assess
and
compare
the
rate
and
extent
of
extinction
across
groups,
the
freezing
data
for
all
animals
within
a
group
were
best
fit
to
an
unconstrained,
single-exponential
decay
function
using
Graphpad
Prism
5.
As
reported
previously
(Green
et
al.,
2011),
freezing
typically
increased
from
tone
1
to
tone
2
in
the
extinction
session.
Therefore,
tone
1
was
not
included
in
the
regression
analysis,
so
that
the
extinction
rate
could
be
calculated
from
the
point
of
maximum
freezing.
From
the
resulting
regression
equa-
tions,
the
decay
constant
(k),
plateau
value,
and
their
stan-
dard
errors
(SE)
were
derived
for
each
group.
Differences
between
groups
were
analyzed
using
an
extension
of
Cochran’s
Q
methodology
(Cochran,
1954)
which
partitioned
the
overall
Chi-square
(df
=
3)
into
independent
factor
com-
ponents
according
to
a
2
(prenatal
CORTexposure)
2
(adult
stress)
design.
The
Q
statistics
were
then
transformed
to
F
values
as
described
(Cochran,
1954).
3.
Results
3.1.
Maternal
and
fetal
corticosterone
levels
Both
PNS
and
CORT
treatment
significantly
increased
mater-
nal
CORT
levels
(Fig.
1A)
measured
at
E16
(F
(3,15)
=
58.07;
p
<
0.01)
and
E20
(F
(3,15)
=
60.89;
p
<
0.01)
when
compared
to
the
unstressed
saline
and
oil-treated
controls
(
p
<
0.01).
At
E16,
there
were
no
differences
in
the
plasma
CORT
levels
of
PNS
dams
compared
to
CORT-treated
dams;
however,
at
E20,
the
PNS-induced
plasma
CORT
level
was
slightly
higher
than
that
of
the
CORT-treated
dams
(
p
<
0.01).
There
were
no
differences
in
CORT
levels
between
saline-
and
oil-treated
dams
at
either
age.
As
with
the
dams,
PNS
and
CORT
treat-
ment
also
increased
fetal
CORT
levels
(Fig.
1B)
at
E16
(F
(3,25)
=
18.71;
p
<
0.01)
and
E20
(F
(3,39)
=
10.41;
p
<
0.01)
when
compared
to
the
respective
unstressed
sal-
ine-
and
oil-treated
control
groups
at
both
ages
(
p
<
0.01).
There
were
no
differences
in
fetal
plasma
CORT
between
saline-
and
oil-treated
groups,
or
between
PNS
and
CORT
groups
at
either
age.
As
expected,
fetal
CORT
increased
from
Figure
1
Maternal
(A)
and
fetal
(B)
plasma
CORT
levels
following
a
60
min
immobilization
stress
or
60
min
after
acute
CORT
injection
on
embryonic
day
ED16
or
ED20.
Both
PNS
and
CORT
increased
plasma
CORTcompared
to
the
respective
control
treatments
in
both
the
dams
and
fetuses
at
each
day
(**p
<
0.01).
On
E20,
PNS
induced
a
slightly
larger
increase
than
CORT
treatment
in
the
dams
(
#
p
<
0.01);
however,
there
were
no
differences
in
the
corresponding
fetal
levels.
(n)
=
number
of
samples
per
group.
Fetal
corticosterone
and
adult
stress
neurobiology
2749
E16
to
E20,
as
the
fetal
adrenal
glands
began
producing
endogenous
CORT
(Dupouy
et
al.,
1975).
This
is
also
reflected
in
the
slight
elevation
of
CORT
in
control
dams
from
E16
to
E20.
As
there
were
no
differences
in
CORT
levels
between
unstressed
saline-
and
oil-injected
controls
in
the
first
experi-
ment,
only
oil-injected
animals
were
used
as
controls
in
subsequent
experiments.
Neither
PNS
nor
CORT
treatment
altered
litter
size
or
the
percentage
of
male
pups
per
litter.
PNS
did,
however,
result
in
a
significant
decrease
in
maternal
weight
gain
from
E14
to
E21,
i.e.,
during
the
period
of
daily
stress
treatment,
when
compared
to
both
the
oil-
and
CORT-
treated
dams
(Table
1).
3.2.
GR
mRNA
and
protein
expression
in
the
adult
mPFC,
hippocampus
and
hypothalamus
Adult
expression
of
GR
mRNA
(Fig.
2A—C)
and
protein
(Fig.
2D—F)
were
measured
in
the
mPFC,
hippocampus,
and
hypothalamus.
One-way
ANOVA
for
mRNA
expression
revealed
significant
treatment
effects
in
all
three
brain
regions
(mPFC:
F
(2,33)
=
5.76,
p
<
0.01;
hippocampus:
F
(2,31)
=
8.74,
p
<
0.01;
hypothalamus:
F
(2,35)
=
3.53,
p
<
0.05).
Post-hoc
analyses
showed
that,
in
the
mPFC
and
hypothalamus,
GR
mRNA
was
reduced
in
the
CORT-treated
group
but
not
in
the
PNS
group
(Fig.
2A
and
C).
In
the
hippocampus,
both
PNS
and
prenatal
CORTexposure
resulted
in
a
significant
reduction
in
GR
mRNA
expression
compared
to
controls
(Fig.
2B).
GR
protein
levels
were
also
significantly
affected
by
the
prenatal
treatment
(for
mPFC:
F
(2,21)
=
24.27;
for
hippocampus:
F
(2,30)
=
26.85;
for
hypotha-
lamus:
F
(2,37)
=
15.61;
all
p
<
0.01).
Further,
post-hoc
com-
parisons
showed
that
GR
protein
expression
was
significantly
reduced
in
all
three
brain
regions
by
both
PNS
and
CORT
treatment
compared
to
controls
(Fig.
2D—F).
3.3.
CRH
expression
in
the
adult
hypothalamus
Both
PNS
and
prenatal
CORT
treatment
resulted
in
a
signifi-
cant
reduction
in
CRH
mRNA
expression
in
the
hypothalamus
(F
(2,35)
=
7.36;
p
<
0.01;
Fig.
3A),
accompanied
by
a
40—60%
decrease
in
CRH
protein
(F
(2,39)
=
23.14;
p
<
0.01;
Fig.
3D).
Table
1
Effects
of
PNS
and
CORT
on
litter
size,
percent
males,
and
maternal
weight
gain
during
treatment.
Treatment
group
Litter
size
(pups)
%
Male
(pups/litter)
Maternal
weight
gain
E14—21
(g)
Oil
9.7
1.5
57.6
8.8
71.1
6.9
PNS
12.3
1.1
54.0
5.3
55.1
5.2
a,b
CORT
11.6
1.1
55.2
4.8
86.3
3.9
a
p
<
0.05
vs.
oil.
b
p
<
0.05
vs.
CORT.
Figure
2
Expression
of
GR
mRNA
and
protein
in
the
adult
offspring
of
control,
PNS,
and
CORT
treated
dams.
Prenatal
CORT
treatment
decreased
GR
mRNA
expression
in
the
mPFC
(A),
hippocampus
(B)
and
hypothalamus
(C).
PNS
decreased
GR
mRNA
in
the
hippocampus
only.
Both
PNS
and
CORT
treatment
decreased
GR
protein
in
all
three
regions
(D—F).
*p
<
0.05,
**p
<
0.01
vs.
control,
(n)
=
number
of
samples
per
group.
2750
B.C.
Bingham
et
al.
We
also
measured
CRH
mRNA
expression
in
the
adult
hippo-
campus
and
found
no
change
in
either
of
the
prenatal
treat-
ment
groups
compared
to
controls
(data
not
shown).
3.4.
BDNF
expression
in
the
adult
hippocampus
No
significant
changes
in
BDNF
mRNA
expression
were
seen
in
the
adult
hippocampus
in
either
treatment
group
(F
(2,28)
=
0.69;
p
=
0.5;
Fig.
3B).
However,
a
modest
but
sig-
nificant
reduction
in
BDNF
protein
level
was
seen
in
the
adult
hippocampus
of
animals
exposed
to
prenatal
CORT
treat-
ment,
but
not
to
PNS
(F
(2,45)
=
6.14;
p
<
0.01;
Fig.
3E).
3.5.
TH
expression
in
the
adult
locus
coeruleus
ANOVA
revealed
a
significant
difference
in
the
expression
of
TH
mRNA
(F
(2,33)
=
9.67;
p
<
0.01;
Fig.
3C)
and
protein
(F
(2,17)
=
48.27;
p
<
0.01;
Fig.
3F)
in
the
region
of
pons
contain-
ing
the
LC.
Post-hoc
analyses
indicate
that
both
PNS
and
CORT
treatment
decreased
TH
expression
when
compared
to
control.
3.6.
Fear
conditioning
and
extinction
Having
established
that
prenatal
CORT
treatment
reproduces
many
of
the
effects
of
PNS
on
several
neurobiological
mea-
sures
in
adults,
we
next
determined
the
impact
of
prenatal
CORT
treatment
on
fear
conditioning
and
extinction
behavior
following
adult
CAPS
stress
(Fig.
4).
CAPS
stress
alone
increased
freezing
behavior
during
fear
conditioning
with
no
significant
effect
of
prenatal
CORT
(Fig.
4A).
A
three-
way
ANOVA
with
repeated
measures
indicated
main
effects
only
for
CAPS
(F
(1,54)
=
5.99;
p
<
0.05)
and
Tone
(F
(3,162)
=
82.64;
p
<
0.01).
By
contrast,
both
CAPS
and
prenatal
CORT
significantly
delayed
the
extinction
of
conditioned
fear.
A
three-way
ANOVA
for
freezing
behavior
during
extinction
learning
(Fig.
4B)
indicated
main
effects
of
prenatal
CORT
(F
(1,54)
=
5.22;
p
<
0.05),
CAPS
(F
(1,54)
=
10.03;
p
<
0.01),
and
Tone
(F
(15,810)
=
37.13;
p
<
0.01)
with
no
significant
interactions.
The
extinction
learning
impairment
was
also
evident
in
the
total
amount
of
freezing
displayed
during
the
extinction
procedure,
measured
by
the
area
under
the
extinction
curve
(AUC).
Two-way
ANOVA
for
AUC
again
revealed
main
effects
of
prenatal
treatment
(F
(1,54)
=
4.77;
p
<
0.05)
and
CAPS
(F
(1,54)
=
10.65;
p
<
0.01)
with
no
inter-
action
(Fig.
4C).
Further,
to
better
assess
the
rate
and
final
degree
of
extinction,
freezing
behavior
across
tones
was
fit
by
non-linear
regression
analysis
to
a
single
exponential
decay
curve
for
each
group
(Fig.
4D).
The
decay
constant
(k)
and
plateau
value
were
then
analyzed.
CAPS
treatment
significantly
reduced
the
decay
constant
(F
(1,87)
=
5.25;
p
<
0.05)
while
prenatal
CORT
tended
to
reduce
the
decay
constant
(F
(1,87)
=
2.67;
p
=
0.11)
(Fig.
4E).
Neither
factor
significantly
altered
the
plateau
parameter
(Fig.
4F).
Combined,
these
analyses
indicate
that
CAPS
and
prenatal
CORT
independently
impair
acquisition
of
extinction
learning
by
increasing
total
freezing
during
training
and
decreasing
the
rate
of
extinction
learning
without
impacting
the
final
level
of
asymptotic
freezing
behavior
reached
at
the
end
of
the
extinction
learning
session.
One
day
after
extinction
learning,
the
animals
were
tested
for
their
ability
to
recall
the
extinction
training
from
the
day
before
(Fig.
5).
CAPS
had
no
significant
effect
on
extinction
retention,
whereas
prenatal
CORT
treatment
sig-
nificantly
impaired
it
(Fig.
5A).
A
three-way
ANOVA
indicated
main
effects
of
prenatal
CORT
(F
(1,54)
=
7.26;
p
<
0.01),
Tone
(F
(15,810)
=
11.49;
p
<
0.01),
and
a
CORT
by
Tone
interaction
Figure
3
Expression
of
CRH,
BDNF,
and
TH
in
the
adult
offspring
of
control,
PNS,
or
CORT
treated
dams.
Both
PNS
and
CORT
treatment
decreased
CRH
and
TH
mRNA
and
protein
content
in
the
hypothalamus
(A
and
D)
and
LC
(C
and
F),
respectively.
CORT
treatment
decreased
hippocampal
BDNF
protein
with
no
effect
on
mRNA
(B
and
E).
*p
<
0.05,
**p
<
0.01
vs.
oil
control,
(n)
=
number
of
samples
per
group.
Fetal
corticosterone
and
adult
stress
neurobiology
2751
(F
(15,810)
=
2.14;
p
<
0.01).
Animals
treated
prenatally
with
CORT
also
demonstrated
significantly
higher
freezing
in
response
to
the
first
retention
tone
presentation
(Fig.
5B),
analyzed
by
2-way
ANOVA
(F
(1,54)
=
6.45;
p
<
0.05).
Analysis
of
AUC
as
a
measure
of
total
freezing
during
extinction
retention
also
showed
a
significant
effect
of
prenatal
CORT
(F
(1,54)
=
7.17;
p
<
0.01;
Fig.
5C).
When
freezing
behavior
across
tones
was
fit
to
a
single
exponential
decay
function
(Fig.
5D),
there
were
no
significant
effects
of
either
CORT
or
CAPS
on
the
rate
constant;
however,
there
was
a
main
effect
of
prenatal
CORT
treatment
on
the
plateau
(F
(1,87)
=
15.27;
p
<
0.01).
Combined,
these
analyses
indicate
that
CORT-treated
animals
are
able
to
re-extinguish
at
the
same
rate
as
controls;
however,
they
are
impaired
in
their
initial
extinction
recall,
and
they
are
ultimately
unable
to
extinguish
to
the
same
extent
as
controls
(Fig.
5E
and
F).
4.
Discussion
We
previously
demonstrated
that
PNS
decreases
GR
and
TH
expression
and
impairs
the
extinction
of
learned
fear
(Green
et
al.,
2011).
The
current
study
tested
the
hypothesis
that
fetal
exposure
to
stress-relevant
CORT
levels
is
sufficient
to
recreate
the
programming
effects
of
PNS
on
brain
stress
systems
that
also
mediate
fear
learning
and
extinction.
Pregnant
dams
were
subjected
to
either
PNS
or
daily
injec-
tions
of
CORT,
titrated
to
match
the
fetal
CORT
levels
induced
by
PNS.
The
offspring
were
then
allowed
to
grow
to
adulthood
for
neurochemical
and
behavioral
testing.
The
results
con-
firmed
and
extended
our
previous
findings.
Prenatal
CORT
exposure
was
sufficient
to
recapitulate
PNS-induced
decreases
in
GR
expression
in
the
mPFC,
hippocampus,
and
hypothalamus
of
the
adult
offspring.
Both
prenatal
treat-
ments
also
decreased
hypothalamic
CRH
and
TH
in
the
LC,
whereas
prenatal
CORT
alone
decreased
hippocampal
BDNF.
Prenatal
CORT
also
mimics
the
previously
described
effect
of
PNS
by
impairing
fear
extinction
and
retention,
alone
and
in
combination
with
subsequent
adult
stress.
With
both
pre-
natal
stress
and
exogenous
CORT
treatment,
circulating
fetal
CORT
levels
did
not
reach
those
measured
in
the
dams,
most
likely
due
to
placental
metabolism
of
some
proportion
of
maternal
CORT
before
it
could
reach
the
fetus,
by
the
enzyme,
11b-hydroxysteroid
dehydrogenase
type
2
(Seckl
and
Meaney,
2004).
Maternal
corticotrophin-binding
globulin
(CBG)
may
also
have
a
role
in
regulating
the
amount
of
circulating
CORT
available
to
impact
the
fetus,
and
prenatal
stress
can
decrease
maternal
CBG
levels
(Takahashi
et
al.,
1998).
However,
this
would,
if
anything,
increase
the
amount
of
free
CORT
to
which
the
fetuses
could
be
exposed
during
stress,
but
is
unlikely
to
have
altered
the
relative
levels
of
circulating
CORTafter
exogenous
administration.
At
any
rate,
similar
fetal
CORT
levels
were
achieved
in
both
conditions,
and
the
results
indicate
that
excessive
CORT
exposure
during
fetal
development
is
sufficient
for
many
of
the
neurochem-
ical
and
behavioral
consequences
of
PNS,
including
those
related
to
associative
fear
extinction.
Figure
4
Fear
conditioning
and
extinction
learning
following
prenatal
CORTand
adult
CAPS
treatment.
(A)
CAPS
stress
(dashed
lines)
significantly
increased
freezing
behavior
during
fear
conditioning.
(B)
Both
prenatal
CORT
and
adult
CAPS
treatment
significantly
delayed
the
extinction
of
cue-conditioned
fear
3
days
after
conditioning.
(C)
Both
PNS
and
CAPS
increased
total
freezing
behavior
during
extinction
as
measured
by
area
under
the
curve.
(D)
Extinction
data
fitted
to
a
single-exponential
decay
curve
for
each
group.
(E)
Analysis
of
the
decay
constants
(k)
derived
from
the
regression
lines
in
(D)
indicates
that
both
prenatal
CORTand
adult
CAPS
stress
reduced
the
rate
of
extinction
learning.
(F)
By
contrast,
analysis
of
the
plateau
value
indicates
no
effect
of
either
prenatal
CORT
or
CAPS
on
the
final
level
of
freezing
behavior
displayed
at
the
end
of
the
extinction
learning
session.
*p
<
0.05
main
effect
of
CAPS
vs.
control.
#
p
<
0.05
main
effect
of
prenatal
CORT
vs.
control.
(n)
=
number
of
subjects
per
group.
2752
B.C.
Bingham
et
al.
4.1.
Long-term
changes
in
gene
and
protein
expression
in
the
brain
Both
PNS
and
CORTsuppressed
the
expression
of
GR
in
multiple
areas
of
the
brain,
providing
evidence
that
fetal
CORT
expo-
sure
is
the
likely
mechanism
underlying
the
long-term
pro-
gramming
effect
of
PNS
on
brain
GR
expression.
PNS
and
prenatal
CORT
also
decreased
CRH
mRNA
and
protein
in
the
hypothalamus.
As
most
hypothalamic
CRH-containing
neurons
reside
in
the
paraventricular
nucleus
(PVN),
it
is
likely
that
the
decrease
occurred
primarily
in
those
HPA-related
neurons.
We
previously
found
that
PNS
increased
basal
CORT
secretion
in
the
adult
offspring
(Green
et
al.,
2011).
Therefore,
one
potential
interpretation
of
these
data
is
that
brain
GR
and
CRH
levels
are
down-regulated
as
a
consequence
of
life-long
exposure
to
elevated
basal
glucocorticoids.
Alternatively,
it
is
possible
that
the
decreases
in
GR
and
CRH
are
established
through
direct
epigenetic
programming
in
utero
and
that
the
increase
in
basal
CORT
is
a
compensatory
response.
While
changes
in
HPA
regulatory
proteins
are
not
always
noted
with
PNS
or
prenatal
glucocorticoids,
others
have
also
found
that
PNS
causes
a
decrease
in
negative
feedback
capability
and
an
increase
in
basal
or
evoked
glucocorticoid
release
(Barba-
zanges
et
al.,
1996;
Weinstock,
2008).
In
human
subjects,
this
pattern
of
CORT
release
and
putative
glucocorticoid
sensitivity
is
more
in
line
with
a
depressive-like
phenotype
than
the
traditional
PTSD-like
phenotype
(Yehuda
et
al.,
1991).
However,
two
factors
are
important
to
consider.
First,
depression
and
PTSD
are
highly
comorbid,
and
the
associated
HPA
activity
is
often
dependent
on
gender,
trauma,
and
developmental
stress
history
(Shea
et
al.,
2005).
Second,
reductions
in
GR
expression
in
the
mPFC
and
hippocampus
may
have
effects
on
stress-related
learning
and
memory
that
transcend
HPA
regulation.
GR
agonist
administration
in
the
prelimbic
subregion
of
the
mPFC
increases
inhibitory
avoidance
memory
and
systemic
administration
of
a
GR
antagonist
blocks
the
reconsolidation
of
fear
memory
(Pitman
et
al.,
2011;
Roozendaal
et
al.,
2009).
GR
activation
increases
the
surface
expression
of
both
NMDA
and
AMPA
receptors
in
the
mPFC,
facilitating
short-
term
memory
(Yuen
et
al.,
2009).
Activation
of
the
mPFC,
specifically
the
infralimbic
(IL)
subregion,
also
facilitates
extinction
learning
and
retention
(Morgan
et
al.,
1993;
Vidal-Gonzalez
et
al.,
2006).
Therefore,
it
is
possible
that
a
decrease
in
GR
expression
within
the
mPFC
and
hippocam-
pus
may
be
more
relevant
to
functional
impairment
of
extinction
learning
and
retention
than
to
HPA
regulation.
We
also
found
that
prenatal
CORT
caused
a
small
but
significant
decrease
in
BDNF
in
the
hippocampus.
Hippocam-
pal
BDNF
has
been
shown
to
play
an
important
role
in
extinction
learning
(Peters
et
al.,
2010).
Thus
the
decrease
in
BDNF
is
consistent
with
the
impairment
in
extinction
retention
in
rats
exposed
to
prenatal
CORT.
However,
the
reduction
in
BDNF
was
relatively
modest,
and
there
was
no
effect
in
PNS
animals.
Thus,
while
reduced
BDNF
expression
may
have
contributed
to
the
extinction
deficit,
it
may
not
be
Figure
5
Effects
of
prenatal
CORT
and
adult
CAPS
treatment
on
the
retention
of
extinction.
(A)
Prenatal
CORT
(black
lines)
significantly
increased
freezing
behavior
during
extinction
retention
testing
24
h
after
extinction
training.
(B)
Prenatal
CORT
significantly
increased
freezing
behavior
in
response
to
the
first
retention
tone
presentation,
indicating
impaired
recall
of
previous
extinction
training.
(C)
Only
prenatal
CORT
increased
total
freezing
behavior
during
extinction
retention,
as
measured
by
area
under
the
curve.
(D)
Extinction
retention
data
fitted
to
a
single-exponential
decay
curve
for
each
group.
(E)
Analysis
of
the
decay
constants
(k)
derived
from
the
regression
lines
in
(D)
indicates
that
all
groups
showed
equivalent
rates
of
extinction
re-learning.
(F)
By
contrast,
analysis
of
the
plateau
term
indicated
that
rats
exposed
to
prenatal
CORT
treatment
were
unable
to
re-extinguish
to
the
same
final
level
of
freezing
behavior
as
controls.
*p
<
0.05
main
effect
of
CORT
vs.
oil-treated
group,
(n)
=
number
of
subjects.
Fetal
corticosterone
and
adult
stress
neurobiology
2753
the
primary
mechanism.
Further,
the
differential
effects
on
BDNF
expression
highlight
the
fact
that
prenatal
CORT
expo-
sure
is
only
one
component
of
prenatal
stress.
It
is
unlikely
that
prenatal
CORT
can
account
for
all
of
the
long-term
consequences
of
PNS.
Consistent
with
our
previous
study
(Green
et
al.,
2011),
both
PNS
and
prenatal
CORT
induced
a
long-term
decrease
in
pontine
TH
mRNA
and
protein
levels.
This
brain
region
con-
tains
the
LC,
the
primary
source
of
norepinephrine
(NE)
innervation
of
the
forebrain,
and
the
sole
source
of
NE
in
the
prefrontal
cortex.
Others
have
demonstrated
a
decrease
in
NE
content,
basal
release,
and
nicotine-evoked
release
in
the
mPFC
as
a
consequence
of
PNS
(Carboni
et
al.,
2010;
Takahashi
et
al.,
1992).
However,
the
mechanisms
underlying
the
long-term
regulation
of
TH
by
prenatal
CORT
remain
unclear.
The
TH
gene
contains
a
composite
GRE/AP-1
site
(Rani
et
al.,
2009)
and
is
responsive
to
regulation
by
gluco-
corticoids;
however,
this
regulation
is
age-
and
brain
region-
specific.
While
direct
comparison
of
fetal
brain
development
in
rats
and
humans
varies
by
brain
region,
the
period
of
maternal
stress
in
this
study
(E14—E21)
corresponds
roughly
to
weeks
6—16
of
human
fetal
brain
development
(Wein-
stock,
2001).
A
recent
study
demonstrated
that
glucocorti-
coids
given
to
rats
at
days
E18—E21
induced
a
marked
increase
in
brainstem
TH
expression,
TH
activity,
and
brain
NE
content;
however,
when
given
postnatally,
glucocorticoids
had
no
effect
(Kalinina
et
al.,
2012).
In
adult
animals,
adrenalectomy
blocked
the
stress-induced
increase
in
TH
expression
in
the
nucleus
of
the
solitary
tract,
whereas
in
the
LC,
exposure
to
CORTeither
had
no
effect
or
inhibited
the
stress-induced
increase
in
TH
expression
(Makino
et
al.,
2002;
Nu´n
˜
ez
et
al.,
2009;
Smith
et
al.,
1991).
Thus,
the
long-lasting
reduction
of
TH
expression
may
be
due
to
epigenetic
mod-
ification
of
transcriptional
regulatory
elements
by
prenatal
CORT
exposure,
rather
than
direct
transcriptional
effects
of
glucocorticoids.
Functionally,
the
long-term
decrease
in
TH
expression
in
the
LC
may
reduce
the
capacity
for
NE
release
in
the
mPFC
during
extinction
training.
4.2.
Fear
conditioning
and
extinction
Prenatal
CORT
and
adult
CAPS
stress
both
altered
fear
con-
ditioning
and
extinction
in
distinct
patterns.
CAPS
stress
increased
freezing
during
fear
conditioning,
and
delayed
the
acquisition
of
fear
extinction,
but
had
no
further
effect
on
the
retention
of
extinction
learning.
By
contrast,
prenatal
CORT
exposure
had
no
significant
effect
on
freezing
behavior
during
fear
conditioning,
but
it
also
impaired
extinction
learning.
Unlike
CAPS
stress,
prenatal
CORT
also
significantly
impaired
the
recall
of
extinction
learning
and
the
final
asymptotic
level
of
freezing
behavior
that
the
animals
achieved.
The
impairment
of
extinction
by
both
prenatal
CORTand
adult
CAPS
stress,
the
lack
of
significant
interaction
between
these
factors,
and
differential
effects
on
condition-
ing
and
extinction
retention
all
suggest
that
the
effects
of
prenatal
CORTand
adult
stress
are
additive
and
independent.
This
is
similar
to
our
previous
findings
with
PNS,
in
which
CAPS
and
PNS
both
impaired
extinction
independently
and
addi-
tively
with
no
interaction
(Green
et
al.,
2011).
Hence,
the
combination
of
prenatal
CORT
exposure
and
adult
stress
creates
an
impaired
phenotype
that
is
greater
than
that
created
by
either
factor
alone.
These
findings
suggest
that
chronic
or
traumatic
stress
in
humans,
when
superimposed
on
a
history
of
high
prenatal
glucocorticoid
exposure,
can
create
an
additive
impairment
on
fear
extinction
even
in
contexts
that
are
distinct
from
the
initial
trauma.
This
type
of
extinc-
tion
deficit
is
a
prominent
and
defining
feature
of
PTSD.
Effective
extinction
learning
occurs,
in
part,
through
activity
in
both
the
IL
subregion
of
the
mPFC
and
the
lateral
amygdala.
Within
the
lateral
amygdala,
extinction
learning
depends
on
NMDA
receptor-mediated
synaptic
plasticity
involving
the
glutamate
NR2B
receptor
(Sotres-Bayon
et
al.,
2007).
In
contrast,
IL
facilitation
of
extinction
learning
does
not
appear
to
be
dependent
on
NMDA
receptor
activity
(Santini
et
al.,
2001).
However,
inactivation
of
the
IL
imme-
diately
prior
to
extinction
training
also
results
in
impaired
extinction
learning
and
retention,
indicating
that
it
has
an
activity-dependent
role
in
extinction
learning
(Sierra-Mer-
cado
et
al.,
2011).
Therefore,
it
is
possible
that
CAPS
stress
and/or
prenatal
CORT
exposure
impairs
extinction
learning
either
through
a
decrease
in
extinction-evoked
IL
activity
or
a
decrease
in
NR2B-mediated
plasticity
in
the
lateral
amyg-
dala.
In
support
of
this
hypothesis,
a
history
of
elevated
CORT
exposure
has
been
shown
to
decrease
both
NMDA
and
AMPA
receptors
in
the
vmPFC
and
impair
contextual
extinction
learning
(Gourley
et
al.,
2009).
In
contrast
to
our
model
of
chronic
stress-induced
impair-
ments
in
extinction
learning,
others
have
found
that
acute
single
prolonged
stress
(SPS)
impairs
extinction
learning
and
retention
and
is
associated
with
an
increase
in
GR
expression
within
the
prefrontal
cortex
and
hippocampus
(Knox
et
al.,
2012).
Although
the
authors
did
not
distinguish
between
subregions
of
the
PFC,
similar
behavioral
outcomes
might
be
achieved
through
differential
modulation
of
GR
expression
patterns
across
PFC
subregions
that
mediate
the
adaptive
responses
to
specific
types
or
durations
of
stress
(Sotres-
Bayon
and
Quirk,
2010).
The
deficit
in
extinction
retention
that
we
previously
described
after
PNS
and
currently
describe
in
rats
exposed
to
prenatal
CORT
may
also
be
related
to
the
reduced
expres-
sion
of
TH
in
the
LC