Fear no more: Beta-Adrenergic blockade affects the neural network of extinction learning

madbrainedmudlickΤεχνίτη Νοημοσύνη και Ρομποτική

20 Οκτ 2013 (πριν από 3 χρόνια και 8 μήνες)

130 εμφανίσεις

1



Fear no more: Beta
-
Adrenergic blockade affects the neural network of extinction learning
and prevents the return of fear in humans




Klodiana Daphne Tona
1

Supervisors:

M.C.W. Kroes
1
and
G. Fernández
1,2

1
Donders Institute for Brain, Cognition and Behaviour, Radboud University Nijmegen,
Nijmegen, The Netherlands;
2
Department for Cognitive Neuroscience, Radboud University
Nijmegen

Medical Centre, Nijmegen, The Netherlands




Correspondence to
:
k.tona@donders.ru.nl,

klodato@gmail.com

Master Thesis Cognitive Neuroscience

July 2012


2


ABSTRACT


Fear and anxiety
-
related

disorders are characterised by malfunctioning extinction learning and
extensive strengthening of emotional memories facilitated by noradrenaline (N
A
)
. In an attempt
to

treat

patients with
t
raumatic memories

interest has switched to mod
ulating emotional
mem
ories by NA

blockade following memory (re)consolidation and
beta blockers have been
proposed for administration in combination with psychotherapy. However, the effect of beta
blockers on extinction learning, and therefore

their
potential
r
ole in psychother
apeutic
interventions and the ultimate fate of fear expression remain unclear.
Therefore
, we conducted an
fMRI study examining
the effects of
beta
-
adrenergic blockade
on extinction learning, extinction
recall and reinstatement of fear, as well as

the
neural networks involved in the regulation of

consolidated fear memories.
Fifty
-
four subjects participated in
a double
-
blind, placebo
-
controlled, between
-
subject study, which took place over three consecutive days.
On
Day 1

participants were conditioned to

two stimuli one of which was
electric shock
reinforced
.


On Day
2
,

after
a single dose administration of β
1
β
2
-
adrenergic receptor antagonist propranolol

or
placebo, participants

were exposed to a context
-
dependent extinction learning paradigm.
On
Day
3
,
extinction recall and reinstatement
of fear

was tested
.
Skin conductance response w
as
measured as an index of conditioned responses

throughout the experiment
and fMRI data was
collected on Day 2 and Day 3
.We report that
beta blockade abolishes differentially conditioned
responses during extinction learning and subsequently prevents the return of fear 24 hours later.
These effects are attributable to changes in the neural network of extinction, where
propranolol

affect
ed

a
ctivity in the
vmPFC
, the
dACC,

and the midbrain
. These findings indicat
e that beta
blockade eliminates
and prevents the return of fear via the impedance of differential responses to
CS+ and CS
-

during extinction learning and provide face validity
to clini
cal interventions
employing beta
-
adrenergic antagonists in conjunction with extinction learning during
psychotherapy

to reduce anxiety disorders

symptomatology
.


Key words
: fear conditioning, extinction learning, extinction recall,

reconsolidation,

vm
PFC,

ACC,

b
eta
-
adrenergic, anxiety disorders,
psychopharmaco
logy



3


INTRODUCTION


Anxiety and fear related disorders are
known as

“the epidemic of the 21
st
century

and are
affecting a high number of people worldwide.

Epidemiological surveys estimate that 18% of
Americans are affected
(
Kessler Rc, 2005
)

and similar numbers are reported for other continents
too. Such a high spread has consequences for patient
s
,
their

family, work and social
environment. It still remains a significant challenge to find an efficient treatment of fear and
anxiety disorders, possibly by optimising the combination of psycho
-

and pharmaco
-
therapy.

Nowadays, a possible
treatment

is p
sychotherapy
, such as cognitive
behavioural

therapy (e.g.
exposure therapy), which incorporates methods of extinction learning
.
Fear and anxiety
-
related

disorders (like PTSD) are thought to be characterised by malfunctioning extinction learning and
extensive strengthening of emotional memories facilitated by

noradrenaline (N
A
) secretion. In an
attempt to prevent traumatic memories

interest has switched to modulating emotional memories
by NA blockade following memory (re)consolidation and
beta blocker
s have been proposed for
administration
to patients with traumatic memories (e.g
.

PTSD patients)

(
Pitman et al., 2002
)

(
Brunet et al., 2008
)
.
Since an effect
ive
treatment

for emotional disorders is psychotherapy, this
would mean that
in most cases

beta blockers
would be administered
in combination with
psychotherapy.

However, the effect of beta blockers on extinction learning, and therefore th
eir

potential
role in psychotherapeutic interventions and the ultimate fate of fear expression remain
unclear.
This study intends to fill this gap.


Investigation of this topic is
timely. In an attempt to modulate

fear memories recent
studies have
shown erasure of fear

memory by combining
noradrenergic blocker

propranolol

and
reconsolidation
processes that are based on a single
, time limited

retrieval of
the fear

memory

(
Kindt, Soeter, & Vervliet, 2009
;
Nader & Hardt, 2009
;
Schiller et al., 2010
)
.
This practice is of
reduced practicality exactly due to this
single

reactivation. Traumatic memories are intrusive
thus
repetitively

retrieved. Therefore, a single reactivation of fear memory, if not impossible,
rarely occurs. Additionally, when it comes to therapy, even if a
psycho
therapist
would
target the
so called reconsolidation window
and only reactivate

a single traumatic memory
, patients wo
uld
be repeatedly retrieving this

experience

when they would go
home:

t
hey would expose
themselves repeatedly to the memory

(
Bos, Beckers, & Kindt, 2012
)
.

This repeated exposure
to
the traumatic memory
without the actual presence of the aversive stimulus thought, opposes

to
the nature of a
single reactivation, and is
related
more
to
the
extinction

process
. Indeed, most
therapeutic practices have their bases on extinction learning (
i.e

exposure based cogn
itive
behavioural therapy) and repeated retrieval is in line with the very nature of emotional disorders
where emotional memories are intrusive.

Despite this controversy, thought, the effects of
noradrenergic blockage on extinction learning remains highly
unknown.





4


Terminology

Extinction has arisen from Pavolvian conditioning research and both conditioning an
d extinction
learning are used as

experimental models for studying fear, anxiety, and safety learning.
In
Pavlovian conditioning

a neutral event
such as an image (conditioned stimulus or CS) is paired
with an aversive event such as a shock (unconditioned stimulus or US). After several pairing
s

of
the image and the
shock
, the presentation of the image itself leads to a fear response
(conditioning re
sponse or CR).

Extinction

is the repeated presentation of the conditioned stimulus (CS) without the
unconditioned stimulus (US), which can lead to reduced conditioned responses (CRs). E
xtinction
is defined as

new learning

of an inhibitory association that

masks the original excitatory
association and therefore
decreases fear responses
. This means

that extinction does not erase a
CS
-
US association
(fear memory)

but creates a competing CS
-
no US association (
extinction
memory
)

(
Milad, Wright, et al., 2007
;
Myers & Davis, 2006
)
.
Thus,
a
fter extinction training, a
new memory trace is created which does not erase the CR memory but inhibits it
(
Milad, Rauch,

Pitman, & Quirk, 2006
;
Myers & Davis, 2006
;
Phelps, Delgado, Nearing, & LeDoux, 2004
)
.

The conditioned memory trace (which expresses fear) and the extinction memory trace (which
inhibits fear) coexist
and compete with each other. Due to the mechanism of extinction learning,
safety

learning occurs. This is the reason why extinction learning is used as a treatment for
anxiety disorders in Cognitive Behavioural Therapy.

After fear extinction, the fear can be
reinstated
(
Rescorla

& Heth, 1975
)
.

Reinstatement

is

the reappearance of extinguished fear
responses following
unsignaled
exposure to

US (shock)
after extinction training

(
Rescorla &
Heth, 1975
)
.



Neuronal circuits of extinction learning

The neuronal circuits involved in extinction learning
include

the amygdala
(
LaBar, Gatenby,
Gore, LeDoux, & Phelps, 1998
;
Milad, Wright, et al., 2007
;
Phelps et al., 2004
)
, the
ventromedial prefrontal cortex (vmPFC),

the

medial prefrontal cor
tex (mPFC)

including
the
dorsal anterior cingualte (
dACC
)

(
(
Lang et al., 2009
)

(
Milad, Quirk, et al., 2007
;
Phelps et al.,
2004
)

the hippocampus

(
Myers & Davis, 2006
)

and the midbrain/ periaquadactal gray (P
A
G)
(
G.
P. McNally, Lee, Chiem, & Choi, 2005
;
Gavan P. McNally, Pigg, & Weidemann, 2004
)
.

In humans, functional connectivity analysis revealed a correlation between the vmPFC
and the hippocampus, as well as between the vmPFC and the amygdala
(
Milad, Wright, et al.,
2007
)
. It has been suggested that during the acquisition of extinction a novel memory is created
through amyg
dala
-
vmPFC interactions, which, following consolidation, lead
s

to vmPFC
inhibition of the central nucleus of
the
amygdala (CEA) and the
PAG
thus preventing the
conditioned response
(
Milad et al., 2006
)
.

Another medial
p
refrontal
region, namely the dACC, is an important hub in the
neural
circuit
underlying
conditioning and extinction.
Dorsal
ACC has been activated in human
5


neuroimaging studies in conditioning and extinction tasks
(
Lang et al., 2009
;
Milad, Quirk, et al.,
2007
;
Phelps et al., 2004
)
. Although dACC consists part of the standard extinction circuit, its
role on emotional memories
and extinction is not clear.
Dorsal
ACC has been proposed to
mediate/ modulate fear responses
(
Milad & Quirk, 2012
;
Milad, Quirk, et al., 2007
)

or exert
control over the amygdala
(
Lang et al., 2009
)
. T
he dACC in humans is homologue to the rodent

prelimbic cortex (PL)

(
Chiba, Kayahara, & Nakano, 2001
;
Stefanacci et al
., 1992
)
. Rodent
studies
suggest

that the PL increases fear responses and impairs extinction
(
Vidal
-
Gonzalez,
Vidal
-
Gonzalez, Rauch, & Quirk, 2006
)
. In line with this, human neuroimaging

studies have
shown that dACC thickness
across individuals
was positively correlated with conditioned fear
responses to the CS+, dACC functional activation increased to the CS+ relative to the CS
-

during
fear conditioning
(
Milad, Quirk, et al., 2007
)

and during extinction
(
Lang et al., 2009
;
Phelps et
al., 2004
)
, and dACC activity was positively correlated with differential SCR
(
Milad, Quirk,

et
al., 2007
)
. The dACC projects to the BLA via excitatory projections
(
Brinle
y
-
Reed, Mascagni, &
McDonald, 1995
)
, therefore it is plausible that it mediates/ modulates fear expression through
excitation of the amygdala.

So, two mPFC regions, the vmPFC and the dACC

(the rodent IL and PL cortices
respectively), can modulate fear expression
through

descending projections to the amygdala.
Whereas PL targets the basal nucleus of the amygdal
a
, IL targets inhibitory areas such as the
lateral
division

of the CEA and ITC. T
herefore
, the
prefrontal cortex is not only

an inhibitor but
exert
s

dual control over fear expression via separate modules, each

with access to separate

inputs
and outputs
(
Milad & Quirk, 2012
;
Sotres
-
Bayon, Bush, & LeDoux, 2004
)


Another key region in emotion processing and extinction is the insular cortex
(
Phelps et
al., 2001
;
Yágüez et al
., 2005
)
. There are numerous reciprocal connections between the insula
and the amygdala
(
Shi & Cassell, 1998
)
. It has been proposed that insula

plays a role in
emotional learning, the acquisition of the cognitive representation of aversive nature of an event
and that conveys a cortical representation of fear to the amygdala
(
Ph
elps et al., 2001
)
.


The role of Noradrenaline on extinction learning and recall

Noradrenaline (NA
) is involved in anxiety and fear related disorders in two ways: first of all
,
noradrenergic

blockers are being prescribed in anxiety disorders because

they can prevent the
physical symptoms that accompany them.
Secondly, N
A

has long been implicated in emotional
expression and the strengthening of memory for emotional and fearful experiences

(
Cahill &
McGaugh, 1998
)
.
Excessive strengthening of emotional memories is a core problem in anxiety
disorders such as
post
traumatic
stress disorder (PTSD), where one of the characteristic features is
a recurrence of intrusive memories for an experienced trauma. The persistence of disturbing
traumatic memories in PTSD is often explained in terms of a trauma
-
induced enhancement of
memory

encoding

(
Pitman, 1989
)
.
An increa
sed noradrenergic activity during trauma enhances
the encoding of emotional/fearful memory

(
O’Donnell, Hegadoren, & Coupland, 2004
)

an
d
6


increased noradrenergic activity is also implied in the maintenance of PTSD symptoms

(
Geracioti et al., 2001
)
.
In light of the findings that involve

NA

on
enhancement

of emotional
and fearful memories, recently, interest has switched to modulating memory for emotional
experiences by NE blockade following consolidation and reconsolidation of fear memory
(
Cahill
& McGaugh, 1998
;
Dębiec & Ledoux, 2004
;
Kroes, Strange, & Dolan, 2010
;
Nader, Schafe, &
Le Doux, 2000
)
.

Based on these findings, NE blockers such as b
eta
-
adrenoreceptor antagonist
propranolol have been proposed for administration in combination with psychotherapy in order
to achieve a more efficient treatment of fear and anxiety disorders

(
Brunet et al., 2008
)
.
However, the ef
fect of propranolol on extinction learning

and
the initial fear
memory trace
, and
therefore its role for psychotherapeutic interventions, remains unclear. Research towards the
effect of
NA
blockade on extinction learning

and extinction memory trace

has lea
d to mixed
results
(
Cain, Blouin, & Barad, 2004
;
Mueller, Porter, & Quirk, 2008
;
Ouyang & Thomas, 2005
)
.
One possibly problema
tic consequence of NE blockers is that they may impair the retention of
extinction memory and as such result in increase of fear responses. In line with this statement,
Muller and colleagues (2008) showed that blockade of

Infralimbic cortex (IL)

beta
-
adrenergic
receptors during extinction learnin
g impairs extinction retention and a recent human study
implicates that propranolol impaired extinction at a cognitive level

(
Bos et al., 2012
)
.
These
findings question the effectiveness

of combining NE blockers with psychotherapy

and call for
further investigation of this topic, as well as of the
underlying
neuronal circuits
.


Therefore, w
e
conducted a study investigating the neuronal mechanisms involved in extinction learning and
examined
the effects of
NE

blockade

on extinction learning, extinction recall and reinstatement
of fear
.

We conducted a double
-
blind, placebo
-
controlled and between
-
subjects experiment,
which took place over three consecutive days. On Day 1 participants w
ere conditioned to a
stimulus via mild shock administration
to

the right index and middle finger, on Day 2
conditioned
fear was extinguished, and on Day 3 extinction recall and fear reinstatement took
place. Skin conductance measures were used to asses con
ditioned fear responses and BOLD
fMRI data was collected during extinction learning, recall and reinstatement to assess neural
activity and modulation by NE. To manipulate
context we displayed visual CSs
within
photographs of two distinct rooms such that conditioning was performed in context A and
extinction recall and reinstatement of fear in context B. An additional manipulation of the
context is related to the fact that context A was presented in the du
mmy scanner while context B
in the fMRI scanning lab. Finally, a partial r
einforcement paradigm was used to slow extinction
learning, which occurs rapidly in humans with 100% reinforcement

(
LaBar et al., 1998
)
.

We hypothesise that extinction learning relies on the “extinction network” involving the
vmPFC, the
d
ACC, the amygdala, the insula and the midbrain and that
raised NA
levels are
critical for retention of safety associations. Therefore,

NA

blockade prior to extinction learning
should impair extinction recall

and lead to increased fear
.





7


MATERIALS AND METHODS


1.

Participants

Fifty four participants (21 males and 33
females) with normal or corrected to normal vision,
without neurological, psychiatric or cardio respiratory history participated in the study. Data of
seven subjects were excluded due to technical probl
ems (MRI data of 6 participants were
excluded due to p
roblems with Matlab server licence or MRI scanner failure and
MRI data of
one participant was

excluded because the participant fell asleep during the task)
; therefore results
from forty seven subjects are reported.

The placebo group comprised 24

subjects
,

11

males and
13 females; the propranolol group comprised 23 subjects, 8 males and 15 females. Participants
age ranged from 19 to 26 years (M= 21.65, S.D= 2.15).



All participants were assessed to be free from any current or previous medical or
psychiatri
c condition that would contraindicate taking a single dose of 40 mg oral dose of
propranolol (i.e., seizure disorder, respiratory disorder, cardiovascular disease

etc.).
Finally,
additional exclusion criteria were applied to guarantee safety in the MRI sca
nner (i.e.: no
claustrophobia, metal objects around the body

etc).
To avoid pregnant women and gonadal
hormone fluctuations, we included women

only if they were r
eceiving hormonal contraceptives.

Menstrual cycle

position is also thought to influence the stress hormone response, with higher
levels of salivary cortisol in response to psychosocial stress during the luteal phase compared to
the follicular phase

(
Kirschbaum, Kudielka, G
aab, Schommer, & Hellhammer, 1999
)
.

Participants were ra
ndomly divided into two groups with different pharmacological
treatment (non
-
selective b
-
adrenergic antagonist propranolol or placebo). Drug administration
was double blind. Volunteers

received either partial course credits or were financially
compensated
for their participation in the study.
Written informed consent was obtained before
the experiment and the experiment was carried in accordance with the local ethical review board
(CMO Region Arnhem
-
Nijmegen, the Netherlands) and in accordance with the Hels
inki
declaration.


2.

Independent variables

a.

Drug administration

Participants were given placebo or propranolol
.
Propranolol is a non
-
selective beta blocker,
which
block
s noradrenergic

action on both
β
1
-

and β
2
-

adrenergic receptors
.

b.

Tasks

The experiment was conducted on three consecutive days, with each session at the same time of
the day.



8


Day 1: Conditioning Task

During the conditioning task a
photographic picture

depicting a room was presented. In this
room a yellow or a blue light turned on constituting the conditioned stimuli: one light colour was
electric shock
-

reinforced (CS+) and the other was not (CS
-
). The stimuli were pseudo
-
randomly
presented with the re
striction that the same number of shocks and same number of CS+ and CS
-

were presented during the first and the second half of the task. The colour light that would be
shock reinforced was randomly selected for each participant and remained constant during

the
task.

Presentation of stimuli followed after

a habituation phase; CS
-

was presented 12 times
and CS+ was presented 18 times (12 times not followed by a shock and 6 shock reinforced).

Habituation phase consisted of six trials where the context pictu
re was presented for 3 to 5
seconds (mean: 4 sec), followed by 4 sec in combination with the CS+ or CS
-
. During the actual
task, the context picture was presented for 11
-
13 sec, (mean: 12 sec), followed by 4 sec in
combination with the CS+ or CS
-
. During t
he task skin conductance and heart rate measures
were obtained. In total, the conditioning task lasted 15 minutes (see figure 1).

Day 2: Extinction Task

Following a habituation phase, participants were presented with the same two visual stimuli as
on Day
1 but within a different contextual background. Although electrodes were attached, no
shocks were administered during the extinction task.
The habituation phase consisted of six trials
where the context picture was presented for 3 to 5 seconds (mean: 4 sec
), followed by 4 sec in
combination with the CS+ or CS
-
. During the actual task, the context picture was presented for
11
-
13 sec, (mean: 12), followed by 4 sec in combination with the CS+ or CS
-
. The mean
duration of fixation cross presentation was 12 sec.

During the task skin conductance and fMRI
data was acquired. In total, the task lasted 15 minutes (see figure 1).

Day 3: Extinction recall and reinstatement Tasks

T
he Extinction recall task

was the same as the extinction learning task (Day 2)
.

During the

reinstatement of fear task,

during presentation of
the contextual background

alone

(without the
presentation of the cued stimuli)
, four

unsignalled shocks were administered
after 30, 100, 115,
and 200 seconds
.
Consecutively,

each stimulus was presented again without any shock
administration. Task duration was 19 minutes.
During all tasks skin conductance and heart rate
was measured and fMRI data was acquired (see figure 1).

c.

Electrical Shock application

Electrical stimulation
was applied using a MAXTENSE 2000 device (ProtechInc, Wonjusi,
Gangwon
-
do, Korea). This device is MRI compatible and provides constant current, high voltage
pulses of brief duration. An electrical shock consisted of one 200 msec

asymmetrical biphasic
pulse

with a pulsewidth of 250 μs at 150 Hz. The electrical shock was delivered transcutaneously
over the volunteers' right index and middle finger.



9


d.

Stimuli

Stimuli were two background pictures of similar and easily distinguishable rooms (Figure 1).
These s
erved as the experimental contexts: one context functioned as a background for
conditioning and the other one functioned as the context for the extinction, and reinstatement
tasks. The order of the context was counterbalanced across participants. In these
contexts a lamp
was present that would turn on to have a blue or a yellow colour. The coloured light served as the
conditioning stimuli. One of the coloured lights (e.g
.,

the yellow) was followed by a shock while
the second light (in this case blue) was n
ever followed by a shock (SC
-
).
The CS type (yellow or
blue) was counterbalanced across participants and the order of CS type (CS+ or CS
-
)
presentation was counterbalanced across all phases of the experiment.


3.

Dependent variables

a.

Vital signs

Blood pressure and heart rate was measured with a sphygmomanometer (OMRON M6,
Intellisense).

b.

Personality questionnaire

State
-
Trait Anxiety Inventory
-

TRAIT (STAI
-
T)

The State
-
Trait Anxiety Inventory (STAI) measures anxiety in a specific situation (state)

or as a
more general long lasting quality (trait). The subscale “Trait Anxiety Inventory” (STAI
-
T)
measures anxiety as a more general long lasting quality (trait). Subjects indicate on a 4
-
point
L
ikert
scale for 20 statements to which degree these applied

to them. Some of the questions
related to absence of anxiety and are reversed scored. The range of scores is 20
-
80; higher scores
indicate greater anxiety (Spielberger et al., 1970).

c.

Mood questionnaires

Positive and negative affect scale (PANAS)

Subjectiv
e mood was assessed by obtaining scores on the positive and negative affect scale
(PANAS). Ten items for positive and ten for negative affect had to be rated on a five
-
point scale
ranging from 1
-
not at all to 5
-
extremely. A mean score was calculated for su
bjective positive and
negative affect (Watson et al., 1988).

State
-
Trait Anxiety Inventory
-

STAIT (STAI
-
S)

The State
-
Trait Anxiety Inventory (STAI) measures anxiety in a specific situation (state) or as a
more general long lasting quality (trait). The subscale “State Anxiety Inventory” (STAI
-
S)
measures anxiety in a specific situation (state). STAI
-
S contains f
our
-
point Likert

items; it serves
as an indicator of the state anxiety and measures the severity of the overall anxiety levels. Some
of the questions are related to the absence of anxiety and are reversed scored. The range of scores
is 20
-
80 and the higher

score indicates greater anxiety (Spielberger et al., 1970).

10


d.

Skin Conductance Response

Galvanic skin conductance was recorded with a Brain amplifier recording device and Brain
vision recorder software (Brain Products GmbH, Gilching, Germany. Skin conductan
ce
responses was measured via two silver
-
nitrate Ag/AgCl sensors on the palmar side of distal
phalanges of the index and middle finger of the non
-
dominant hand with standard NaCl
electrolyte gel. The skin conductance signal was amplified using MR compatibl
e BrainAmp MR
and BrainAmpExG MR (BrainProducts GmbH) within the MR environment, transmitted through
an optical cable, and recorded outside the MR environment using BrainVision Recorder
software. Data was continuously recorded at 5000 samples per second.

e.

fMRI

MRI data acquisition was performed on a 3 Tesla Siemens Magnetom Trio scanner equipped
with 32
-
channel transmit
-
receiver head coil (Siemens Medical System, Erlangen, Germany). The
manufacturer’s automatic 3D
-
shimming procedure was performed at the be
ginning of each
experiment. Subjects were placed in a light head restraint within the scanner to limit head
movements during acquisition. The functional scans for the tasks and the localizer were acquired
using a multi
-
echo gradient pulse sequence.

Field
Maps

A B
o
Field Map was acquired using a gradient echo field map sequence using the following
parameters: 64 axial slices aligned with AC
-
PC plane, slice thickness = 2.0 mm, interslice gap =
50%, repetition time (TR) = 1020 ms, echo time (TE1) = 10 ms, (TE2
)= 12,45, flip angle α =
90°, Voxel size= 3.5 x 3.5 x 2.0, FOV = 224.

Functional Image acquisition

Functional images were acquired with single
-
shot gradient echo
-
planar imaging (EPI) sensitive
to the blood
-
oxygenation level dependent (BOLD) response. We

used parallel
-

acquiring
inhomogeneity
-
desensitized fMRI (Poser et al, 2006). This is a multiecho planar imaging
sequence, in which images are acquired at multiple time echos (TEs) using the following
parameters: 37 axial slices aligned with AC
-
PC plane,
slice thickness = 2.5 mm, interslice gap =
17%, repetition time (TR) = 2320 ms, echos time: TE1 = 9ms, TE2= 19.3 ms , TE3=30 ms,
TE4= 40ms, flip angle α = 90°, Voxel size= 3.3 x 3.3 x 2.5, FOV = 211, fat suppression. The
benefit of multi echo imaging is th
at it reduces image artifacts and obtains signal from areas
sensitive to distortion (e.gvMPFC). The number of slices did not allow acquisition of a full brain
volume in most participants. We made sure that the entire frontal and temporal lobes fitted
withi
n the field of view because these were the regions where the fMRI effects of interest were
expected. This meant that data from the superior parietal lobe (data from the top of the head) was
not acquired in several participants.

Structural scan

A structura
l scan with a 3D magnetisation prepared rapid gradient echo (MPRAGE) sequence
was acquired using the following parameters: 192 sagittal slices slice thickness= 1 mm, no slice
gap, TR = 2300 ms, TE = 3.03 ms, voxel size= 1 x 1 x1 mm, FOV = 256
.


11


4.

Procedure

a.

Visit 1: Safety Screening

Subjects signed an informed consent form upon which they filled out a demographics and
medical history questionnaire (Figure 1). Vital signs were measured: heart
-
rate, blood pressure
and base heart rate (
10 sec and
60 sec).
Bloo
d pressure and heart rate during screening was
measured as a safety measurement. Consecutively, personality questionnaires (including STAI
-
T) were administered.
When subject met all inclusion and none of the exclusion criteria, they
were randomly assigned
to a placebo or propranolol group in a double blind procedure.

b.

Procedure common to all experimental days

Prior to the task participants’ hands were cleaned with alcohol and consecutively entered the
dummy scanner where skin conductance and shock electrodes were attached to their fingers. The
electrodes for the electrical shock were attached to the volunteers'

dominant hand’s index and
middle finger. The SCR electrodes were attached to the volunteers' non dominant hand’s index
and middle finger. Consecutively, the electric shock device was set: participants underwent a
staircase procedure to adjust the strength

of electrical shocks in order to determine the level of
the shock that would be administered during the experiment. The intensity level was set
individually prior to the task to be maximally uncomfortable without being painful. After the
shock level was
set, the tasks took place. During the tasks subjects lay in the dummy scanner or
the fMRI scanner. All stimuli were visual material presented via a projector outside the scanner.
Subjects viewed the screen via a nonmagnetic mirror.
Stimuli were presented u
sing Matlab
software
2009b
.
Subjects were instructed to pay attention to the computer screen and were told
that there is a relationship between the stimuli appearing on the screen and the shocks that they
would receive. Throughout the sessions, the partici
pants were attached to the SCR and shock
electrodes.

c.

Visit 2: Experimental Day1

On first experimental day, after the standard preparation described above and electric shock
adjustment, the conditioning task took place in the dummy scanner.

d.

Visit 3: Expe
rimental Day2

On the second experimental day, upon arrival (time point T1) participants’

vital signs were
measured (blood pressure and heart rate) and their mood state was assessed using STAI
-
S,

and
PANAS. Consecutively, propranolol or placebo pill was ad
ministered. The influence of
propranolol on blood levels, heart rate and mood was assessed at different time points T1 (prior
adminstartion), T2 (30 min after administration), T3 (60 min after administration) (Figure 1).
Consecutively, participants underwe
nt the standard procedure for finger cleaning, electrode
attachment and electrical shock adjustment
before the extinction learning task started in the
fMRI scanner.
Although electrodes were attached, no shocks were administered. During the task,
skin condu
ctance and heart rate were measured and fMRI data was acquired. Following the task
12


completion, a structural scan (T1) was obtained. Upon scanner exit, mood and vital signs were
measured once more (time point T4), (Figure 1).Participants departed from the
research centre
only when vital signs were within safety limits and participants were encouraged to contact the
researcher in case of queries or possible health related issues occurring on that day.

e.

Visit 4: Experimental Day3

At the beginning of the thir
d experimental day, vital signs and mood state were assessed at time
point 1 (T1). Prior to the task participants underwent the standard procedure for finger cleaning,
electrode attachment and electrical shock adjustment. Consecutively, participants enter
ed the
MRI scanner and tasks were conducted in the following order: recall of extinction task, resting
state scan, reinstatement and reextinction task (Figure 1). Upon exiting the scanner, at time point
2 (T2) vital signs and mood were measured and persona
lity questionnaires were filled in.
Following completion of the study subjects were debriefed on the aims and details of the study.



Figure
1
.

Experimental design
.
Safety screening,
Conditioning on
Experimental
Day 1, drug supplementation and
Extinction learning on
Experimental
Day 2, Extinction recall and Reinstatement of fear on

Experimental

Day 3.
Time
points (
T1, T2, T3,
and T4
) present the time points at which
vital signs were

assessed and mood/personality
q
uestionnaires were administered


13


DATA ANALYSIS

1.

Analysis of
Questionnaires

All data
analyses
from question
naires were performed with SPSS
for Windows software package
PASW Statistics 18
. Significance criterion of α= 0.05 was used.


2.

Skin Conductance
Analysis

Skin conductance data was assessed using an in
-
house analysis programme written in Matlab

2009band using FieldTrip.
Data was low
-
pass filtered at 5Hz and resampled to 100Hz. The level
of skin conductance responses was determined for each trial as
the peak
-
to
-
peak amplitude
difference in skin conductance of the largest deflection in the latency window from 0
-
8 s after
stimulus onset. The raw skin conductance responses were square root transformed to normalize
the distributions.


3.

fMRI

Analysis

Pre
-
processing

fMRI data were processed and
analyzed

using the statistical software package SPM8 (Wellcome
Trust Centre for Neuroimaging, London, UK;
http://www.fil.ion.ucl.ac.uk/spm
)
. The first six
volumes were excluded to control for T1 equilibration eff
ects.
First
field

Maps were created

to
correct for
effects of
field
(Bo) inhomogeneity and reduce image distortion and blurring (for
toolbox and methods see

(
Cusack, Brett, & Osswald, 2003
;
Chloe Hutton et al., 2002
;
C. Hutton,
Deichmann, Turner, & Andersson, 2004
)
. For the fieldmaps, six movement parameters (three
translations and three rotations) were extracted from the first echo of each volume and
subsequently used to

correct for rigid head movements in all four echoes of each volume.
Subsequently, all four echoes were combined into a single volume using a weighted average
.
This was done in the following way: 26 brain volumes were used to calculate mean activation
and
standard deviation in each voxel for the four echoes. Consequently all four echoes were
combined into a single volume using a weighted average. In the end these parameters were
applied to the volumes acquired during the task. This was done in the following

way: 26 brain
volumes were used to calculate mean activation and standard deviation in each voxel for the four
echoes. Consequently all four echoes were combined into a single volume using a weighted
average. In the end these parameters were applied to th
e volumes acquired during the task.
F
or
each participant,
the structural image was corregistered to template image, then the functional
image to the template image and in the end the structural image to the mean functional image.
Following this process, th
e corregistered images were segmented.
After segmentation into grey
and white matter,
weighted
and
corregistered
functional
images were spatially normalized
-

by

applying the normalization par
ameters from the segmentation procedure

-
into a common
stereotactic space (MNI 152 T1
-
template) and resampled to 2 × 2 × 2
-
mm
3

isotropic voxels using
14


trilinear

interpolation. Finally, spatial smoothing to the functional images was applied with an
isotropic 3D Gaussian kernel of 8 mm full
-
width ha
lf
-
maximum (FWHM). The distortion
-
corrected, realigned, weighted, normalized and smoothed images served as input to further
statistical analysis.

General Statistics

Statistical analysis was performed within the
framework

of the general linear model. The
co
mbined and
preprocessed

time series were
analyzed

as an event related design and
each
condition was convolved with a
hemodynamic

response function (HRF) and used as a regressor.
Additionally, realignment parameters were included to model potential movement

artefacts
. The
data were high
-
pass filtered (cut
-
off 128s) to remove low
-
frequency signal drifts, and a first
-
order autoregressive model was used to model the remaining serial correlations
(
Friston et al.,
2002
)
.
Contrast images of testing parameter estimates encoding condition
-
specific effects were
created for each subject.

The single
-
subject contrast images

(conditioned stimuli vs. fixat
ion)

were entered into
voxel
-
wise one
-
sample t
-
tests to assess main effects of task, a
nd
implemented in a second
-
level
random effects analysis.

This group analysis was a 2 (drug) x 2 (phase) x 2 (conditioned
stimulus) mixed model analysis of variance.
We r
eport regions that survive cluster
-
level
correction for multiple
-
comparisons (family
-
wise error, FWE) across the whole brain at p<0.05
using an initial height threshold of p < 0.001, unless otherwise indicated.
Correction for multiple
comparisons was also

conducted
for the search volume of the vmPFC
and the midbrain
by using
a small volume correction which was functionally defined based on prior
literature

(
Kalisch,
Korenfeld, et al., 2006
)

and
(
Hermans et al., 2011
)

respectively.




RESULTS


Participants that did not show conditioned response to the reinforced stimuli on the first
experimental day were excluded and replaced (eight in total).
Out of

the fifty four participants
that finally participated in the study,
MRI
data of
7

participants were excluded due to technical
failure o
n one of the experimental days

(
for details
see also subjects)
.
The placebo group
comprised 24

subjects, 10 males and 13 females; the propranolol group comprised 23 subjects, 8
males and 15 females. Participants age ranged from 19 to 26 years (M
= 21.65, S.D= 2.
15).
SCR
data from two participants were excluded due to
high noise SCRs
that made the as
sessment of
fear impossible
.


15


Test for group differences in

physiological

and psychiatric background

Separate ANOVAs were performed to verify that there were no inherent differences between
populations in each pharmacological group. No significant
differences were observed
;

age (
F
1
=
0.062,
p=0.805), Trait anxiety as measured with STAI
-
T (
F
1
=
0.531,
p=0.47), baseline heart rate
(Welch test,
F
38.66
=
0.703,
p=0.407), systolic and diastolic Blood Pressure (
F
1
=
0.200,
p= 0.657
,

and

F
1
=
0.055,

p=0.815).

Effects of drug on blood pressure

There was no baseline systolic
or diastolic
BP differences between groups (
t
(46)
=
-
0.
124;
p=0.902,

and t
(46)
= 0.589; p=0.559
two
-
tailed

respectively
).
In line with previous findings
(
Kroes et al., 2010
;
Tollenaar, Elzinga, Spinhoven, & Everaerd, 2009
)

we observe a drop in
systolic BP in the propranolol group (
t
(46)
=
-
1
.
934; p= 0.
0
29
, one tailed
) after 60 min but no drop
in

diastolic BP
(
t
(46)
=
-
0.354; p= 0.362, one tailed
)
. This effect persisted also 120 min after
administration, (
t
(46)
=
-
1.922; p=0.030
; for systolic
)

and
(
t
(46)
= 0.397; p=0.346

for diastolic)

(
Fig
ure 2
).

Figure

2
.

Systolic and diastolic BP (in millimetres mercury) before placebo or propranolol administration (0 min),
60 min and 120 minutes
later. No baseline difference in systolic or diastolic BP is present between groups. The
propranolol

group displays a significant reduction in systolic BP 60 and 120 min after administration.

Error bars
indicate SEM.
*p< 0.05, one
-
tailed.


16


Effects of drug on
Mood Questionnaire
s

Positive and Negative Affect

Repeated measures ANOVA was performed on PANAS scores, by entering the factor time, with
three levels and group as between
-

subjects factor to monitor mood over time and to see if there
was an effect of the drug induction on positive or negative affect. Fi
rst we tested negative affect.
The factor time had a

significant
effect (F
1.60
= 14.05, p=0.000). This was further analysed with
single contrasts. We found that at the first time point of the day, negative affect was significantly
higher than
at
the second
time point
(p= 0.003) and
at
the second
time point
was significantly
higher than
at the
third time point (p= 0
.005). The drug group had no

effect on negative affect
(F
1
= 1.806, p=0.186) and neither had the interaction between drug and time (F
1.60

=0.035,
p
=0.940).These results suggest that there were no effects of drug m
anipulation on negative affect.

We performed the same tests for positive affect. There was an effect of time (F
42
=6.057,
p=0.003). This was further analyzed with single contrasts. We found that at the first time point of
the day, positive affect was significantly higher than the second time point (p= 0.003) but the
second time point was not significantly higher than
third one (p= 0.199).
D
rug had no sig
nificant

effect on positive affect (F
1
=
0,868 p=0.357) and neither had the interaction between drug and
time (F
2
=
1.070, p=0.347).

These results suggest that there were no effects of drug manipulation
on positive affect.

Anxiety (STAI
-
S)

Repeated measures ANOVA was performed on STAI state scores, by entering the factor time,
with three levels and group as between
-

subjects factor to monitor state

anxiety

over time and to
see if there was an effect of drug on state anxiety. The factor time had a
n
effect (F
1.75
=
5.962,
p=0.005). This was further analyzed with single contrasts. We found that at the first time point of
the day, state anxiety was significantly

higher than
at
the third time point (p= 0.
005). The drug
group had no
effect on state anxiety (F
1
=
0.841 p=0.364) and neither had the interaction between
drug and time (F
1.755
=
0.210, p=0.782).

These results suggest that there were no effects of drug
manip
ulation on state anxiety.


Physiological Assessment
:
Skin Conductance Response

Conditioning

For each CS type (CS+, CS
-
) skin conductance responses were
averaged
over the first six trials
(Early phase) or last six trials (Late Phase). A CStype

(CS+, CS
-
) x Phase (Early, Late) with
Group (Propranolol, Placebo) as a between
-
subjects factor 2x2x2 repeated measure ANOVA
revealed

differential

fear conditioning on day 1 (CStype x Phase (F
1, 47

= 5.191,
p

< 0.027) and
no difference in fear learning b
etween the propranolol and the placebo group

(Figure 3)
.


17


Extinction learning

For each CS type (CS+, CS
-
) skin conductance responses were averages over the first six trials
(Early phase) or last six trials (Late Phase). A CStype

(CS+, CS
-
) x Phase (Early, Late) with
Group (Propranolol, Placebo) as a between subjects variable 2x2x2 repeated measure ANOVA
revealed a main effect of CStype (F
1, 47

= 13.531,
p

= 0.001), a main effect of Phase (F
1, 46

=
37.192,
p

< 0.001), an interacti
on effect of CStype x Group at trend (F
1, 46

= 3.609,
p

= 0.064), an
interaction effect of CStype x Phase (F
1, 46

= 4.854,
p

= 0.033), an interaction effect of CStype x
Phase x Group (F
1, 46

= 9.890,
p

= 0.003), with no other main effects or interactions.

Paired samples t
-
tests responses revealed that, in contrast to the placebo group, the
administration of propranolol led to elimination of differentiation between CS+ and CS
-

1,5 h
after administration (
t
(22
)
=
0.914,
p

= 0.371; Early CS+ Mean: 0.598, s.e.
m.: 0.094, Early CS
-

Mean: 0.549, s.e.m.: 0.098) On the contrary, differential SCRs remain stable for the placebo
group where greater responses to the CS+ compared to the CS
-

during the Early phase were
observed (
t

(24
)
=
5.427,
p

= 0.075; Early Phase CS+ M
ean: 0.738, s.e.m.: 0.101, Early Phase CS
-

Mean: 0.455, s.e.m.: 0.068).

Since the two groups differed already at the beginning of the task, they also differed over
the co
urse of extinction learning
.
We calculated the difference in average responses to the CS+
and the CS
-

for
each phase (
Early Phase Diff= meanEarlyCSp
-

meanEarly CSm and Late Phase
diff= meanLateCSp
-

meanLate CSm
the early and late phase.
Paired T
-
tests reveal greater
differential resp
onding in the Early compared to the Late phase in the Placebo Group (t

(24
)
=
3.101, p

= 0.05; Early Phase Diff : 0.283, s.e.m., 0.052, Late Phase Diff : 0.080, s.e.m., 0.063)
but no change was found for the propranolol group (
t

(24
)
=
-
1.053,
p

= 0.312; Ear
ly Phase Diff :
0.040, s.e.m., 0.044, Late Phase Diff : 0.076, s.e.m., 0.050).

Additionally,
there was a significant decrease in fear SCR responses at the second half of
the extinction learning task for both groups. Both groups showed lower responses to
the CS+
during the late phase as compared to the early phase [Placebo Group (
t
(24
)
= 4.718,
p
< 0.001;
Early CS+ Mean: 0.738, s.e.m.: 0.101, Late Phase CS+ Mean: 0.424, s.e.m.: 0.074), and
Propranolol Group (
t
(22
)
= 4.452,
p
< 0.001; Early CS+ Mean: 0.598, s.e
.m.: 0.094, Late Phase
CS+ Mean: 0.396, s.e.m.: 0.083)] and lower responses to CS
-

during the late phase as compared
to the early phase [Placebo Group at trend (
t

(24
)
= 1.862,
p

= 0.075; Early CS
-

Mean: 0.455,
s.e.m.: 0.068, Late Phase CS
-

Mean: 0.343, s.e
.m.: 0.050), and Propranolol Group (
t
(22
)

= 4.569,
p
< 0.001; Early CS
-

Mean: 0.549, s.e.m.: 0.098, Late Phase CS
-

Mean: 0.286, s.e.m.: 0.063)]. At
the end of the task (late phase) both groups showed no response differences to the CS+ and CS
-

[Placebo Group

(
t

(24
)
=
1.266,
P

= 0.218; Late Phase CS+ Mean: 0.424, s.e.m.: 0.074, Late Phase
CS
-

Mean: 0.344, s.e.m.: 0.050), and Propranolol Group (
t

(22
)
=
1.529,
P

= 0.141; Late Phase
CS+ Mean: 0.396, s.e.m., 0.083, Late Phase CS
-

Mean: 0.286, s.e.m.: 0.063)].

Finally, calculating the average skin conductance responses to context presentations over
the Early phase (trial 1
-
12) and Late Phase (trial 13
-
24) and testing a Phase (Early, Late) x Group
(Placebo, Propranolol) 2x2 ANOVA revealed a main effect of Phase (
F
1, 46

= 5.568,
p

= 0.023),
with critically no main effect of Group and no interactions. Paired T
-
test reveal a reduction in
18


responses to context presentations in the Late Phase compared to the Early Phase (
t

(47
)
=
2.403,
p

= 0.020; Early Phase Context Mea
n: 0.266, s.e.m.: 0.036, Late Phase Context Mean: 0.193,
s.e.m.: 0.025).

Thus during the early Phase the Placebo group shows greater responses to the CS+
compared to the CS
-
, whereas this effect is absent in the Propranolol group. Both groups exhibit
no di
fferential responses in the Late Phase. Further, the groups show no difference in responses
to cont
ext presentations

(Figure 4).



Figure 3.

Skin conductance Responses for fear conditioning.

Square root transformed SCRs for the placebo (thick
line) and
propranolol group (dotted line) for CS+ (red) and CS
-

(blue). Both groups were equally conditioned at the
end of the task.


Figure 4
.

Skin conductance Re
sponses for extinction learning. Square root transformed SCRs for the placebo (thick
line) and
propranolol group (dotted line) for CS+ (red) and CS
-

(blue). Contrary to the placebo group, the
propranolol group does not diffferernatiate between CS+ and CS
-
.

0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Plac CSp
Plac CSm
Prop CSp
Prop CSm
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Plac CSp
Plac CSm
Prop CSp
Prop CSm
19


Extinction Recall

For each CS type (CS+, CS
-
) skin conductance responses were averages over
the first six trials
(Early phase) or last six trials (Late Phase). A CStype (CS+, CS
-
) x Phase (Early, Late) with
Group (Propranolol, Placebo) as a between subjects variable 2x2x2 repeated measure ANOVA
revealed a main effect of CStype (F
1, 46

= 14.817,
p

< 0.001), a main effect of Group (F
1, 46

=
4.335,
p

= 0.043), a main effect of Phase (F
1, 46

= 22.638,
p

< 0.001), an interaction effect of
Phase x Group (F
1, 46

= 6.678,
p

= 0.013), an interaction effect of CStype x Phase at trend (F
1, 46

=
3.975,
p

= 0.052), with no other main effects or interactions.

As the recall paradigm is principally a second extinction session and follows the actual
extinction task, extinction learning can be expected to occur rapidly. Therefore we also
recalculated for each CS type (CS+, CS
-
) the Early Phase as the average skin c
onductance
response over the trials 1
-
4, a Middle Phase as the average skin conductance response over the
trials 5
-
8, and the Late Phase as the average response over trial 9
-
12. First, a CStype (CS+, CS
-
)
x Phase (Early, Middle, Late) with Group (Proprano
lol, Placebo) as a between subjects variable
2x3x2 repeated measure ANOVA revealed a main effect of CStype (F
1, 46

= 14.817,
p

< 0.001), a
main effect of Phase (F
2, 92

= 19.372,
p
< 0.001), a main effect of Group (F
1, 46

= 4.335,
p

=
0.043), an interaction

effect of Phase x Group (F
2, 92

= 7.387,
p

= 0.001), an interaction effect of
CStype x Phase (F
2, 92

= 3.379,
p

= 0.038), with no other main effects or interactions
.

Next, we calculated the differences between the average responses to the CS+ and CS
-

per
Phase for each subject. A one
-
way ANOVA on the difference scores reveal a Group
difference in the Early Phase (F
1, 46

= 4.335,
p

= 0.043; Placebo Early Diff Mean: 0.234, s.e.m.:
0.072, Propranolol Early Diff Mean: 0.066, s.e.m.: 0.032) but not the Middle
Phase (F
1, 46

=
0.000,
p

= 0.994; Placebo Middle Diff Mean: 0.074, s.e.m.: 0.060, Propranolol Middle Diff
Mean: 0.074, s.e.m.: 0.036), and not in the Late Phase (F
1, 46

= 0.044,
p

= 0.834; Placebo Late
Diff Mean: 0.033, s.e.m.: 0.047, Propranolol Late Diff

Mean: 0.022, s.e.m.: 0.025).

Independent samples T
-
tests reveal that the Placebo group exhibited greater response
s

to
the CS+ compared to the CS
-

than the Propranolol group during the Early Phase (
t

(33.155
)
=

2.144,
p

= 0.039; Early Phase Diff Placebo Gro
up: 0.234, s.e.m., 0.072, Early Phase Diff Propranolol
Group: 0.066, s.e.m., 0.032), but not during the Middle Phase (
t

(46)
=

-
0.008,
p
= 0.994; Middle
Phase Diff Placebo Group: 0.074, s.e.m., 0.060, Middle Phase Diff Propranolol Group: 0.074,
s.e.m.,
0.036), and not during the Late Phase (
t

(36.297
)
=
0.216,
p

= 0.830; Late Phase Diff Placebo
Group: 0.033, s.e.m., 0.047, Late Phase Diff Propranolol Group: 0.022, s.e.m., 0.025).

F
ollowing up on these results, paired T
-
tests reveal greater responses to th
e CS+
compared to the CS
-

during the Early phase in the Placebo Group (t
(24
)
=
3.266, p

= 0.003; Early
Phase CS+ Mean: 0.631, s.e.m.: 0.094, Early Phase CS
-

Mean: 0.397, s.e.m.: 0.068), and
response differences between the CS+ and CS
-

during the Early phase

in the Propranolol Group
at trend (t

(22
)
=
2.044, P = 0.053; Early CS+ Mean: 0.275, s.e.m.: 0.070, Early CS
-

Mean: 0.210,
s.e.m.: 0.060).
Thus a spontaneous recovery of fear is observed in the placebo group but not for
t
he propranolol

group at the beginning of the task.

20


Finally, calculating the average skin conductance responses to context presentations over
the Early phase (trial 1
-
8), Middle Phase (trial 9
-
16) and Late Phase (trial 17
-
24) and testing a
Phase (Early, Middle, Late) x G
roup (Placebo, Propranolol) 3x2 ANOVA revealed no main
effects and no interactions

Thus during the Early Phase the Placebo group shows greater responses to the CS+
compared to the CS
-
, whereas this effect is absent in the Propranolol group. Both groups exh
ibit
no differential responses in the Late Phase. Groups show no difference in responses to context
presentations (
Figure 5).

Reinstatement

Reinstatement effects are generally only observed for the first few trials
(
Kindt et al., 2009
)
.

Therefore we
calculated for each CS type (CS+, CS
-
) the Early Phase as the average skin
conductance response over the trials 1
-
4, a Middle Phase as the average skin conductance
response over the trials 5
-
8, and the Late Phase as the average response over trial 9
-
12. F
irst, a
CStype (CS+, CS
-
) x Phase (Early, Middle, Late) with Group (Propranolol, Placebo) as a
between subjects variable 2x3x2 repeated measure ANOVA revealed a main effect of CStype
(F
1, 45

= 7.206,
p

= 0.010), a main effect of Phase (F
2, 90

= 4.247,
p

=
0.017), an interaction effect
of Phase x Group (F
1, 46

= 3.280,
p

= 0.042), with no other main effects or interactions.

Again, we calculated the difference in average responses to the CS+ and CS
-

for the
Early, Middle, and Late phase. Paired T
-
tests revealed no differences in the differential responses
in either group for each of the Phases. However, independent samples T
-
t
ests reveal that the
Placebo group exhibits during the Early Phase greater response to the CS+ than the Propranolol
group at trend (
t

(41.895
)

= 1.712,
p

= 0.094; Early Phase Placebo Group: 0.460, s.e.m., 0.084,
Early Phase Propranolol Group: 0.284, s.e.m.
, 0.059), as well as to the CS
-

(
t

(45
)
=
1.687,
p

=
0.099; Early Phase Placebo Group: 0.338, s.e.m., 0.055, Early Phase Propranolol Group: 0.206,
s.e.m., 0.055).

Finally, calculating the average skin conductance responses to context presentations over
the
Early phase (trial 1
-
8), Middle Phase (trial 9
-
16) and Late Phase (trial 17
-
24) and testing a
Phase (Early, Middle, Late) x Group (Placebo, Propranolol) 3x2 ANOVA revealed a main effect
of Phase at trend (F
1.754, 78.947
= 3.232,
p

= 0.051), with
no
other main effects and no interactions.

Thus during the Early Phase the Placebo group shows greater reinstatement of fear,
expressed as increased responses to both the CS+ and CS
-

compared to the Propranolol group in
the Early Phase. Both groups exhibit n
o differential responses in the Late Phase. Further, the
groups show no difference in responses to context
presentations (
Figure 6)
.

21



Figure 5
.

Skin conductance Responses for extinction recall
. Square root transformed SCRs for the placebo (thick
line) and propranolol group (dotted line) for CS+ (red) and CS
-

(blue). Contray to the placebo group which shows a
spontaneous recovery of fear, the propranolol group does not differentiate between CS+
and CS
-

and shows an
reduction of fear to both CS+ and CS
-
.



Figure 6
.

Skin conductance Responses for reinstatement of fear
. Square root transformed SCRs for the placebo
(thick line) and propranolol group (dotted line) for CS+ (red) and CS
-

(blue). The
propranolol group shows
decreased fear responses even after the unsignaled presentation of shocks.

Neural Activation

Fearful vs safe stimuli (CS+ >CS
-

and
CS
-
>CS
+)

During extinction learning the
differential
SCRs observed between the CS+ and the CS
-

gradually
diminished
and no difference
s

were

observed
on
the last trials
(Figure 4)

proving that
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Plac CSp
Plac CSm
Prop CSp
Prop CSm
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
2
3
4
5
6
7
8
9
10
11
12
Plac CSp
Plac CSm
Prop CSp
Prop CSm
22


at
behavioural

level extinction learning took place.
At

a neural level, this can be investigated
with the contrasts CS+ vs CS
-

and CS
-

vs CS+. The first contrast
is related to fear responses and
revealed significant activation in the following brain regions: the right Insula (44, 2, 0, p= 0.000),
the left Insul
a (
-
40, 8,
-
2, p=0.003), and the
dorsal
Anterior Cingulate

cortex

(
-
6, 32, 20, p= 0.007)
and
the Midbrain (
-
2
-
26
-
6, p=0.021 after SVC based on
(
Hermans et al., 2011
)

and application
of 10mm radius sphere
)
(Figure 7
; Table 1
)
. The second contrast
drives safety responses

and
revealed significant effect in vmPFC (
-
2, 42,
-
22, p= 0.002 after SVC based on
(
Kalisch,
Korenfeld, et al., 2006
)

and
application of 10
mm
radius

sphere
)
,
(Figure
8;

T
able 1
)
.

Time effects (Time x CS)

To investigate the influence of time on extinction learning, we calculated brain activation for the
contrasts (Early Phase > Late Phase) and (Late Phase >

Early Phase). A significant activation
was found only for the first contrast. More specifically brain regions involved in the (Early Phase
> Late Phase) contrast were: the
Superior temporal gyrus
, the
Midle temporal gyrus

(Right
angular gyrus
), the Right
Superior temporal Gyrus
, and the Right
Angular Gyrus

[IPC (PGp)].
No other interaction revealed significant results
(see T
able 1).

Drug effects


At a behavioural level we found that
the two groups differed during the early phase of the
extinction learning task:
during the early Phase the Placebo group shows greater responses to the
CS+ compared to the CS
-
, whereas this effect is absent in the Propranolol group:

the propranolol
group d
id not differentiate between
CS
+ and C
S
-
.
Additionally, b
oth groups exhibit no
differential responses in the Late Phase.
As a

matter of fact
,

t
he two groups

also

differed over th
e
course of extinction learning: the propranolol

group did not differentiate between the CS+ and
CS
-

all over the extinction task, while the placebo group initially showed
greater responses to
CS+ vs CS
-

and consecutively this

differentiation was eliminated
.
To investigate drug effects at
neuronal level
, we tested for the three way interaction Drug x Phase x CS and a two way
interaction effect Drug x CS. These interactions didn't reveal significant results (see table 1).
Nonetheless,

since the two groups differed

over t
he course of extinction learning

re
sults at a
behavioural level show that the drug effects exist, but might be difficult to detect the effects at a
neuronal level. Therefore, we next investigated brain activation to the conditioned stimuli
separately for each drug group.

More specifically t
he contrast
CS+ vs CS
-

for the placebo group
revealed significant activation in the Anterior Cingulate cortex (
-
8, 32, 18, p= 0.036)

(Figure 9)

and
the contrast
CS
-

vs CS+

for the placebo group revealed activation that did not survive the
significant threshold. As for the

propranolol group, the contrast
CS+ vs CS
-


revealed
activation
in

the midbrain (
-
4
-
18
-
14, p= 0,017,
after SVC based on
(
Hermans et al., 2011
)

and application
of 10mm radius sphere),

(Figure 10). Finally, the contrast
CS
-

vs CS+

for the propranolol group
revealed
no significant
activation.

Since at behavioural level the two drug groups differed during the early phase of the extinction
learning, and because,
as
mentioned above, the contrast Early > Late p
hase led to significant
23


results at a

neural level,

we
investigating neuronal activity during the early phase. Critically, the
contrast early CS+ vs CS
-

showed significant activation in the dACC and the midbrain.
Next

we
calculated the neural activity for this contrast separately for each group (contrast early

CS+ vs
CS
-
). Once more we found involvement of the dACC

only
for the placebo group
and at the

midbrain
only
for the propranolol group. Critically, activation at the dACC is absent for the
propranolol group even when lowe
ring the threshold at p< 0.001

whil
e for the
placebo
group
all
involved regions (not s
ign
ificant

activation) are in brain regions
of the ACC.
This signifies a
possibility of double dissociation in neuronal activity between the two groups with the dACC
being
important for the placebo group
and absent/ overruled by the mid
brain in the propranolol
group.



Figure 7
.

Extinction learning. Activation associated with extin
ction
learning on experimental D
ay 2. Images show group
-

level estimates
for the contrast (SC+>CS
-
), display threshold p,0.01
. Activation in
Right Insula, Left Insula, and the Midbrain.





Figure 8
.

Extinction learning. Activation associated with extin
ction
learning on experimental D
ay 2. Images show group
-

level
estimates for the contrast (SC
-
>CS+), display threshold p,0.01
,
masked for vmPFC at the coordinates(
-
2 42
-
22) with a sphere of
10mm radius. Activation in vmPFC.





Effects of
interests’ analysis

Since our critical contrasts revealed significant activation in the vmPFC,
the left and right Insula,
the m
idbrain and the dACC, we wanted to see wh
ether and in which way these

areas are
differentially activated for the two drug groups. We therefore performed an effect of interest
analysis for these regions. dACC and vmPFC

showed increased differential activat
ion between
24


CS+ and CS
-

in the placebo group (four last bars, Figure 9 & 11) while this effect is absent in
the propranolol group (four last bars, Figure 9 & 11). This finding is in line with the SCR results
(Figure 4) and suggest that the
dACC and the vm
PFC play an important role in the SCR
observed for the placebo group but not for the propranolol group.
Activation in the left and right
Insula
shows same
pattern

of activation for both placebo and propranolol group
, therefore no
group differences are obse
rved for insula cortex
.

Finally, a
ctivation
and
differential

activation
in

the m
idbrain

is higher for the propranolol
.







Figure 9
.

Extinction learning. Activation associated with extin
ction learning on experimental D
ay 2.
a,

Images show
group
-

level estimates for the contrast (SC+>CS
-
) for the placebo group only; display threshold <p,0.01. Activation
in the ACC.
b,

Effects of interest analysis. Estimation of ACC activation for CS+ >CS
-
, in early (E) and late (L)
phase, for pla
cebo and propranolol group.
c,

SCR

for the Extinction learning task. SCR has the same pattern as
brain activity in
d
ACC.



25







Figure 10

Extinction
learn
ing. Activation associated with
extin
ct
ion learning on experimental D
ay 2.
a,
Images show
group
-

level estimates for the contrast (SC+>CS
-
) for the propranolol group alone; display threshold<p,0.01.
Activation in the midbrain.
b,

Effects of interest analysis.
Estimation of midbrain activation for CS+, CS
-
, in early (E)
and late (L) phase, for the placebo and the propranolol group.

c,
Activity in the midbrain
does not correspond to the
SCR

pattern
.



Figure
11
.

Extinction learning.

Activation associated with extin
ction learning on experimental D
ay 2.
a,
Images
show group
-

level estimates for the contrast (SC
-
>CS+) ; display threshold <p,0.01. Activation in the vmPFC.
b,

Effects of interest analysis. Estimation of vmPFC activation for
CS+, CS
-
, in early (E) and late (L) phase, for the
placebo and the propranolol group.
c,

SCRfor the Extinction learning task. SCR has the same pattern as brain activity
in vmPFC.

26



Figure 12
.

Extinction learning.

Activation
associated with extinction learning on
experimental day 2.
a,
Images show group
-

level

estimates for the contrast (SC+>CS
-
)
for the placebo group alone; display
threshold
<p,0.01. Activation in the right
and left insula
.
b,

Effects of interest an
alysis.
Estimation of vmPFC activation for CS+,
CS
-
, in early (E) and late (L) phase, for the
placebo and the propranolol group.
c,

SCRfor the Extinction learning task. SCR
does not correspond to the

brain activity

pattern observed in right and left the i
nsula.













Figure 13
. Extinction learning: Drug effects

in the Early phase
. Activation associated with extinction learning on
experimental day 2.
a,
Images show group
-

level estimates for the contrast early (SC+>CS
-
); display threshold
<p,0.01.
Activation in the dACC and in
the Mi
dbrain.

B

Images show group
-

level estimates for the contrast early
(SC+>CS
-
) separately for the placebo group

(A)

and for the propranolol group

(B)
; display threshold <p,0.01.
Activation

only

in the dACC

for the placebo group and in the Midbrain for the propranolol group

(in
line with a
double dissociation trend)
.

27


DISCUSSION


Conclusions


In line with our hypothesis, our results suggest that extinction learning is mediated by the
vmPFC, the dACC, the
insula, and the midbrain. Contrary to our

expectations that NA

blockade
prior to extinction learning would impair extinction re
call and thus lead to increased fear,
we
report that
beta blockade abolishes differentially conditioned responses during extincti
on
learning and subsequently prevents the return of fear 24 hours later. These effects are attributable
to changes in the neural network of extinction, where we find propranolol to affect activity in the
dACC
and the midbrain.


Summary

Both groups were
equally conditioned on Day1 and showed differential SCR responses to CS+
and CS
-
.

On Day 2, administration of propranolol led to elimination of differentiation between
CS+ and CS
-

at the beginning

of the extinction training

and influenced the course of ext
inction
learning. While the placebo group showed differential responses to the CS+ and CS
-

during the
early phase of the extinction learning task, the propranolol group showed no differential
responses. Over the course of the extinction task
,

the placebo g
roup learned

to inhibit their fear
responses,

therefore

both groups
show
ed

no differential fear responses at the end of the
extinction task.

The effects of propranolol endured over time.

On Day 3, fear was still reduced for the

propranolol

group while the placebo group exhibit
ed

a spontaneous recovery of fear.
Interestingly, while the placebo group showed greater responses to the CS+ compared to the CS
-
,
the propranolol group

not only did not differentiate between CS+ and CS
-
, but also show
ed a
general reduction of fear as signalled by the increased SCR reduction to both CS+ and CS
-
.
Finally, fear remain
ed

low for the propranolol group
even after unsignaled administration of
shocks, while

the placebo group showed greater reinstatement of fea
r expressed as increased
responses to both the CS+ and CS
-

.

At a neural level, during extinction learning
,

activation was found in the vmPFC for the
contrast CS
-

vs CS+, and in the insular cortex, the midbrain and the dACC for the contrast CS+
vs CS
-
. Th
is is in line with previous neuroimaging studies.
As for

the drug effects, propranolol
seems to affect the dACC
, thereby

lead
ing

to reduction and elimination of fear 24 hours later.
Whereas

involvement of the dACC

appeared
t
o be blocked and a significant
activation was
found in the midbrain

for the propranolol group
, for the placebo group significant activation was
found in the dACC.




28


Neural mechanisms of extinction learning

There is a notable similarity between the location of the brain areas activated during extinction
learning in the present study and locations previously reported in structural and functional
neuroimaging studies
(
Milad, Wright, et al., 2007
;
Phelps et al., 2004
)
. More concretely, the
contrast CS
-

vs CS+
revealed involvement of vmPFC. Although it has been suggested that
vmPFC is particularly involved during recall of extinction,
therefore

after

initial acquisition of
extinction learning
(
Phelps et al., 2004
)
,

our results
indicate
that vmPFC invol
vement during
extinction learning may be necessary for l
ong
-
term encoding and

retrieval of extinction

and are
in line with work conducted by

(
Milad, Wright, et al., 2007
)
. Contrary to
(
Milad, Wright, et al.,
2007
)

and
(
Milad & Quirk, 2002
)
,

and in agreement with
(
Phelps et al., 2004
)
,

the vmPFC was
deactivated during extinction learning. The responses observed in the vmPFC were primarily
expressed as a decrease to the CS+. Although this has been assessed relatively to the CS
-
,

the
effect of interest analysis pattern

indicates that this di
fferential response was mostl
y driven by
depression in BOLD to the CS+.

D
ecrease in BOLD response is difficult to interpret;
probably it
reflect
s

a reduction in neuronal activity
(
Shmuel et al., 2002
)
. For the contr
ast CS+ vs CS
-

(which signals fear) there was activation
in

the dACC

(in line with
(
Phelps et al., 2004
)

(
Lang et
al., 2009
)
, the midbrain and the bilateral insular cortex

(
Gottfried & Dolan, 2004
;
Phelps et al.,
20
04
;
Yágüez et al., 2005
)
.
Activation in the midbrain was recently shown to be involved in
exposure of fear
-
related acute stressors
(
Hermans et al., 2011
)
,

but

it

has not been yet shown in
other fMRI studies examining fear extinction. It might be that the high resolution analysis used
in our experiment capture
d

activ
ation in the PAG

that drives fear or the locus coeruleus

(LC)
, the
main generator of
noradrenaline. Note that both regions are very small and difficult to capture
with fMRI.

We replicate the finding that
,

within
m
PFC
,

the dACC and the vmPFC play different
roles in fear extinction. Our results are in line with a current review paper stati
ng that during
extinction learning there is a differentiation between dorsal ACC and mPFC

subregions
, which
are implicated in threat appraisal and the expression of fear, and ventral ACC and mPFC

subregions, which are involved in the inhibition of conditio
ned fear through extinction
(
Etkin,
Egner, & Kalisch, 2011
)
.


Drug effects

Since at

a

behavioural level the two drug groups differed over the course of extinction learning,
with the
propranolol group not showing differential responses between CS+ and CS
-

throughout
the extinction learning task, and

with

the placebo group differentiating between the CS+ and CS
-

at the beginning of the task and not at the end, we investigated the
drug e
ffects at a neural level

for each group separately. Critically
,

the contrast CS+ vs CS
-

revealed significant effect
in

the
dACC

for the placebo group
, and significant effect
in

the midbrain for the propranolol group,
thereby
showing a double dissociation i
n neuronal activity between the two groups.

29


At behavioural level
,

the two drug groups differed during the early phase of the
extinction learning
.

T
herefore, what is happening in a neuronal level at this early phase is
of
much importance
. The early phase of extinction expresses recall of the conditioned fear memory,
as the new, safe associations are not created yet: they are only created during the task and
especially
in

the late phase
(
Milad, Wright, et al., 2007
)
.
Crucially
, the contrast early CS+ vs CS
-

showed again significant activation in the dACC and the midbrain.

W
e

then

calculated the neural activity for this contrast

(early CS+ vs CS
-
)

separately for
each group. Once more we found involvement of the dACC for the placebo group and the
midbrain for the propranolol group.
A
ctivation
in

the dACC is absent for the propra
nolol group
even when lowering the threshold at p< 0.001. For the propranolol group we find involvement of
the midbrain, while for the placebo group all
the

regions

involved

(even when not significantly
activated) are part of the ACC. These findings suggest a quasi
-
double dissociation between the
two groups: dACC is activated in the placebo group but is not present in the propranolol group,
and the midbrain is
up regulated

in

the propranolol group but not in the placebo group.

Our contrasts did not reveal drug effects
in

the vmPFC.
Still
, the region of interest
analysis suggests that differentia
l

activity

of the vmPFC

to CS
-

and CS+ is decreased for the
propranolol group. So
although the placebo group vmPFC activation shows a clear differential
pattern between CS+ and CS
-
, the propranolol group does not differentiate
substantially
.


What is propranolol doing and how?

In line with other studies

(
Kindt et al., 2009
)
,
(
Rodriguez
-
Romaguera, Sotres
-
Bayon, Mueller, &
Quirk, 2009
)

we showed that propranolol led to elimination of fear 24 hours after administration.
What could

then

be the mechanism of propranolol action? What happens during e
xtinction
learning that leads to loss of fear the next day?

To begin with,

we see that
in

the early phase of extinction learning the propranolol group
does not differentiate between CS+ and CS
-
.A
t a behavioural level, only the placebo group
shows differential activation for CS+ vs CS
-
. At a neural level
,

this contrast le
ads

to activation in
the dACC only for the placebo group,
indicating

that dACC

may

play an important role in the
differential SCR
s observed for the placebo group but is absent in the propranolol group. Could
propranolol effects on dACC during extinction learning lead to the decreased fear responses
observed the next day?

According to

Rainbow and colleagues

(
Rainbow, Parso
ns, & Wolfe, 1984
;
Reznikoff,
Manaker, Rhodes, Winokur, & Rainbow, 1986
)
,

there are noradrenergic receptors in the human
and animal dACC. Influence of dACC adrenergic receptors by propranolol

could lead to
dysfunction of dACC.
Moreover
, recent studies have revealed direct anatomical projections from
the ACC to the LC, the mi
d
brain

neuromodulatory nucleu
s responsible for the majority of the
NA

release in the brain
(
Aston
-
Jones & C
ohen, 2005
)
. At the same time, the dACC receives
signals from the midbrain (LC) via noradrenergic signalling
(
Aston
-
Jones & Cohen, 2005
)
, so
noradrenergic blockade
may

influence the signals sent to the dACC by the LC, which is
in
30


agreem
ent

with our
finding

that in the propranolol group there was increased activation
in

the LC
(as a compensatory mechanism)
,

but not
in

the dACC. Our results
do not

disclose

the
directionality of this impact (whether LC affects dACC

or blockade of noradrenergic receptors
in

the dACC affect LC), and this is out of the scope of the current study.
However
, our results
suggest that dACC is affected in one way or the other
,

and this is in line with
several

theories
that highlight the rela
tionship between dACC and LC
:

“Gain modulation theory”
(
Aston
-
Jones &
Cohen, 2005
)
;“unexpected uncertainty theory”
(
Yu & Dayan, 2005
)
.

The dACC plays an important role in

valence appraisal
(
Etkin et al., 2011
;
Kalisch,
Wiech, Critchley, & Dolan, 2006
)
, error detection

(
Bush, Luu, & Posner, 2000
;
Gehring, Goss,
Coles, Meyer, & Donchin, 1993
)
, expectancy
(
Olive
ira, McDonald, & Goodman, 2007
)

and
predictions
(
Brown & Braver, 2007
)
.

Impairment of dACC function by propranolol would impai
r
differential valence appraisal of CS+ and CS
-
, and would lead to same response to both CS+ and
CS
-
.
O
ur results support this
hypothesis on the basis of the fact that

there is no differentiation
between CS+ and CS
-

for the propranolol group. In line with
this, prior work has shown that
lesions in ACC attenuate SCRs
(
Tranel & Damasio, 1994
)
.

U
nder circumstances of affected ACC
,

the extinction learning task
takes

place:
in

the
early phase of the extinction task the fearful memory is retrieved, and
in

the late phase creation
of safety trace occurs
(
Myers & Davis, 2006
)
. Retrieval of a consolidated fear memory can
return the memory to a labile state
(
Lee et al., 2008
;
Nader et al., 2000
)
. During this vulnerability
phase there is a ubiquitin
-

and proteosome
-

dependent protein degradation disrupting the pre
-
existing memory
(
Lee et al., 2008
)
,

followed by a protein synthesis dep
e
ndent process of
restabilis
ation that returns the memory to a fixed state (
reconsolidation). T
he transient memory
labiali
s
ation phase allows for the modification of memory

(
either strengthening or weakening
).

A
s mentioned
, this

occurs when a memory is being reactivated
(
Nader et al., 2000
)

and when a
new information is being integrated
(
Sevenster, Beckers, & Kindt, 2012
;
Wang & Morris, 2010
)
.
This process

can take place during extinction learning: the old fear memory is retrieved (early
phase of extinction learning), is updated and ‘in dialog’ with the new safety memory
and
a new
safety trace is created (late phase of extinction learning). The idea that du
ring extinction
unlearning

(e.g.
,

weakening of fear memory) and
new learning

(e.g.
,

creation of safe
ty trace) can
coexist has been proposed many decades ago.
Research ha
s

suggested
that a strict inhibitory
view

of extinction is
an oversimplification
because

ignores that CR recovery is
not always
complete
. This suggests

that some degree of erasure occurs

always

(
Delamater, 2004
;
Myers &
Davis, 2006
;
Rescorla, 2001
)
.
I
n line with this,
(
Lee et al., 2008
)

demonstrated that extinction
processes and memory reconsolidation share common molecular mechanisms at the synaptic
level
,

and stressed the fact that new learning underlies certain forms of memory extinction
(
Kaang, Lee, & Kim, 2009
)
.

It has been shown that updating of a fear memory trace occurs when an outcome is not
fully predictable and does

not

occur when an outcome is fully predictable

(
Pedreira, Pérez
-
Cue
sta, & Maldonado, 2004
;
Sevenster et al., 2012
)
. Thus
,

ambiguity increases chances for
reconsolidation. In our paradigm
,

the outcome was not fully predictable for both groups, but it
31


seems that propranolol action over dACC increased ambiguity and
likewise the

updating time
window,

thus

making
the modulation of fear memory more
likely
. Indeed, it has been shown that
propranolol affects reconsolidation of fear memories and leads to loss of fear
(
Dębiec & Ledoux,
2004
;
Kindt et al., 2009
)
.

Since for the propranolol group fear memory retrieval occurs while the dACC is affected
(
i.e., under

circumstances of emotional valence ambigui
ty) and no shock is administered,
weakening of prior fear memory is facilitated. Consecutively, inhibition of fear memory is
further enhanced by

the

creation of a safety trace (extinction trace). Note that
,

although

beta
adrenergic receptors are blocked, e
xtinction learning can still happen (at least partially) via other
receptors
, for example

D
-
5 dopaminergic receptors
(
Ouyang, Young, Lestini, Schutsky, &
Thomas, 2012
)
. If during this phase of liability and ambiguity a shock
had

been administered to
the propranolol

group, the fear trace would have been strengthened instead of weakened. This
might explain the discrepancy observed in the field where propranolol has been shown to
increase fear responses in some cases, and especially in cases of spaced conditioning
(
Cain et al.,
2004
)
. In cases of spaced extinction
,

the reassurance of no shock reception would not be strong
enough to drive ambiguity towards the direction of safety trace. In line with this argumentation
,
Cain and colle
a
gues showed that spaced extinction training leads to increased fear incubation
and under such circumstances propranolol increases fear
(
Cain et al., 2004
)
.

To sum up, for the
propranolol group retrieval and updating of fear memory occurs under
circumstances where dACC is affected
,

thus promoting weakening of fear trace during recoding.
Additionally, creation of safety trace (partially) happens. A summation of these two events l
eads
to enhanced safety memory and decreased fear observed 24 hours later and explains the loss of
fear observed for the propranolol group.


Alternative explanations

Increased extinction learning

An alternative explanation of why the propranolol

group shows loss of fear on Day 3

may

be that
propranolol
improves

extinction learning. This explanation

is problematic

because it has been
long shown that
NA
improves learning

(
Hu et al., 2007
;
Mohammed, Jonsson, & Archer, 1986
)
.
When exploring for drug effects at

a
neural level,

we found

the dACC

to be

activated for the
placebo group
in

the early phase and during the course of the extinction learning task
.

T
his area
was
not activated

for the propranolol group, signalling that propranolol is affecting/blocking
dACC

activity. For the propra
nolol group
,

we found significant activation in the midbrain
,

possibly as an upregulating/o
vercompensating effect due to NA

blockage and dACC influence.
These findings suggest the important role of dACC and midbrain (LC) interaction during
emotional learni
ng. Yu and Dayan have proposed that estimates of unexpected uncertainty
-
which promote a revision of expectations
-

are meditated by NE
-
LC
-
ACC system and that ACC
plays an important role in detecting conflict in incongruent/unexpected trials,

thereby

promo
ting
learning
(
Yu & Dayan, 2005
)
. Thus noradrenergic blockage affects the ACC
-

midbrain
32


interaction and impairs extinction learning (at least up to a level). On the other hand, as
mentioned above, extinction learning can still partially occur: when beta adrenergic receptors are
blocked,
extinction of fear can partially take place via the D
-
5 dopaminergic receptors

(
Ouyang
et al., 20
12
)
. The argument that extinction learning is partially
increased

is in line with our
results on Day 3: if there w
ere

no extinction trace created on Day 2, we might have observed
reinstatement of fear for the propranolol group on Day 3. A summation of a

(partially occurring)

extinction (safety) trace and an impaired re
-
encoded fear trace (as analysed
in

the
previous
section)

leads to the enhanced extinction trace on Day 3 (impairment of fear trace + creation of a
weak extinction trace

(on Day 2)

= increa
sed fear
reduction
on Day 3).

Impairment of contextual fear

Some studies have suggested that propranolol acts via reducing contextual conditioned fear
(
Grillon, Cordova, Morgan, Charney, & Davis, 2004
;
Ji, Wang, & Li, 2003
)
. Reduction of
contextual fear would indeed lead to fear reduction for the
propranolol group if the
non
differentiation between the CS+ and CS
-

(the cues) in a specific context is taken into account.
The new context is shown during extinction learning (Day 2). In order for context B to become
associated with a differential conditi
oned response, on Day 2 a differential conditioned response
has to be expressed in context B.
Consecutively,
the differential response becomes associated
with context B as a result of new

learning and generalization. Since
p
ropranolol blocked

the
different
ial expression to CS+ and CS
-
, context B
would not
become associated with a
condit
ioned response (on Day 2), therefore leading

also

to

the
no
n

differential responses

on Day
3. Contrary to
(
Grillon et al., 2004
)
,

who argued that acquisition and retention of cued fear
conditioning were not affected by propranolol, our results take into account the cued and
contextual responses
jointly
.



Discussion of
the
validity of our findings and

of the

limitations of the st
udy

We hypothesise that extinction learning relies on the “extinction network” involving the vmPFC,
the ACC, the amygdala, the insula and the midbrain and that raised noradrenaline levels are
critical for retention of safety associations. Therefore,

noradrenaline blockade prior to extinction
learning should impair extinction recall and lead to increased fear.

We found that vmPFC is

indeed

involved in extinction learning together with

standa
rd extinction regions”
including

the
dACC, the insular cortex

and the midbrain.
C
ontrary to our expectations,

however,

blockade
prior to extinction learning did not impair extinction recall: fear was decreased in the propranolol
group. It might be that noradrenergic blockade reduces extinction learning
.

B
ut since it

also
affects retrieval of fear memory, the
summation of these competing

forces
eventually
leads to
reduced fear and thus stronger extinction trace.

We did not find activation in amydgala. There are two explanations for this event
.

The
first one is that
amygdala gets habituated fast
(
Breiter et al., 1996
)
.

For this reason, a
lthough

present, it cannot be captured by neuroimaging methods. Indeed activation in amygdala during
33


conditioning and e
xtinction degrades as time passes by
(
LaBar et al., 1998
)
. Secondly, it has
been shown that propranolol affects amygdala
(
Dębiec & Ledoux, 2004
)
. Since propranolol was
administered systematically, amygdala
may

have been affected thus
becoming

not active. If the
latter

were true, though, one would expect to find amygdala activation in the placebo group as
opposed to the propranolol group.
With all that said
, none of the abovementioned explanations
can be excluded with certainty
.


Finally, a
further

limitation of this study is that propranolol was administered
systematically; therefore drug effects in specific regions cannot be fully controlled as it is the
case with animal studies where drugs are injected.
It should be pointed out
,

however, t
hat

since
injection of propranolol in human brain is not ethically allowed and the effect of propranolol is
quiet specific (block
ade of

beta adrenergic receptors), this study control
led

for drug specificity as
much as possible.


Conclusions and

implicatio
ns

Anxiety and fear related disorders have been called “the epidemic of the 21
st

century” and are
affecting a large number of people worldwide.

It still remains a significant challenge to find the
most efficient treatment for fear and anxiety disorders, possibly by
optimising
the combination of
psycho
-

and pharmaco
-
therapy.
The findings of our study contribute to the current scientific
knowledge
o
f

a highly interesting and much
-
debated topic; namely the erasure of fearful and
stressful memories by pharmacological and behavioural agents

(
such as (re)consolidation
paradigms
)

and extends these finding from single reactivation to repeated memory reacti
vation
.
It is the first study that shows erasure of fear when combining propranolol with multiple
memory
reactivations (
extinction
)

instead of single reactivation.
In addition,

o
ur research
bridges the gap
between fundamental research and preclinical
research

that

focus
es

on patients with anxiety
disorders.

The e
ffectiveness of combining NA

blockers with extinction and thus psychotherapy
has been questioned due to mixed results of research investigating this topic. Yet,
our findings
indicate that beta

adrenergic blockade eliminates fear and prevents

the return of fear via the
impedance of differential responses to CS+ and CS
-

during extinction learning
.
Moreover
, it
highlight
s

the involvement of dACC in this
process
. These results provide face validity
to
clinical interventions employing beta
-
adrenergic antagonists in conjunction with extinction
learning during psychotherapy

to
increase the effectiveness of the treatment of stress and anxiety
disorders.








34


Contrast


MNI Coordinates

Peak
voxel



Region

BA

x

y

z

T

z
-
score
s

cluste
r
size

CS+ vs CS
-

Insula

(right)

area 13

44

2

0

4.74

4.59

581

Anterior c
ingulate

area 24

-
6

32

20

4.56

4.43

361

Insula

(left)


-
40

8

-
2

4.36

4.25

418

Midbrain


-
2

-
26

-
6

3.45

3.39

29

CS
-

vs CS+

Rectal g
yrus


-
4

38

-
22

4.19

4.09

119

Parietal Lobe (Post central

g
yrus)


28

-
34

54

4.07

3.98

92

Early vs Late

Superior temporal gyrus


60

-
40

16

5.38

5.18

62

Middle temporal gyrus


area 39

44

-
60

24

5.22

5.03

56

Right
superior temporal g
yrus


46

-
44

16

5.02

4.85

2

Right
angular g
yrus



48

-
68

32

4.87

4.71

1

Insula

(right)


30

20

4

4.79

4.64

1

Early (CS+ vs CS
-
)

Left anterior cingulated cortex


-
6

32

22

3.87

3.79

127

Midbrain


-
8

-
22

-
14

3.49

3.43

9

Plac
ebo

(CS+ vs CS
-
)

Anterior c
ingulate


-
8

32

18

4.29

4.18

239

Inferior frontal gyrus


36

24

-
4

3.62

3.55

81

Insula (l
eft)


-
38

14

-
4

3.61

3.54

121

Insula (r
ight)


44

2

0

3.29

3.24

3

Caudate nucleus


10

38

18

3.18

3.13

1

Prop
ranolol

(CS+ vs CS
-
)

Midbrain


-
6

-
14

-
14

4.29

4.18

189

Insula

(right)


42

6

-
2

3.70

3.63

38

Insula (left)


-
42

4

-
2

3.54

3.48

20

Placebo
(
early CS + vs CS
-
)

A
nterior cingulate


-
10

32

20

3.40

3.34

9

A
nterior cingulate

(left)


0

42

18

3.28

3.23

14

A
nterior cingulate



-
10

24

24

3.16

3.11

1

Propranolol
(
early CS+ vs CS
-
)

Midbrain


8

-
14

-
16

3.49

3.43

13

Midbrain


-
8

-
12

-
18

3.45

3.39

11

Left temporal g
yrus


-
48

6

0

3.15

3.11

1

35


Table 1:
Neural activation during extinction learning task.

Activity differences survived multiple comparisons
correction, except from the areas listed in italics which were significant at a lenient threshold of p<0.001 but did not
survive correction for multiple comparisons. Therefore, these areas are solely list
ed for reference. Cluster size is
given in number of voxels.



ACKNOWLEDGEMENTS


I would like to thank the Memory and Emotions group for many inspiring discussions and for
fostering my critical thinking. I am grateful to Atsuko Takashima, Niels ter Huurne

and Sabine
Kooijman for their medical advice, to Susanne Vogel for assisting with data collection, and to
my colleagues at the Donders Center for Cognitive Neuroimaging

and Giovanni Rossi

for
helpful discussion
s

and support when wo
rking during

late hours.

Last but certainly

not least
,
special thanks go to
my supervisors Guillén
Fernández

and Marijn Kroes for

guiding

me through
the world of cognitive neuroscience and
teaching me how research life is.






36


REFERENCES

Aston
-
Jones, G., &
Cohen, J. D. (2005). An integrative theory of locus coeruleus
-
norepinephrine function:
adaptive gain and optimal performance. [Research Support, N.I.H., Extramural

Review].
Annu Rev Neurosci, 28
(1), 403
-
450. doi: 10.1146/annurev.neuro.28.061604.135709

Bos,

M. G. N., Beckers, T., & Kindt, M. (2012). The effects of noradrenergic blockade on extinction in
humans.
Biological Psychology, 89
(3), 598
-
605. doi: 10.1016/j.biopsycho.2012.01.007

Breiter, H. C., Etcoff, N. L., Whalen, P. J., Kennedy, W. A., Rauch, S. L
., Buckner, R. L., . . . Rosen, B. R.
(1996). Response and Habituation of the Human Amygdala during Visual Processing of Facial
Expression.
Neuron, 17
(5), 875
-
887. doi: 10.1016/s0896
-
6273(00)80219
-
6

Brinley
-
Reed, M., Mascagni, F., & McDonald, A. J. (1995).

Synaptology of prefrontal cortical projections
to the basolateral amygdala: an electron microscopic study in the rat.
Neuroscience Letters,
202
(1

2), 45
-
48. doi: 10.1016/0304
-
3940(95)12212
-
5

Brown, J., & Braver, T. (2007). Risk prediction and aversion by
anterior cingulate cortex.
Cognitive,
Affective, & Behavioral Neuroscience, 7
(4), 266
-
277. doi: 10.3758/cabn.7.4.266

Brunet, A., Orr, S. P., Tremblay, J., Robertson, K., Nader, K., & Pitman, R. K. (2008). Effect of post
-
retrieval propranolol on psychophysi
ologic responding during subsequent script
-
driven
traumatic imagery in post
-
traumatic stress disorder.
Journal of Psychiatric Research, 42
(6), 503
-
506. doi: 10.1016/j.jpsychires.2007.05.006

Bush, G., Luu, P., & Posner, M. I. (2000). Cognitive and emotional

influences in anterior cingulate cortex.
Trends in Cognitive Sciences, 4
(6), 215
-
222. doi: 10.1016/s1364
-
6613(00)01483
-
2

Cahill, L., & McGaugh, J. L. (1998). Mechanisms of emotional arousal and lasting declarative memory.
Trends in Neurosciences, 21
(7), 2
94
-
299. doi: 10.1016/s0166
-
2236(97)01214
-
9

Cain, C. K., Blouin, A. M., & Barad, M. (2004). Adrenergic Transmission Facilitates Extinction of
Conditional Fear in Mice.
Learning & Memory, 11
(2), 179
-
187. doi: 10.1101/lm.71504

Chiba, T., Kayahara, T., &

Nakano, K. (2001). Efferent projections of infralimbic and prelimbic areas of
the medial prefrontal cortex in the Japanese monkey, Macaca fuscata.
Brain Research, 888
(1),
83
-
101. doi: 10.1016/s0006
-
8993(00)03013
-
4

Cusack, R., Brett, M., & Osswald, K. (200
3). An Evaluation of the Use of Magnetic Field Maps to Undistort
Echo
-
Planar Images.
NeuroImage, 18
(1), 127
-
142. doi: 10.1006/nimg.2002.1281

Dębiec, J., & Ledoux, J. E. (2004). Disruption of reconsolidation but not consolidation of auditory fear
conditioni
ng by noradrenergic blockade in the amygdala.
Neuroscience, 129
(2), 267
-
272. doi:
10.1016/j.neuroscience.2004.08.018

Delamater, A. R. (2004). Experimental extinction in Pavlovian conditioning: behavioural and
neuroscience perspectives. [10.1080/02724990344
000097].
Q J Exp Psychol B, 57
, 97
-
132.

Etkin, A., Egner, T., & Kalisch, R. (2011). Emotional processing in anterior cingulate and medial prefrontal
cortex.
Trends in Cognitive Sciences, 15
(2), 85
-
93. doi: 10.1016/j.tics.2010.11.004

Friston, K. J., Glaser
, D. E., Henson, R. N. A., Kiebel, S., Phillips, C., & Ashburner, J. (2002). Classical and
Bayesian Inference in Neuroimaging: Applications.
NeuroImage, 16
(2), 484
-
512. doi:
10.1006/nimg.2002.1091

Gehring, W. J., Goss, B., Coles, M. G. H., Meyer, D. E., &
Donchin, E. (1993). A Neural System for Error
Detection and Compensation.
Psychological Science, 4
(6), 385
-
390. doi: 10.1111/j.1467
-
9280.1993.tb00586.x

Geracioti, J. T. D., Baker, D. G., Ekhator, N. N., West, S. A., Hill, K. K., Bruce, A. B., . . . Kasckow
, J. W.
(2001). CSF Norepinephrine Concentrations in Posttraumatic Stress Disorder.
American Journal
of Psychiatry, 158
(8), 1227
-
1230. doi: 10.1176/appi.ajp.158.8.1227

37


Gottfried, J. A., &

Dolan, R. J. (2004). Human orbitofrontal cortex mediates extinction learning while
accessing conditioned representations of value. [10.1038/nn1314].
Nat Neurosci, 7
(10), 1144
-
1152.

Grillon, C., Cordova, J., Morgan, C. A., Charney, D. S., & Davis, M. (200
4). Effects of the beta
-
blocker
propranolol on cued and contextual fear conditioning in humans.
Psychopharmacology, 175
(3),
342
-
352. doi: 10.1007/s00213
-
004
-
1819
-
5

Hermans, E. J., van Marle, H. J. F., Ossewaarde, L., Henckens, M. J. A. G., Qin, S., van Kes
teren, M. T. R., . .
. Fernández, G. (2011). Stress
-
Related Noradrenergic Activity Prompts Large
-
Scale Neural
Network Reconfiguration.
Science, 334
(6059), 1151
-
1153. doi: 10.1126/science.1209603

Hu, H., Real, E., Takamiya, K., Kang, M.
-
G., Ledoux, J., Huga
nir, R. L., & Malinow, R. (2007). Emotion
Enhances Learning via Norepinephrine Regulation of AMPA
-
Receptor Trafficking.
Cell, 131
(1),
160
-
173. doi: 10.1016/j.cell.2007.09.017

Hutton, C., Bork, A., Josephs, O., Deichmann, R., Ashburner, J., & Turner, R. (20
02). Image Distortion
Correction in fMRI: A Quantitative Evaluation.
NeuroImage, 16
(1), 217
-
240. doi:
10.1006/nimg.2001.1054

Hutton, C., Deichmann, R., Turner, R., & Andersson, J. L. R. (2004). Combined correction for geometric
distortion and its interacti
on with head motion in fMRI.
Proceedings of ISMRM, 12
, 1084.

Ji, J.
-
Z., Wang, X.
-
M., & Li, B.
-
M. (2003). Deficit in long
-
term contextual fear memory induced by
blockade of β
-
adrenoceptors in hippocampal CA1 region.
European Journal of Neuroscience,
17
(9),

1947
-
1952. doi: 10.1046/j.1460
-
9568.2003.02620.x

Kaang, B.
-
K., Lee, S.
-
H., & Kim, H. (2009). Synaptic Protein Degradation as a Mechanism in Memory
Reorganization.
The Neuroscientist, 15
(5), 430
-
435. doi: 10.1177/1073858408331374

Kalisch, R., Korenfeld, E.
, Stephan, K. E., Weiskopf, N., Seymour, B., & Dolan, R. J. (2006). Context
-
Dependent Human Extinction Memory Is Mediated by a Ventromedial Prefrontal and
Hippocampal Network.
The Journal of Neuroscience, 26
(37), 9503
-
9511. doi:
10.1523/jneurosci.2021
-
06.2
006

Kalisch, R., Wiech, K., Critchley, H. D., & Dolan, R. J. (2006). Levels of appraisal: A medial prefrontal role
in high
-
level appraisal of emotional material.
NeuroImage, 30
(4), 1458
-
1466. doi:
10.1016/j.neuroimage.2005.11.011

Kessler Rc, C. W. D. O. W.

E. E. (2005). PRevalence, severity, and comorbidity of 12
-
month dsm
-
iv
disorders in the national comorbidity survey replication.
Archives of General Psychiatry, 62
(6),
617
-
627. doi: 10.1001/archpsyc.62.6.617

Kindt, M., Soeter, M., & Vervliet, B. (2009). B
eyond extinction: erasing human fear responses and
preventing the return of fear. [10.1038/nn.2271].
Nat Neurosci, 12
(3), 256
-
258. doi:
http://www.nature.com/neuro/journal/
v12/n3/suppinfo/nn.2271_S1.html

Kirschbaum, C., Kudielka, B. M., Gaab, J., Schommer, N. C., & Hellhammer, D. H. (1999). Impact of
gender, menstrual cycle phase, and oral contraceptives on the activity of the hypothalamus
-
pituitary
-
adrenal axis. [Comparativ
e Study

Kroes, M. C. W., Strange, B. A., & Dolan, R. J. (2010). β
-
Adrenergic Blockade during Memory Retrieval in
Humans Evokes a Sustained Reduction of Declarative Emotional Memory Enhancement.
The
Journal of Neuroscience, 30
(11), 3959
-
3963. doi: 10.1523/j
neurosci.5469
-
09.2010

LaBar, K. S., Gatenby, J. C., Gore, J. C., LeDoux, J. E., & Phelps, E. A. (1998). Human Amygdala Activation
during Conditioned Fear Acquisition and Extinction: a Mixed
-
Trial fMRI Study.
Neuron, 20
(5),
937
-
945. doi: 10.1016/s0896
-
6273(
00)80475
-
4

Lang, S., Kroll, A., Lipinski, S. J., Wessa, M., Ridder, S., Christmann, C., . . . Flor, H. (2009). Context
conditioning and extinction in humans: differential contribution of the hippocampus, amygdala
and prefrontal cortex. [Research Support, N
on
-
U S Gov't].
Eur J Neurosci, 29
(4), 823
-
832.

38


Lee, S.
-
H., Choi, J.
-
H., Lee, N., Lee, H.
-
R., Kim, J.
-
I., Yu, N.
-
K., . . . Kaang, B.
-
K. (2008). Synaptic Protein
Degradation Underlies Destabilization of Retrieved Fear Memory.
Science, 319
(5867), 1253
-
1256.
doi: 10.1126/science.1150541

McNally, G. P., Lee, B. W., Chiem, J. Y., & Choi, E. A. (2005). The midbrain periaqueductal gray and fear
extinction: opioid receptor subtype and roles of cyclic AMP, protein kinase A, and mitogen
-
activated protein kinase. [10.
1037/0735
-
7044.119.4.1023].
Behav Neurosci, 119
, 1023
-
1033.

McNally, G. P., Pigg, M., & Weidemann, G. (2004). Opioid Receptors in the Midbrain Periaqueductal Gray
Regulate Extinction of Pavlovian Fear Conditioning.
The Journal of Neuroscience, 24
(31), 691
2
-
6919. doi: 10.1523/jneurosci.1828
-
04.2004

Milad, M. R., & Quirk, G. J. (2002). Neurons in medial prefrontal cortex signal memory for fear
extinction. [10.1038/nature01138].
Nature, 420
(6911), 70
-
74.

Milad, M. R., & Quirk, G. J. (2012). Fear Extinction a
s a Model for Translational Neuroscience: Ten Years
of Progress.
Annual Review of Psychology, 63
(1), 129
-
151. doi:
doi:10.1146/annurev.psych.121208.131631

Milad, M. R., Quirk, G. J., Pitman, R. K., Orr, S. P., Fischl, B., & Rauch, S. L. (2007). A Role for
the Human
Dorsal Anterior Cingulate Cortex in Fear Expression.
Biological Psychiatry, 62
(10), 1191
-
1194.
doi: 10.1016/j.biopsych.2007.04.032

Milad, M. R., Rauch, S. L., Pitman, R. K., & Quirk, G. J. (2006). Fear extinction in rats: Implications for
human b
rain imaging and anxiety disorders.
Biological Psychology, 73
(1), 61
-
71. doi:
10.1016/j.biopsycho.2006.01.008

Milad, M. R., Wright, C. I., Orr, S. P., Pitman, R. K., Quirk, G. J., & Rauch, S. L. (2007). Recall of Fear
Extinction in Humans Activates the Ven
tromedial Prefrontal Cortex and Hippocampus in
Concert.
Biological Psychiatry, 62
(5), 446
-
454. doi: 10.1016/j.biopsych.2006.10.011

Mohammed, A. K., Jonsson, G., & Archer, T. (1986). Selective lesioning of forebrain noradrenaline
neurons at birth abolishes
the improved maze learning performance induced by rearing in
complex environment.
Brain Research, 398
(1), 6
-
10. doi: 10.1016/0006
-
8993(86)91243
-
6

Mueller, D., Porter, J. T., & Quirk, G. J. (2008). Noradrenergic Signaling in Infralimbic Cortex Increases
Cel
l Excitability and Strengthens Memory for Fear Extinction.
The Journal of Neuroscience, 28
(2),
369
-
375. doi: 10.1523/jneurosci.3248
-
07.2008

Myers, K. M., & Davis, M. (2006). Mechanisms of fear extinction.
Mol Psychiatry, 12
(2), 120
-
150.

Nader, K., & Hardt
, O. (2009). A single standard for memory: the case for reconsolidation.
[10.1038/nrn2590].
Nat Rev Neurosci, 10
(3), 224
-
234. doi:
http://www.nature.com/nrn/journal/v10/n3/su
ppinfo/nrn2590_S1.html

Nader, K., Schafe, G. E., & Le Doux, J. E. (2000). Fear memories require protein synthesis in the amygdala
for reconsolidation after retrieval. [10.1038/35021052].
Nature, 406
(6797), 722
-
726.

O’Donnell, T., Hegadoren, K. M., & Coupl
and, N. C. (2004). Noradrenergic Mechanisms in the
Pathophysiology of Post
-
Traumatic Stress Disorder.
Neuropsychobiology, 50
(4), 273
-
283.

Oliveira, F. T. P., McDonald, J. J., & Goodman, D. (2007). Performance monitoring in the anterior
cingulate is not al
l error related: Expectancy deviation and the representation of action
-
outcome
associations.
J. Cognitive Neuroscience, 19
(12), 1994
-
2004. doi: 10.1162/jocn.2007.19.12.1994

Ouyang, M., & Thomas, S. A. (2005). A requirement for memory retrieval during and a
fter long
-
term
extinction learning.
Proceedings of the National Academy of Sciences of the United States of
America, 102
(26), 9347
-
9352. doi: 10.1073/pnas.0502315102

Ouyang, M., Young, M. B., Lestini, M. M., Schutsky, K., &

Thomas, S. A. (2012). Redundant
Catecholamine Signaling Consolidates Fear Memory via Phospholipase C.
The Journal of
Neuroscience, 32
(6), 1932
-
1941. doi: 10.1523/jneurosci.5231
-
11.2012

39


Pedreira, M. E., Pérez
-
Cuesta, L. M., & Maldonado, H. (2004). Mismatch

Between What Is Expected and
What Actually Occurs Triggers Memory Reconsolidation or Extinction.
Learning & Memory,
11
(5), 579
-
585. doi: 10.1101/lm.76904

Phelps, E. A., Delgado, M. R., Nearing, K. I., & LeDoux, J. E. (2004). Extinction Learning in Humans:

Role of
the Amygdala and vmPFC.
Neuron, 43
(6), 897
-
905.

Phelps, E. A., O'Connor, K. J., Gatenby, J. C., Gore, J. C., Grillon, C., & Davis, M. (2001). Activation of the
left amygdala to a cognitive representation of fear. [10.1038/86110].
Nat Neurosci, 4
(
4), 437
-
441.

Pitman, R. K. (1989). Post
-
traumatic stress disorder, hormones, and memory.
Biological Psychiatry,
26
(3), 221
-
223. doi: 10.1016/0006
-
3223(89)90033
-
4

Pitman, R. K., Sanders, K. M., Zusman, R. M., Healy, A. R., Cheema, F., Lasko, N. B., . . . O
rr, S. P. (2002).
Pilot study of secondary prevention of posttraumatic stress disorder with propranolol.
Biological
Psychiatry, 51
(2), 189
-
192. doi: 10.1016/s0006
-
3223(01)01279
-
3

Rainbow, T. C., Parsons, B., & Wolfe, B. B. (1984). Quantitative autoradiogra
phy of beta 1
-

and beta 2
-
adrenergic receptors in rat brain.
Proceedings of the National Academy of Sciences, 81
(5), 1585
-
1589.

Rescorla, R. A. (2001). Experimental extinction (pp. 119
-
154).

Rescorla, R. A., & Heth, C. D. (1975). Reinstatement of fear to
an extinguished conditioned stimulus.
[10.1037/0097
-
7403.1.1.88].
J Exp Psychol: Anim Behav Process, 1
, 88
-
96.

Reznikoff, G. A., Manaker, S., Rhodes, C. H., Winokur, A., & Rainbow, T. C. (1986). Localization and
quantification of beta

adrenergic receptors

in human brain.
Neurology, 36
(8), 1067.

Rodriguez
-
Romaguera, J., Sotres
-
Bayon, F., Mueller, D., & Quirk, G. J. (2009). Systemic Propranolol Acts
Centrally to Reduce Conditioned Fear in Rats Without Impairing Extinction.
Biological Psychiatry,
65
(10), 887
-
892. doi: 10.1016/j.biopsych.2009.01.009

Schiller, D., Monfils, M.
-
H., Raio, C. M., Johnson, D. C., LeDoux, J. E., & Phelps, E. A. (2010). Preventing
the return of fear in humans using reconsolidation update mechanisms. [10.1038/nature08637].
Nature, 463
(7277), 49
-
53. doi:
http://www.nature.com/nature/journal/v463/n7277/suppinfo/nature08637_S1.html

Sevenster, D., Beckers, T., &

Kindt, M. (2012). Retrieval per se is not sufficient to trigger reconsolidation
of human fear memory.
Neurobiology of Learning and Memory, 97
(3), 338
-
345. doi:
10.1016/j.nlm.2012.01.009

Shi, C. J., & Cassell, M. D. (1998). Cascade projections from somatos
ensory cortex to the rat basolateral
amygdala via the parietal insular cortex.
The Journal of Comparative Neurology, 399
(4), 469
-
491.
doi: 10.1002/(sici)1096
-
9861(19981005)399:4<469::aid
-
cne3>3.0.co;2
-
#

Shmuel, A., Yacoub, E., Pfeuffer, J., Van de Moortele
, P.
-
F., Adriany, G., Hu, X., & Ugurbil, K. (2002).
Sustained Negative BOLD, Blood Flow and Oxygen Consumption Response and Its Coupling to
the Positive Response in the Human Brain.
Neuron, 36
(6), 1195
-
1210. doi: 10.1016/s0896
-
6273(02)01061
-
9

Sotres
-
Bayon,

F., Bush, D. E. A., & LeDoux, J. E. (2004). Emotional Perseveration: An Update on
Prefrontal
-
Amygdala Interactions in Fear Extinction.
Learning & Memory, 11
(5), 525
-
535. doi:
10.1101/lm.79504

Stefanacci, L., Farb, C. R., Pitkänen, A., Go, G., Ledoux, J. E
., & Amaral, D. G. (1992). Projections from the
lateral nucleus to the basal nucleus of the amygdala: A light and electron microscopic PHA
-
L
study in the rat.
The Journal of Comparative Neurology, 323
(4), 586
-
601. doi:
10.1002/cne.903230411

Tollenaar, M. S
., Elzinga, B. M., Spinhoven, P., & Everaerd, W. (2009). Immediate and prolonged effects
of cortisol, but not propranolol, on memory retrieval in healthy young men.
Neurobiology of
Learning and Memory, 91
(1), 23
-
31. doi: 10.1016/j.nlm.2008.08.002

40


Tranel, D
., & Damasio, H. (1994). Neuroanatomical correlates of electrodermal skin conductance
responses.
Psychophysiology, 31
(5), 427
-
438. doi: 10.1111/j.1469
-
8986.1994.tb01046.x

Vidal
-
Gonzalez, I., Vidal
-
Gonzalez, B., Rauch, S. L., & Quirk, G. J. (2006). Microsti
mulation reveals
opposing influences of prelimbic and infralimbic cortex on the expression of conditioned fear.
Learning & Memory, 13
(6), 728
-
733. doi: 10.1101/lm.306106

Wang, S. H., & Morris, R. G. (2010). Hippocampal
-
neocortical interactions in memory fo
rmation,
consolidation, and reconsolidation.
Annual Review of Psychology, 61
, 49
-
79, C41
-
44.

Yágüez, L., Coen, S., Gregory, L. J., Amaro Jr, E., Altman, C., Brammer, M. J., . . . Aziz, Q. (2005). Brain
Response to Visceral Aversive Conditioning: A Functio
nal Magnetic Resonance Imaging Study.
Gastroenterology, 128
(7), 1819
-
1829. doi: 10.1053/j.gastro.2005.02.068

Yu, A. J., & Dayan, P. (2005). Uncertainty, Neuromodulation, and Attention.
Neuron, 46
(4), 681
-
692. doi:
10.1016/j.neuron.2005.04.026