Serotonin transporter polyadenylation polymorphism
modulates the retention of fear extinction memory
Catherine A. Hartleya,1, Morgan C. McKennab, Rabia Salmana, Andrew Holmesc, B. J. Caseyd, Elizabeth A. Phelpsa,e,
and Charles E. Glattb,1
aDepartment of Psychology andeCenter for Neural Science, New York University, New York, NY 10003;bDepartment of Psychiatry anddThe Sackler Institute
for Developmental Psychobiology, Weill Cornell Medical College, New York, NY 10065; andcLaboratory of Behavioral and Genomic Neuroscience, National
Institute on Alcohol Abuse and Alcoholism, Bethesda, MD 20852-9411
Edited by Michael Posner, University of Oregon, Eugene, OR, and approved February 21, 2012 (received for review February 3, 2012)
Growing evidence suggests serotonin’s role in anxiety and depres-
sion is mediated by its effects on learned fear associations. Pharma-
cological and genetic manipulations of serotonin signaling in mice
alter the retention of fear extinction learning, which is inversely
associated with anxious temperament in mice and humans. Here,
we test whether genetic variation in serotonin signaling in the form
of a common human serotonin transporter polyadenylation poly-
morphism (STPP/rs3813034) is associated with spontaneous fear
recovery after extinction. We show that the risk allele of this poly-
morphism is associated with impaired retention of fear extinction
memory and heightened anxiety and depressive symptoms. These
STPP associations in humans mirror the phenotypic effects of sero-
tonin transporter knockout in mice, highlighting the STPP as a
potential genetic locus underlying interindividual differences in se-
rotonin transporter function in humans. Furthermore, we show
that the serotonin transporter polyadenylation profile associated
with the STPP risk allele is altered through the chronic administra-
tion of fluoxetine, a treatment that also facilitates retention of
extinction learning. The propensity to form persistent fear associ-
ations due to poor extinction recall may be an intermediate pheno-
type mediating the effects of genetic variation in serotonergic
function on anxiety and depression. The consistency and specificity
of these data across species providerobust support for this hypoth-
esis and suggest that the little-studied STPP may be an important
risk factor for mood and anxiety disorders in humans.
An early “chemical hypothesis” that depression stems from a se-
rotonin deficiency was belied by observations that reducing se-
rotonin levels does not induce depression in healthy individuals
(1) and that inhibition of serotonin reuptake fails to yield rapid
antidepressant effects (2). These findings suggest that necessary
intervening processes mediate serotonin’s role in anxiety and
depression. The “network hypothesis” proposes that serotonin
fosters neural plasticity that supports adaptive processing of af-
fective information (3). Under this view, anxiety and depression
stem from aberrant integration of salient environmental in-
formation due to altered serotonergic function. Consistent with
fear associations as a central underlying mechanism by which
serotonin contributes to mood and anxiety disorders (4).
Extinction learning is a primary means of regulating conditioned
fear responses (5). During extinction, fear expression decreases,
reflecting new learning that a once-threatening stimulus now sig-
nals safety. The efficacy of extinction at reducing fear depends on
the ability to retrieve an extinction memory upon subsequent
encounters with a conditioned stimulus. However, failure to recall
extinction learning may occur with the passage of time, resulting in
“spontaneous recovery” of the original fear memory (6). Research
suggests that the degree of extinction retention is a relatively stable
individual trait with corresponding neurobiological substrates (7–
9). Recent reports in humans that spontaneous recovery of fear is
erotonin is implicated in the etiology of anxiety and de-
of extinction memory may underlie serotonin’s effects on anxiety
Research in animal models provides strong evidence for this
hypothesis. Genetic knockout of the serotonin transporter (5-
HTT) increases anxiety and depression-related behavior in the
mouse(12).Furthermore, duringfearconditioning,5-HTT knock-
out mice show normal acquisition and initial extinction learning,
but exhibit a selective deficit in extinction recall (13, 14) accom-
panied by abnormal neuronal morphology in regions that support
extinction retention (9, 13). These data suggest that 5-HTT down-
regulation impairs extinction retention, resulting in persistent fear
no study has yet examined whether normal variation in seroto-
nergic function modulates extinction memory in humans.
The 5-HTT is expressed as two mRNA species that differ in
the use of two polyadenylation signals (15, 16). Polydenylation
is a posttranscriptional modification of the 3′ end of the tran-
script that occurs in the majority of protein-coding mRNAs.
Alternative polyadenylation forms are common in the brain and
can lead to diversity in the regulation of gene expression (17).
The two reported 5-HTT polyadenylation forms differ by a 123-
bp element, and the human gene contains a common T/G single
nucleotide polymorphism (rs3813034) in the more distal of the
two polyadenylation signals. We have termed rs3813034 the se-
rotonin transporter polyadenylation polymorphism (STPP) be-
cause it alters the use of the polyadenylation signal in which it
occurs, influencing the balance of the two polyadenylation forms
in the brain (15, 16). G-allele carriers have a reduced fraction of
5-HTT mRNA containing the distal polyadenylation sequence
(distal polyadenylation fraction; DPF), and also exhibit increased
risk for panic disorder (15). Of functional interest, the distal
polyadenylation sequence element is positively correlated with
the steady-state level of 5-HTT mRNA (15) and occurs adjacent
to a microRNA (miR-16) binding site (18), suggesting that
binding of regulatory proteins to the distal sequence element
may modulate miR-16 binding (19) and alter 5-HTT translation
(18), as has been described in miR-16–mediated regulation of
cyclooxygenase 2 gene expression.
Here, we genotyped participants (SI Appendix, Table S1) in
a two-day fear conditioning paradigm (Fig. 1A) for the STPP to
influence extinction retention. During fear acquisition, a colored
square (the conditioned stimulus +, or CS+) was paired with
Author contributions: C.A.H., B.J.C., E.A.P., and C.E.G. designed research; C.A.H., M.C.M.,
R.S., and C.E.G. performed research; C.A.H. and C.E.G. analyzed data; A.H. contributed
previously published data in the 5-HTT knockout mouse; and C.A.H., A.H., B.J.C., E.A.P.,
and C.E.G. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
1To whom correspondence may be addressed. E-mail: email@example.com or ceg2004@med.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.
| April 3, 2012
| vol. 109
| no. 14
a mild electric shock on one-third of presentations, whereas an-
other square, conditioned stimulus − (CS-), was never paired with
shock. Differential skin conductance response (SCR) to the two
stimuli (CS+ minus CS-) served as our measure of the condi-
tioned fear response (CR). Acquisition was immediately followed
by extinction, during which the CS+ was no longer paired with
shock. A day later, a second extinction phase enabled evaluation
of whether participants exhibited spontaneous fear recovery
(defined as the increase in the CR from the final block of day one
extinction to the initial block of extinction recall). After condi-
tioning, participants completed questionnaires assessing anxiety
and depressive symptoms (20, 21). We found that the STPP
showed a dose-dependent association with the spontaneous re-
covery of fear, anxious temperament, and depressive symptoms.
To further relate the 5-HTT distal polyadenylation fraction to
anxiety-related emotional processing, we treated mice with the
selective serotonin reuptake inhibitor (SSRI) fluoxetine, a treat-
ment that facilitates extinction retention (22–24), alleviates anxi-
etyanddepressive symptoms(25), andis the most commonlyused
pharmacological treatment for panic disorder. Chronic fluoxetine
increased the 5-HTT DPF in mouse brain, suggesting that bi-
ological factors that increase the 5-HTT DPF are anxiolytic and
those that decrease it are anxiogenic.
Collectively, these data suggest that the 5-HTT DPF is asso-
ciated with extinction retention and that common genetic vari-
ation in 5-HTT polyadenylation may modulate the retention of
fear extinction memory in humans, influencing the risk of anxiety
disorders and depression.
Genetic Variation in 5-HTT Polyadenylation Selectively Modulates
Extinction Retention in Humans. Conditioned responses during late
acquisition and late extinction for participants who exhibited
CS- SCR during the final two blocks of acquisition; n = 110) are
shown in Fig. 1B. Because extinction and spontaneous recovery
cannot be assessed in those who do not show initial acquisition,
theseindividuals were excludedfrom thephysiological analyses. A
paired Student’s t test revealed a decrease in conditioned fear
from late acquisition to late extinction (t = 4.868, df = 109, P =
0.0000038). As a group, participants showed modest spontaneous
fear recovery (differential CR from the first block extinction recall
minus the final block of extinction) the following day (Fig. 1C)
(one-sample t test; t = 1.88, df = 109, P = 0.063).
To ascertain whether STPP genotype influenced fear expres-
sion, we conducted linear regressions with each participant’s
number of putative “risk” (G) alleles as an independent variable,
and fear acquisition, extinction, and spontaneous recovery
that the STPP G allele, which has been associated with a lower
5-HTT DPF and increased risk of panic disorder (15), would be
associated with impaired extinction retention. We found that the
number of G alleles showed a dose-dependent linear relationship
with the magnitude of spontaneous fear recovery (β = 0.191, P =
0.046) (Fig. 2A, see SI Appendix, Fig. S1 for scatterplot). This
association was unaffected by participants’ age, sex, and race (SI
Appendix, Tables S2 and S3). Demonstrating the selectivity of this
effect, there was no relationship between the number of G alleles
and either fear acquisition (β = −0.077, P = 0.422) or extinction
(β = −0.105, P = 0.266) (SI Appendix, Fig. S2). This STPP asso-
mice on extinction retention (13, 14) (Fig. 2B).
For comparison purposes, we also genotyped samples at the
highly studied 5-HTTLPR, which has been inconsistently asso-
ciated with 5-HTT expression (26, 27) and a variety of behavioral
phenotypes (28–32). There was no significant relationship be-
tween 5-HTTLPR genotype and fear expression during any phase
of conditioning (all P > 0.05) (SI Appendix, Fig. S3 A and B and
Table S6), and there were no STPP-5-HTTLPR genotype inter-
actions upon any of our fear conditioning measures (all P > 0.05).
Genetic Variation in 5-HTT Polyadenylation Is Associated with In-
dividual Differences in Anxiety and Depressive Symptoms. Consis-
tent with previous studies linking poor extinction retention to
heightened anxiety (10, 11), participants’ degree of spontane-
ous fear recovery was positively correlated with self-reported
state anxiety [State-Trait Anxiety Inventory (STAI)-S subscale;
mean (M) = 38.61, SD = 9.95] (r = 0.193, P = 0.043) and
showed a trending relationship with trait anxiety (STAI-T
subscale; M = 39.76, SD = 10.86) (r = 0.159, P = 0.096), in-
dependent of STPP genotype.
We then examined whether individual differences in trait
anxiety and depressive symptomatology (BDI-II; M = 8.38; SD =
7.05) were related to STPP genotype. Participants’ (n = 141)
number of STPP G alleles showed linear associations with both
trait anxiety (β = 0.320, P = 0.00011) (Fig. 2B; see SI Appendix,
Fig. S4 for scatterplot) and depressive symptoms (β = 0.293, P =
0.0004) (Fig. 2C; see SI Appendix, Fig. S5 for scatterplot), which
were unaffected by age, sex, and race (SI Appendix, Tables S4 and
S5) and remained significant after excluding subjects reporting
high levels of depressive symptoms and trait anxiety (SI Appendix,
SI Text). The number of G alleles also showed a trending re-
lationship with state anxiety (STAI-S; β = 0.166, P = 0.083).
By comparison, participants’ number of 5-HTTLPR S′ (tri-
allelic “short”, S and LG) alleles showed a correlation with de-
pressive symptoms(β = 0.170, P = 0.044)(SI Appendix, Fig. S6A),
anda trendingassociation withtrait anxiety(β = 0.121,P =0.153)
(SI Appendix, Fig. S6B) (see SI Appendix, Tables S7 and S8 for
demographic controls). There were no STPP-5-HTTLPR inter-
actions upon trait anxiety or depressive symptoms (both P > 0.05).
However, in our sample, participants’ number of STPP G and
5-HTTLPR S′ alleles were correlated (r = 0.296, P = 0.00037),
a phenomenon referred to in population genetics as linkage dis-
equilibrium. We hypothesized that the observed associations of
the 5-HTTLPR with trait anxiety and depressive symptoms might
which was more strongly predictive of these measures. Including
(A) Experimental design. (B) Summary graphs illustrating group conditioned
response (mean differential SCR to the CS+ versus the CS-) during late acqui-
sition (final two blocks) and late extinction (final two blocks). Participants
acquired a conditioned fear response that was significantly decreased via
extinction learning. (C) Conditioned responses increased (P = 0.063) from the
final block of extinction to the first block of extinction recall (spontaneous
recovery). *****P < 0.00001. Error bars ± SEM.
| www.pnas.org/cgi/doi/10.1073/pnas.1202044109Hartley et al.
participants’ allele counts for both polymorphisms in a multiple
regression to control for their correlation revealed that only
number of STPP G alleles was a significant predictor of both trait
anxiety (β = 0.311, P = 0.0003) and depressive symptoms (β =
0.266, P = 0.0021). A bootstrap mediation analysis (SI Appendix,
SI Text) confirmed that the apparent relationship between the 5-
HTTLPR and these phenotypes was mediated by the STPP
(STAI-T: z = 3.26, P = 0.0011; BDI-II: z = 3.13,P = 0.0017) (Fig.
3 A andB). Reversing the position of the STPP and5-HTTLPR in
this mediation does not yield a significant reduction in the direct
effect of the STPP (STAI-T: P = 0.71; BDI-II: P = 0.92).
Chronic Fluoxetine Alters 5-HTT Polyadenylation in the Mouse Brain.
Administration of SSRIs is a common pharmacological approach
to the treatment of mood and anxiety disorders including panic
disorder. Recent studies indicate that chronic administration of
the SSRI fluoxetine improves extinction retention (22–24), sug-
gesting a potential mechanism underlying its anxiolytic and anti-
depressant effects(25).Here, wetestwhether fluoxetinealters the
5-HTT DPF, which would further implicate the 5-HTT DPF as
a mediator of both altered extinction retention and the thera-
peutic effects of SSRIs. We found that brains of mice after 3 wk of
fluoxetine administration (n = 5) had a significantly higher DPF
may stem from changes in 5-HTT polyadenylation.
Our results suggest that the STPP selectively influences the recall
of fear extinction memory in humans. Consistent with the hy-
pothesis that poor extinction retention confers vulnerability to
anxiety and fear-related affective disorders (11, 12), the STPP
was also associated with trait anxiety and depressive symptoms.
Fig. 2 highlights the striking parallel between these STPP asso-
ciations and reported behavioral effects of genetic knockout of 5-
HTT in the mouse. In both species, extinction retrieval, but not
initial fear acquisition or extinction, is selectively impaired (13,
14) (Fig. 2D). Furthermore, the modulation of anxiety and de-
pressive symptoms by STPP genotype in humans parallels the
behavioral impairments of knockout mice in the elevated plus
maze and forced swim tests (13, 33), rodent assays of anxiety, and
depression-related behavior (Fig. 2 E and F). We note that gene
knockout may not fully recapitulate the effects of the STPP be-
cause the two are very different forms of genetic variation. None-
theless, the consistency and specificity of the phenotypes reported
here across species lend confidence to the validity of the STPP
associations and support the theory that 5-HTT contributes to
the development of anxiety and depression via modulation of
Our findings suggest that extinction retention may be an endo-
phenotype mediating the effects of genetic variability in 5-HTT
a heritable and quantifiable trait that genetically cosegregates with
disease (34). Both fear acquisition and initial extinction learning
Rodent data adapted from refs. 13 and 33. (A) Participants showed a linear increase (β = 0.191, P = 0.046) in spontaneous fear recovery as a function of
number of STPP G alleles (TT vs. GG: t = 2.109, df = 61, P = 0.039). Trait anxiety (B) and depressive symptoms (C) increased linearly (STAI-T: β = 0.320, P =
0.00011; BDI-II: β = 0.293, P = 0.0004) as a function of number of STPP G alleles [GG vs. TT: STAI-T: t(79) = 3.576, P = 0.0006; BDI-II: t(79) = 3.264, P = 0.002; GG vs.
GT: STAI-T: t(105) = 3.468, df =, P = 0.0007; BDI-II: t(105) = 2.673, P = 0.009]. (D) 5-HTT knockout mice (KO) show higher levels of freezing during extinction
recall (first block) compared with wild-type mice (WT) (E) Percentage of time spent in open arms of the elevated plus maze task was reduced in KO versus WT
mice. (F) Percentage of time spent immobile during the second day of exposure to the forced swim task was greater in KO versus WT mice.
*P < 0.05, **P < 0.01, ***P < 0.001. Error bars ± SEM.
STPP modulates extinction retention, anxiety, and depressive symptoms in humans and parallels reported effects of 5-HTT knockout in the mouse.
HTTLPR and anxiety and depressive symptoms. Bootstrap mediation analyses
indicate that the influence of 5-HTTLPR on trait anxiety (A) and depressive
symptoms (B) is mediated by the STPP. Path a shows the estimated co-
efficient of the relationship between the 5-HTTLPR and STPP. Path b shows
the coefficient for the relationship between STPP and each self-report
measure. Paths c and c′ show coefficients for the total (dashed line) and
direct (solid line) effects of 5-HTTLPR on each measure. All coefficients are
unstandardized, SEM in parentheses. **P < 0.01, ***P < 0.001 (Mediation
Effects: STAI-T: P = 0.0011; BDI-II: P = 0.0017).
STPP polymorphism mediates apparent relationship between 5-
Hartley et al. PNAS
| April 3, 2012
| vol. 109
| no. 14
has quantified the heritability of extinction retention, specific in-
terstrain behavioral differences in extinction recall (8) suggest that
genetic background also influences extinction retention. Here, we
identify a common genetic factor associated with extinction re-
the STPP is associated with risk for panic disorder (15), it is not
known whether individuals with this illness show impaired extinc-
tion recall. A more robust future validation of this endophenotype
hypothesis would require a demonstration of a genetically driven
impaired extinction retention in a clinical population.
The deficit inextinction recall associated withtheSTPPGallele
could stem from a failure to either consolidate or retrieve extinc-
tion learning. These processes depend on interaction between the
amygdala, the ventromedial prefrontal cortex (vmPFC), and the
hippocampus (9), regions that receive serotonergic innervation
Morphological abnormalities of neurons in the basolateral amyg-
dala and the infralimbic cortex of 5-HTT knockout mice implicate
these regions as potential substrates of the associated impairment
in extinction retention (13, 14). Successful retention of extinction
in humans is associated with increased cortical thickness in the
vmPFC (7, 38), a putative human homolog of the rodent infra-
limbic region thought to facilitate fear inhibition during extinction
retrieval (9). Because no studies have yet examined differences in
brain structure or function associated with the STPP, further re-
spontaneous fear recovery.
Anxiety disorders are commonly treated with extinction-based
exposure therapy (39). Extinction retention across exposure ses-
sions predicts the degree of reduction in anxiety symptoms, high-
lighting its importance for persistent fear reduction (40). Our data
suggest that individuals who may be most at risk for anxiety dis-
orders might benefit the least from this therapeutic approach,
because reduced extinction retention would limit its efficacy. Ad-
ministration of SSRIs is another common approach to treating
clinical anxiety and depression. Recent reports that chronic SSRI
administration reduces spontaneous recovery (22, 24), contextual
renewal (22), and reinstatement (23) of extinguished fear while
leaving fear acquisition intact provide further evidence for the
specificity of the effects of serotonin on extinction recall. In con-
junction with our present finding that chronic fluoxetine amelio-
rates the polyadenylation risk profile associated with reduced
extinction retention in humans (15), these data suggest that the
therapeutic efficacy of SSRIs may stem from molecular alterations
that facilitate consolidation and retrieval of extinction learning.
The mechanism by which the distal polyadenylation sequence
alters 5-HTT function is unknown. The steady state level of total
5-HTT mRNA expression is correlated with the presence of the
distal polyadenylation sequence (15), suggesting that it may
stabilize 5-HTT mRNA. Furthermore, a recent report that miR-
16 binding to a site directly adjacent to the distal polyadenylation
sequence causes translational repression of the 5-HTT suggests
that an RNA binding protein may modulate the effects of miR-
16 on 5-HTT expression, as has been reported for the cyclo-
oxygenase2 gene (19).
The 5-HTTLPR is presently the most widely studied poly-
morphism in human genetics. This common polymorphism has
been reported to alter transcription of 5-HTT (26, 28) and has
been associated with a range of phenotypes including anxious
temperament (28), frontoamygdala responses to emotional stim-
uli (41–43), and risk of depression in response to stressful life
events (29). These associations have not been consistently repli-
phenotypes present conflicting evidence for the 5-HTTLPR’s role
in behavioral variability (30–32) These inconsistent findings sug-
gest that the link between the 5-HTTLPR and anxiety-related
phenotypes may be complex, with its detection dependent upon
the methodologies used as well as factors including alternative
genetic variants such as the STPP (44, 45). In our data, we find
that the 5-HTTLPR does not phenocopy the effects of 5-HTT
knockout in mice on extinction retention and we do not replicate
an earlier report of a 5-HTTLPR association with human fear
acquisition (46). Furthermore, an apparent relationship of the
5-HTTLPR to trait anxiety and depressive symptoms is explained
strongly associated with these phenotypes. Thus, inconsistent
phenotypes might be due to variable correlation in experimental
samples between the two polymorphisms. Given the history of
nonreplication in the field of behavioral genetics and the limited
statistical power of our sample, the conclusions we draw from our
data are speculative. We encourage future studies to test this
linkage disequilibrium hypothesis by genotyping behavioral sam-
ples for both the 5-HTTLPR and the STPP, clarifying their rela-
tive phenotypic effects and attempting to resolve the conflicts
in the present literature. We strongly suggest the use of pooled
genetic samples, consistent phenotypic criteria, and a priori
hypotheses in such efforts.
A shortcoming of our study was the failure to fully control for
psychiatric history and drug use of participants. Although our sub-
jects were not presently taking psychiatric medication and reported
subclinical symptoms of depression, they were not formally char-
acterized as free of psychiatric disorders. Given the high lifetime
our sample. Furthermore, our sample might include subjects taking
nonpsychiatric medications known to influence memory consoli-
dation (e.g., beta blockers). These confounds may have influenced
our reported effects and should be controlled for in future studies.
Depression and anxiety disorders show substantial heritability
(48, 49) and high comorbidity (50), suggesting common patho-
physiological mechanisms. Based on the broad efficacy of medi-
cations targeting 5-HTT and the marked behavioral effects of
genetic knockout, the 5-HTT gene has been widely proposed as
a candidate gene conferring vulnerability to these disorders (12,
37). Our present findings, as well as a previous association with
panic disorder (15), indicate that the STPP may be an important
genetic locus of risk for anxiety and depression. Furthermore, our
data suggest that the retention of fear extinction memory may be
a critical mediator of the influence of serotonin on individual vul-
nerability to these disorders in humans.
87 female), were recruited at New York University for a 2-d fear conditioning
study. All participants gave informed consent and were paid for participation.
brain. Brain tissue of fluoxetine-treated mice had a higher fraction of 5-HTT
mRNA containing the distal polyadenylation sequence than that of control
animals. Values for fluoxetine group are normalized to controls. ****P <
Fluoxetine increases 5-HTT distal polyadenylation fraction in mouse
| www.pnas.org/cgi/doi/10.1073/pnas.1202044109 Hartley et al.
Participants were not taking any psychiatric medication. Participants were
racially diverse (63 Caucasian, 58 Asian, 3 Black/African-American, 6 Hispanic,
11 mixed race). Genotypes for both the STPP (34 TT, 60 GT, 47 GG) and the
S′S′, 57 S′L′, 28 L′L′) were in Hardy Weinberg equilibrium within Caucasian
(5-HTTLPR P = 0.7; STPP P = 0.14) and Asian (5-HTTLPR P = 0.33; STPP P = 0.28)
ethnic groups, who collectively constituted 85.8% of the sample. Subject de-
mographics by STPP genotype are shown in SI Appendix, Table S1.
Fear Conditioning. During fear conditioning, participants saw a sequence of
trials in which two colored squares were presented for 4 s each. During the
fear acquisition phase, one square (the CS+) coterminated with a mild electric
wrist shock on 33% of trials, whereas the other (the CS-) was never paired
with shock. Acquisition was followed immediately by extinction, in which
neither CS was paired with shock. The next day, a second extinction phase
enabled assessment of whether participants retained extinction learning or
exhibited spontaneous recovery of the conditioned fear response. Trials in
each phase were grouped into “blocks.” In every block, the CS+ and CS-
stimuli were presented four times each without shock. During acquisition,
there were two additional trials per block in which the CS+ was paired with
shock. Acquisition consisted of four blocks of trials (four CS+, four CS-, and
two CS+ with shock per block), and each extinction phase consisted of five
blocks of trials (four CS+ and four CS- per block) (Fig. 1A). Two randomized
trial orders were counterbalanced across subjects. The first trial of the day 2
extinction session began with a CS+ for one-half of subjects and a CS- for the
other. The skin conductance response (SCR) to each stimulus was assessed.
Participants showing a greater mean SCR to the CS+ versus CS- during the
final two blocks of acquisition were considered to have acquired a fear re-
sponse (only CS+ presentations not paired with shock were analyzed). Our
recovery measure was the increase in the CR from the final block of day 1
extinction to the initial block of extinction recall.
Shocks were delivered via a stimulator (Grass Instruments) connected to
a bar electrode attached to the right wrist. Participants selected shock levels
via a procedure in which the shock (200-ms duration) was gradually increased
until it was deemed “uncomfortable, but not painful” (maximum shock level
was 60 V). There were no genotype differences in shock level (SI Appendix,
Table S1). SCR was recorded through shielded Ag-AgCl electrodes attached
to the second and third fingers of the left hand by using a Biopac Systems
module. AcqKnowledge software (Biopac Systems) was used to conduct
offline analysis of the SCR waveforms. SCR data were low-pass filtered and
smoothed. The greatest base to peak change in SCR in a 0.5- to 4.5-s window
after each CS onset was assessed. These values were then square-root
transformed to normalize the distribution and divided by the mean SCR to
the shock to enable between-subject comparisons. Only unreinforced CS+
presentations were included in the analysis.
Eight participants were excluded from the physiological analysis because
of experimental error in physiological recording or a failure to display
a variable skin conductance signal. Another 23 participants were excluded
because of failure to exhibit a discriminative CR (mean CS+ SCR > mean CS-
SCR) during late acquisition. The remaining 110 participants (STPP: 28 TT, 47
GT, 35 GG; 5-HTTLPR: 22 L′L′, 50 S′L′, 38 S′S′) were included in the physio-
logical analyses. χ2tests showed no significant differences between the STPP
and 5-HTTLPR genotypes of those who did not acquire a CR and the geno-
type frequencies of the complete sample (all P > 0.05).
Questionnaires. After conditioning, participants completed the STAI (20), the
Beck Depression Inventory II (BDI-II) (21), and the Life Experience Survey
(LES) (51), assessing recent life experiences. There were no significant dif-
ferences in STAI-T or BDI-II scores as a function of sex or age (all P > 0.05) or
between Caucasian and Asian participants (both P > 0.05). There was no
significant interaction of either genotype and negative life events (LES
scores) upon anxiety and depressive symptoms, or any fear conditioning
measures (all P > 0.05).
Genotyping. DNA was prepared from saliva samples by using standard proce-
dures (Oragene;DNA Genotek). The STPP (rs3813034) was genotypedby using
The 5-HTTLPR was genotyped by using a protocol that allows assignment of
both the biallelic and triallelic genotype categorizations (52). Samples whose
initial genotype was not clear were regenotyped in duplicate, and all such
samples provided clear concordant genotypes on rerun.
Statistical Analysis. All tests were two-tailed by using an α level of 0.05. Using
SPSS 18.0 (IBM), we conducted simple linear and multiple regressions to
assess the relationships between our fear conditioning measures, question-
naire scores, genotypes, and potential covariates. Independent samples t
tests were used in pairwise genotype comparisons of conditioning measures
and questionnaire scores, and comparison of DPF in control and fluoxetine
treated brains. We conducted mediation analyses and bootstrap significance
testing by using a custom Matlab toolbox (53, 54) (SI Appendix, SI Text).
Rodent Fluoxetine Distal Polyadenylation Analysis. C57BL/6J mice were given
160 mg/L fluoxetine (n = 5) in their drinking water for 21 d, an approximate
dose of 18 mg/kg per day under ad libitum access, which yields therapeutic
libitum access to tap water. Whole brain tissue was harvested and homoge-
nized in Tri reagent (Molecular Research Center). Total RNA was isolated
according to the manufacturer’s standard protocol. Total RNA (2 μg) was
treated with DNase I and heat-inactivated then reversed transcribed by using
oligo dT primers and Moloney Murine Leukemia Virus reverse transcriptase
(200 units). Total 5-HTT and distal sequence containing 5-HTT messages were
quantified by using SYBR green-based quantitative PCR assays as described
UTR were as follows: Forward: 5′-CCAAGCTGATGATGTAAGGTCTTT-3′, Re-
verse: 5′-GTCACCAGCTAATGTGGCAGTAA-3′. Standard curves for each assay
were prepared by using serial dilutions of pooled cDNA samples to allow
relative quantification. Distal polyadenylation fraction was calculated by di-
viding the relative amount of distal sequence containing 5-HTT messages by
the relative amount of total 5-HTT message determined for each sample. All
samples were reverse transcribed twice, and each cDNA sample was run in
5-HTT Knockout Mouse Data. Behavioral data from two published studies (13,
33) in the 5-HTT knockout mouse are included here to show cross-species
phenotypic correspondence in fear conditioning (13) and anxiety and de-
pression-like behavior [elevated plus maze (33) and forced swim (13) tasks, in
the mouse]. Details of the experimental protocols and analyses appear in the
methods sections of these manuscripts.
ACKNOWLEDGMENTS. We thank A. Gorun and A. Richman for assistance
with data collection and analysis. A. Izquierdo and A. Holmes collected the
original 5-HTT knockout mouse data. This research was supported by the
National Science Foundation (C.A.H.); the James S. McDonnell Foundation,
and National Institutes of Health (NIH) Grants MH072279 and MH80758 (to
E.A.P.); The National Institute on Alcohol Abuse and Alcoholism Intramural
Research Program (A.H.); a generous gift by the Mortimer D. Sackler, MD
family (to B.J.C.); NIH Grant MH079513 (to B.J.C. and C.E.G.); and The
Hartwell Foundation, the Pritzker Neuropsychiatric Disorders Research
Consortium, and NIH Grant MH091401 (to C.E.G.).
1. Booij L, Van der Does AJ, Riedel WJ (2003) Monoamine depletion in psychiatric and
healthy populations: Review. Mol Psychiatry 8:951–973.
2. Nestler EJ (1998) Antidepressant treatments in the 21st century. Biol Psychiatry 44:
3. Castrén E (2005) Is mood chemistry? Nat Rev Neurosci 6:241–246.
4. Deakin JFW, Graeff FG (1991) 5-HT and mechanisms of defence. J Psychopharmacol 5:
5. Hartley CA, Phelps EA (2010) Changing fear: The neurocircuitry of emotion regula-
tion. Neuropsychopharmacology 35:136–146.
7. Milad MR, et al. (2005) Thickness of ventromedial prefrontal cortex in humans is
correlated with extinction memory. Proc Natl Acad Sci USA 102:10706–10711.
8. Hefner K, et al. (2008) Impaired fear extinction learning and cortico-amygdala circuit
abnormalities in a common genetic mouse strain. J Neurosci 28:8074–8085.
9. Quirk GJ, Mueller D (2008) Neural mechanisms of extinction learning and retrieval.
10. Rauch SL, et al. (2005) Orbitofrontal thickness, retention of fear extinction, and ex-
traversion. Neuroreport 16:1909–1912.
11. Milad MR, et al. (2009) Neurobiological basis of failure to recall extinction memory in
posttraumatic stress disorder. Biol Psychiatry 66:1075–1082.
12. Murphy DL, et al. (2008) How the serotonin story is being rewritten by new gene-
based discoveries principally related to SLC6A4, the serotonin transporter gene,
which functions to influence all cellular serotonin systems. Neuropharmacology 55:
13. Wellman CL, et al. (2007) Impaired stress-coping and fear extinction and abnormal cor-
ticolimbic morphology in serotonin transporter knock-out mice. J Neurosci 27:684–691.
14. Narayanan V, et al. (2011) Social defeat: Impact on fear extinction and amygdala-
prefrontal cortical theta synchrony in 5-HTT deficient mice. PLoS ONE 6:e22600.
Hartley et al. PNAS
| April 3, 2012
| vol. 109
| no. 14
15. Gyawali S, et al. (2010) Association of a polyadenylation polymorphism in the sero-
tonin transporter and panic disorder. Biol Psychiatry 67:331–338.
16. Battersby S, et al. (1999) Presence of multiple functional polyadenylation signals and
a single nucleotide polymorphism in the 3′ untranslated region of the human sero-
tonin transporter gene. J Neurochem 72:1384–1388.
17. Di Giammartino DC, Nishida K, Manley JL (2011) Mechanisms and consequences of
alternative polyadenylation. Mol Cell 43:853–866.
18. Baudry A, Mouillet-Richard S, Schneider B, Launay JM, Kellermann O (2010) miR-16
targets the serotonin transporter: A new facet for adaptive responses to anti-
depressants. Science 329:1537–1541.
19. Shanmugam N, Reddy MA, Natarajan R (2008) Distinct roles of heterogeneous nuclear
ribonuclear protein K and microRNA-16 in cyclooxygenase-2 RNA stability induced by
S100b, a ligand of the receptor for advanced glycation end products. J Biol Chem 283:
20. Spielberger CD, Gorsuch RL, Lushene PR, Vagg PR, Jacobs AG (1983) Manual for the
State-Trait Anxiety Inventory (Form Y) (Consulting Psychologists, Palo Alto, CA).
21. Beck AT, Steer RA, Ball R, Ranieri W (1996) Comparison of Beck Depression Inventories
-IA and -II in psychiatric outpatients. J Pers Assess 67:588–597.
22. Karpova NN, et al. (2011) Fear erasure in mice requires synergy between antide-
pressant drugs and extinction training. Science 334:1731–1734.
23. Deschaux O, Spennato G, Moreau JL, Garcia R (2011) Chronic treatment with fluox-
etine prevents the return of extinguished auditory-cued conditioned fear. Psycho-
pharmacology (Berl) 215:231–237.
24. Camp M, et al. (February 15, 2012) Genetic strain differences in learned fear inhibition
associated with variation in neuroendocrine, autonomic, and amygdala dendritic
phenotypes. Neurophyschopharmocology, 10.1038/npp.2011.340.
25. Dulawa SC, Holick KA, Gundersen B, Hen R (2004) Effects of chronic fluoxetine in
animal models of anxiety and depression. Neuropsychopharmacology 29:1321–1330.
26. Heils A, et al. (1996) Allelic variation of human serotonin transporter gene expression.
J Neurochem 66:2621–2624.
27. Parsey RV, et al. (2006) Effect of a triallelic functional polymorphism of the serotonin-
transporter-linked promoter region on expression of serotonin transporter in the
human brain. Am J Psychiatry 163:48–51.
28. Lesch KP, et al. (1996) Association of anxiety-related traits with a polymorphism in the
serotonin transporter gene regulatory region. Science 274:1527–1531.
29. Caspi A, et al. (2003) Influence of life stress on depression: Moderation by a poly-
morphism in the 5-HTT gene. Science 301:386–389.
30. Willis-Owen SA, et al. (2005) The serotonin transporter length polymorphism, neu-
roticism, and depression: A comprehensive assessment of association. Biol Psychiatry
31. Risch N, et al. (2009) Interaction between the serotonin transporter gene (5-HTTLPR),
stressful life events, and risk of depression: A meta-analysis. JAMA 301:2462–2471.
32. Karg K, Burmeister M, Shedden K, Sen S (2011) The serotonin transporter promoter
variant (5-HTTLPR), stress, and depression meta-analysis revisited: Evidence of genetic
moderation. Arch Gen Psychiatry 68:444–454.
33. Holmes A, Yang RJ, Lesch KP, Crawley JN, Murphy DL (2003) Mice lacking the sero-
tonin transporter exhibit 5-HT(1A) receptor-mediated abnormalities in tests for
anxiety-like behavior. Neuropsychopharmacology 28:2077–2088.
34. Flint J, Munafò MR (2007) The endophenotype concept in psychiatric genetics. Psychol
35. Hettema JM, Annas P, Neale MC, Kendler KS, Fredrikson MA (2003) A twin study of
the genetics of fear conditioning. Arch Gen Psychiatry 60:702–708.
36. Azmitia EC, Gannon PJ (1986) The primate serotonergic system: A review of human
and animal studies and a report on Macaca fascicularis. Adv Neurol 43:407–468.
37. Holmes A (2008) Genetic variation in cortico-amygdala serotonin function and risk for
stress-related disease. Neurosci Biobehav Rev 32:1293–1314.
38. Hartley CA, Fischl B, Phelps EA (2011) Brain structure correlates of individual differ-
ences in the acquisition and inhibition of conditioned fear. Cereb Cortex 21:
39. Foa EB (2006) Psychosocial therapy for posttraumatic stress disorder. J Clin Psychiatry
40. Berry AC, Rosenfield D, Smits JAJ (2009) Extinction retention predicts social anxiety
symptom reduction. Depress Anxiety 26:22–27.
41. Munafò MR, Brown SM, Hariri AR (2008) Serotonin transporter (5-HTTLPR) genotype
and amygdala activation: A meta-analysis. Biol Psychiatry 63:852–857.
42. Pezawas L, et al. (2005) 5-HTTLPR polymorphism impacts human cingulate-amygdala
interactions: A genetic susceptibility mechanism for depression. Nat Neurosci 8:
43. Heinz A, et al. (2005) Amygdala-prefrontal coupling depends on a genetic variation
of the serotonin transporter. Nat Neurosci 8:20–21.
44. Caspi A, Hariri AR, Holmes A, Uher R, Moffitt TE (2010) Genetic sensitivity to the
environment: The case of the serotonin transporter gene and its implications for
studying complex diseases and traits. Am J Psychiatry 167:509–527.
45. Rutter M, Thapar A, Pickles A (2009) Gene-environment interactions: Biologically valid
pathway or artifact? Arch Gen Psychiatry 66:1287–1289.
46. Lonsdorf TB, et al. (2009) Genetic gating of human fear learning and extinction:
Possible implications for gene-environment interaction in anxiety disorder. Psychol Sci
47. Kessler RC, et al. (2005) Lifetime prevalence and age-of-onset distributions of DSM-IV
disorders in the National Comorbidity Survey Replication. Arch Gen Psychiatry 62:
48. Kendler KS, Gatz M, Gardner CO, Pedersen NL (2006) A Swedish national twin study
of lifetime major depression. Am J Psychiatry 163:109–114.
49. Hettema JM, Neale MC, Kendler KS (2001) A review and meta-analysis of the genetic
epidemiology of anxiety disorders. Am J Psychiatry 158:1568–1578.
50. Merikangas KR, et al.; Zurich Cohort Study (2003) Longitudinal trajectories of de-
pression and anxiety in a prospective community study: The Zurich Cohort Study. Arch
Gen Psychiatry 60:993–1000.
51. Sarason IG, Johnson JH, Siegel JM (1978) Assessing the impact of life changes: De-
velopment of the Life Experiences Survey. J Consult Clin Psychol 46:932–946.
52. Wendland JR, Martin BJ, Kruse MR, Lesch KP, Murphy DL (2006) Simultaneous gen-
otyping of four functional loci of human SLC6A4, with a reappraisal of 5-HTTLPR and
rs25531. Mol Psychiatry 11:224–226.
53. Atlas LY, Bolger N, Lindquist MA, Wager TD (2010) Brain mediators of predictive cue
effects on perceived pain. J Neurosci 30:12964–12977.
54. Davidson ML, Atlas LY, Lindquist MA, Bolger N, Wager TD (2008) The M3 Toolbox:
The Multi-level Mediation/Moderation Framework for Connectivity Analyses in fMRI
Data (Cognitive Neuroscience Society, Davis, CA).
| www.pnas.org/cgi/doi/10.1073/pnas.1202044109Hartley et al.