Antidepressants and the resilience to early-life stress ininbred mouse strainsElke Binderb*, Karim Malkia*, Jose L. Paya-Canoa, Cathy Fernandesa,Katherine J. Aitchisona, Aleksander A. Mathec, Frans Sluytera

and Leonard C. Schalkwyka

Rationale Selecting an effective treatment for patients

with major depressive disorder is a perpetual problem

for psychiatrists. It is of particular interest to explore the

interaction between genetic predisposition and

environmental factors.

Objectives Mouse inbred strains vary in baseline

performance in depression-related behaviour tests,

which were originally validated as tests of antidepressant

response. Therefore, we investigated interactions between

environmental stress, genotype, and drug response in a

multifactorial behaviour study.

Method Our study design included four inbred mouse

strains (129S1/SvlmJ, C57LB/6J, DBA/2J and FVB/NJ)

of both sexes, two subjected to environmental

manipulations (maternal separation and unpredictable

chronic mild stress) and two representative of treatment

with antidepressants (escitalopram and nortryptiline vs.

vehicle). The mice treated with antidepressants were

further divided into those administered acute (1 day)

and subchronic (14 days) regimes, giving 144 experimental

groups in all, each with at least seven animals. All animals

were tested using the Porsolt forced-swim test (FST) and

the hole-board test.

Results Despite a 24-h maternal separation (MS) or a

14-day unpredictable chronic mild stress protocol, most

animals seemed to be resilient to the stress induced.

One compelling finding is the long-lasting, strain-specific

effect of MS resulting in an increased depression-like

behaviour in the Porsolt FST and elevated anxiety-related

behaviour in the hole-board test seen in 129S1/SvImJ

mice. Nortriptyline was effective in reversing the effect

of MS in the FST in 129S1/SvlmJ male mice.

Conclusion A single 24-h maternal separation of pups

from their mother on postnatal day 9 is a sufficient insult

to result in a depression-like phenotype in adult

129S1/SvImJ mice but not in C57LB/6 J, DBA/2 J,

and FVB/NJ mice. Pharmacogenetics and Genomics

21:779–789 �c 2011 Wolters Kluwer Health | Lippincott

Williams & Wilkins.

Pharmacogenetics and Genomics 2011, 21:779–789

Keywords: antidepressants, depression, escitalopram, Genome-basedTherapeutic Drugs for Depression, maternal deprivation, maternalseparation, nortriptyline, pharmacogenetics

aKing’s College London, Social, Genetic and Developmental Psychiatry Centre,Institute of Psychiatry, London, UK, bINSERM U862, Avenir group Physio-pathology of Energy Balance and Obesity, Universite de Bordeaux 2, Bordeaux,France and cKarolinska Institutet-Clinical Neuroscience, Karolinska University,Hospital Huddinge, Stockholm, Sweden

Correspondence to Leonard C. Schalkwyk, PhD, Social, Genetic andDevelopmental Psychiatry Centre, Institute of Psychiatry, Box P082De Crespigny Park, London SE5 8AF, UKTel: + 44 0 207 848 0279; fax: + 44 0 207 848 0866;e-mail: Leonard.schalkwyk@kcl.ac.uk

*Elke Binder and Karim Malki have contributed equally to this study.

Received 10 June 2011 Accepted 23 July 2011

IntroductionThe most popularly prescribed class of antidepressant

drugs is the selective serotonin reuptake inhibitor

(SSRI), represented in this study by the drug escitalo-

pram. Individual response to SSRIs is highly hetero-

geneous, with up to half of the treated patients finding no

relief or only partial relief from their symptoms [1].

Moreover, many patients have reported low tolerabil-

ity [2]. Therefore, a variety of alternative antidepressant

drugs can be prescribed in the hope of increasing efficacy

or tolerance. These include second-generation tricyclics,

such as nortriptyline (a noradrenaline reuptake inhibitor).

The genetic and environmental etiologies of individual

differences moderating the response and tolerance to

common antidepressant medications remain largely un-

known. It is, therefore, important to investigate the

behavioural differences in drug response in connection

with environmental factors and genetic background. This

will lead to better prediction of drug response and a

reduction in the lag time produced by the current trial-

and-error procedure.

Various attempts have been made to generate animal

models of depression [3]. One strategy is to use selectively

bred lines of rats or mice [4,5] and another is to use proto-

cols such as unpredictable chronic mild stress (UCMS)

and maternal separation (MS) to induce depression-

related behaviour in healthy animals.

Supplemental digital content is available for this article. Direct URL citationsappear in the printed text and are provided in the HTML and PDF versions of thisarticle on the journal’s website (www.pharmacogeneticsandgenomics.com).

Original article 779

1744-6872 �c 2011 Wolters Kluwer Health | Lippincott Williams & Wilkins DOI: 10.1097/FPC.0b013e32834b3f35

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

UCMS has been reported to lead to reductions in

reactivity to rewards and to a variety of other depres-

sion-like behaviours in rats and mice [6,7]. In humans, it

is known that the increase in stress in the working

environment correlates with the rising number of patients

who are suffering from depression [8,9].

The dependency of pups on their mother during early

postnatal stages is well documented [10,11], and MS can

result in a variety of long-lasting changes on adult

phenotypes [12–22]. Previous studies have suggested

that a single 24-h separation from the mother during the

first 2 postnatal weeks activates the hypothalamic-

pituitary-adrenal axis system of the neonate and results

in long-lasting consequences, such as stress and anxiety in

adulthood [23,24], whereas mild neonatal stress has been

associated with ‘positive’ effects of resilience [25].

Response to different MS protocols, however, is hetero-

geneous, as recently reported by Millstein and Holmes [26].

The study was unable to produce long-lasting behavioural

changes using either MS or handling. Reasons for this

have been attributed to methodological differences and

also to genetic variation between strains of animals

used [27,28].

This study was designed to investigate strain-specific

differences towards depression-related behaviour and anti-

depressant action using two different models: UCMS

and an MS protocol of one separation period lasting

for 24 h on postnatal day 9 [13]. The hypothesis driving our

approach is that different genetic backgrounds will play a

role in the modulation of environmental and pharmacolo-

gical treatment response in a mouse model of depression.

Materials and methodsAnimals

A total of 1166 mice from four inbred strains were used

for this study. The strains used were C57BL/6J (C57),

DBA/2J (DBA), 129S1/SvImJ (129) and FVB/NJ (FVB).

The animals were specific pathogen free, bred in the

barrier unit at the Institute of Psychiatry (London, UK).

Weaning took place when the animals were 21–28 days

old. They were group housed with same-sex siblings until

they were 8 weeks old. They were then singly housed and

moved to the testing facility. Mice were allowed to

habituate to the individual plastic cages (30.5� 13� 11

cm) and to the testing facility for 2 weeks before testing.

A 12-h light/dark cycle was used with lights on at 08.00 hr

and lights off at 20.00 hr Temperature and humidity

were controlled at 211C ± 21C and 45%, respectively.

Water and food were freely available throughout the ex-

periment. All housing and experimental procedures were

carried out in accordance with the UK Home Office

Animals (Scientific Procedures) Act 1986.

Experimental design

This study adopted a five-way multifactorial design

(Fig. 1). The factors were as follows:

(1) Strain with four levels: C57, DBA, 129, and FVB;

(2) Environment with three levels: MS (described

below), UCMS (described below), and control;

(3) Sex with two levels: male and female;

(4) Drug treatments with three levels: nortriptyline

(4 mg/kg, intraperitoneally), escitalopram (5 mg/kg,

intraperitoneally), or saline (vehicle); and

(5) Treatment duration of drug administration with two

levels: subchronic (14 daily doses) and acute

(single dose).

This resulted in 144 experimental cells with a group size

of seven to eight animals. Group assignment at weaning

was pseudorandom, except for mice undergoing the MS

protocol as entire litters underwent MS; therefore, the

allocation of dams to MS or control was randomized.

Behavioural measurements consisted of the hole-board

test (HBT) used as an arena for general activity, anxiety

and exploration and the Porsolt forced-swim test (FST)

for measuring depression-related behaviour.

Unpredictable chronic mild stress

Starting at the age of 10 weeks, the animals were exposed

to a stressor each day in pseudorandom order. The

stressors in the UCMS regime were: 2 h of home cage

tilting at 451, damp bedding for 4 h, cage switching for

Fig. 1

C57BL/6J

Male

cms

Nortry-ptiline

Chronic Acute

Female Male Female Male Female Male Female

4

8

24

72

144

DBA/2J 129SvemJ FVB/NH

s-cit

msControl

Control

Experimental design. A full factorial design with five factors yielding 144 experimental groups, each with seven or eight animals.

780 Pharmacogenetics and Genomics 2011, Vol 21 No 12

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

2 h, 10 min of flooded cage, reversed light cycle and

airpuff. Each day consisted of one to two stressors applied

at different times of the day (Table 1). The UCMS

regime lasted for a period of 14 days and was concurrent

with the intraperitoneal injections for the animals

assigned to the subchronic treatment groups.

Maternal separation

Litters of each strain were randomly allocated to the MS

group. On postnatal day 9, the dam was removed from the

litter for 24 h. The litter was kept on a heating pad in

their home cage at 331C ± 21C in a different room from

the dam to avoid contact through vocalization. Litters

were always separated and reunited with the mother

during the first half of the light phase. The first hour after

reuniting, the litter with the mother was videotaped.

Litters were of different sizes and, when possible, each

litter came from a different breeding pair (see Table 2,

supplemental digital content 1, http://links.lww.com/FPC/A319and Table 3, supplemental digital content 2, http://links.lww.com/FPC/A320). When the pups grew up, the

littermates were separated into those that would undergo

acute and those that would undergo subchronic treatment

of the antidepressant.

Drugs

Escitalopram and nortriptyline were provided by Lund-

beck (Copenhagen, Denmark). The drugs were dissolved

in saline (Aquapharm, Dunnington, York, UK) every day

fresh before use. The injection volume was 10 ml/kg of

bodyweight; 4 mg/kg of escitalopram, 5 mg/kg of nortrip-

tyline, or the corresponding volume of saline was given.

The doses of the drugs were chosen following data

analysis collected from a pilot study (data not shown).

After 2 weeks of habituation, the animals received one

daily intraperitoneal injection of nortriptyline, escitalo-

pram, or saline for 14 days for the subchronic treatment.

Animals were injected and tested during the light phase.

Blood samples were analysed to investigate the drug

levels and drug metabolism of one animal per strain

per sex per drug (Fig. 8). For the acute treatment, animals

were handled every day (except for weekends) for 14 days

without receiving any drugs. This was done to accustom

the animals to human contact in order to minimize the

stress effect of the acute injection before testing. On the

day of testing, mice received one acute intraperitoneal

injection 30 min before the HBT.

Behavioural test procedure

At the age of 12 weeks, the animals were tested using the

HBT for 5 min, directly followed by the FST for 6 min.

Hole-board test

Mice were tested in a square, evenly lit arena (25.4� 25.4

� 40.6 cm, 300 lux illumination) with clear plastic walls. A

metal floor was used, containing 16 holes (2.2 cm in

diameter), evenly distributed over the floor (4� 4 holes).

The setup was equipped with infrared photocell sensors.

Animals were placed in a corner of the field and allowed

to freely explore for 5 min. The distance travelled (gene-

ral activity), the number of holes visited (exploration)

and the time spent in the centre (anxiety, 17.8� 17.8 cm,

mouse arena E63–10) were monitored by infrared sensor

rings connected to a computer equipped with TruScan

Software Version 2.0 (Coulbourn Instruments, Allentown,

Pennsylvania, USA). Analysis was conducted using R

(Release 2.9–2; R Development Core Team, Vienna,

Austria) and Statistica version 6.1 (StatSoft, Inc., Tulsa,

Oklahoma, USA).

The Porsolt forced-swim test

Directly after the HBT, animals were placed in a clear

Plexiglas cylinder (49-cm high� 15-cm diameter) filled

with water (40-cm high, 211C). The test was conducted

in a similar manner as described elsewhere [29]. Animals

were observed for 6 min and their behaviour was hand-

coded using Ratontime 1.0 (University of Valencia,

Faculty of Psychology). The behaviour during the last

4 min was used for analysis. The behaviour was classified

Table 1 Unpredictable chronic mild stress protocol

Monday Tuesday Wednesday Thursday Friday Saturday Sunday

First weekMorning 11.00 a.m.–

13.00 p.m.,cage tilted at 451(in rack)

10.00 a.m.–14.00p.m.,damp bedding(add 200 ml ofcold water)

13.00 p.m.–15.00 p.m.,cage tilted at451 (in rack)

14.00 p.m.–14.10 p.m.,forced bath in 351 water(approximately 1 cm of waterin an empty cage)

12.00 p.m.–14.00 p.m.,empty cage

Reverselight

Reverselight

Afternoon 15.00 p.m.–16.30 p.m.,empty cages

14.00 p.m.,airpuff (three times/animal)move animals to testing rooms

Second weekMorning 11.00 a.m.–

15.00 p.m.,damp bedding(add 200 ml ofcold water)

10.00 a.m.–12.00p.m.,cage switching(back in originalhomecage)

10.00 a.m.–12.00 p.m.,empty cage

14.00 p.m. –15.00 p.m.,cage tilted 451 (in rack)

14.00 p.m.–14.10 p.m.,forced bath in 351 water (approximately1 cm of water in an empty cage), moveanimals to testing rooms

Reverselight

Reverselight

Afternoon 12.00 p.m.,airpuff (threetimes)

Antidepressants and early life stress Binder et al. 781

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

as climbing (characterized by vertical motion of the front

paws, directed at the walls), swimming (characterized by

horizontal movement), or immobile (defined as the

absence of all movement except that necessary to

maintain balance). The parameter of mobility was the

sum of climbing and swimming.

ResultsFigure 2 presents an overview of the data. Each row of

panels presents a behaviour test, and for each test there is

a characteristic strain pattern; for example, FVB shows

approximately 100% results in swimming behaviour in the

FST and DBA also ranked high (top row), whereas C57

and FVB are the most active in the hole board (middle

row) and DBA spends strikingly less time at the centre

compared with other strains (lower row). Similarly, each

column represents a drug treatment and these clearly

modify the behaviour profiles. These effects are explored

in more detail in an analysis of variance (ANOVA)

treating each factor as a fixed-effect categorical variable

and each behavioural measure as a continuous variable,

without transformation.

Fig. 2

100

Saline Nortriptyline Escitalopram

80

60

Mea

n ps

wim

[4m

in]

40

20

0

2500

2000

1500

Mea

n di

stan

ce

1000

500

0

150

120

90

Mea

n ce

ntre

tim

e

60

30

0

150

120

90

60

30

0

150

120

90

60

30

2500

2000

1500

1000

500

0

2500

2000

1500

1000

500

0

100

80

60

40

20

0

100

80

60

40

20

0C57 FVB 129 DBA

Strain

C57 FVB 129 DBAStrain

Error bars: ± 1 SE

Error bars: ± 1 SE

Error bars: ± 1 SE

C57 FVB 129 DBAStrain

Error bars: ± 1 SE

C57 FVB 129 DBAStrain

Error bars: ± 1 SE

Error bars: ± 1 SE Error bars: ± 1 SE

Error bars: ± 1 SE Error bars: ± 1 SE

C57 FVB 129 DBAStrain

C57 FVB 129 DBAStrain

C57 FVB 129 DBAStrain

C57 FVB 129 DBAStrain

C57 FVB 129 DBAStrain

ConditionMDControl CMS

Data overview: mean ± standard error (SE) of the mean for the complete experimental data, collapsed by sex and treatment duration (seesupplementary data for separate plots). CMS, chronic mild stress; MD, maternal deprivation.

782 Pharmacogenetics and Genomics 2011, Vol 21 No 12

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

The Porsolt forced-swim test

Animals were observed for 6 min, and mobility during the

last 4 min was used for analysis. Immobility in the FST

suggests that the animal displays behavioural despair and

is considered to be a depression-related behaviour.

Increased mobility after drug treatment reflects an

antidepressant effect.

A factorial ANOVA revealed a significant strain effect for

both male [F(3,57) = 11.112, P < 0.001] and female

animals [F(3,56) = 16 433, P < 0.001], but no significant

effect for the factor treatment duration or the interaction

strain� treatment duration (Fig. 3). A Bonferroni post-

hoc test indicated a significant difference among all four

strains, but the main effect was largely seen in strain 129

differing from FVB and DBA. However, measurements of

FST showed a ceiling effect in two of the strains: DBA

and FVB. Therefore, parametric assumptions for homo-

geneity of variance across strains were violated and results

must be interpreted with caution.

We further investigated the effect of the pharmacological

and environmental challenges in strain 129. In male

animals, strain 129 showed no significant difference

between subchronic and acute drug administration in

relation to the percent of time spent mobile; therefore,

the factor treatment duration was ignored for the

following calculations. The next series of analysis looked

at the effects of MS in male 129 mice. An independent-

sample t-test confirmed a significant difference between

MS and control [t(30) = 3.13, P < 0.003], with vehicle

MS animals showing decreased mobility compared with

vehicle control 129 mice. However, there was a floor

effect in mobility behaviour, resulting in a negatively

skewed distribution. Therefore, to confirm our findings

we used survival statistics as a distribution-free analysis to

investigate the effect of MS versus control in 129 male

mice treated with saline. The results confirmed a

significant difference between MS and control mice

[log-rank Mantel Cox X2 = 4.190, degree of freedom

(df) = 1, P < 0.041].

Nortriptyline was shown to be effective in reversing the

depressant-like effect of MS on male 129 mice. There

was a significant difference between animals treated

with vehicle and those treated with nortriptyline [t(30) = 27.83, P < 0.001]. The effect of nortriptyline was

specific to MS as the drug did not significantly affect

animals in the control or UCMS group (Fig. 4).

The magnitude of the effect of escitalopram on strain 129

is both striking and pervasive across sexes, treatment

duration and environment. The directionality of the

effect is, however, largely unexpected as it suggests that

escitalopram has a depressogenic rather than an anti-

depressant effect on 129 mice. The effect of escitalopram

on mobility measurements was not dependent on the en-

vironmental manipulations, as this drug reduced mobility

in control mice as much as it did in MS and UCMS mice.

Survival analysis shows that there is no difference

between subchronic and acute treatment in male (Mantel

Cox X2 = 2.774, df = 1, P > 0.09) and female animals

(X2 = 0.670, df = 1, P < 0.410). Overall, the effect of

escitalopram contrasted with the ameliorative effect of

nortiptyline in strain 129.

The Hole-board test

In addition to the nose-poke measure of exploration, we

derived measures of anxiety (time spent in the central

area of the hole board) and activity (distance travelled in

the hole board). To obtain the maximum value from the

testing, we were able to extract an open-field-like

locomotion measure from the hole-board data. Admit-

tedly, this is not a proper open field because of its smaller

size, and there will be an effect of the presence of holes;

however, it is a well-defined experimental measure of

Fig. 3

FST (male mice) FST (female mice)120

(a)

100

80

Per

cent

mob

ile

Per

cent

mob

ile

60

40

20

129m C57m DBAm FVBm

Strain

129f C57f DBAf FVBf

Strain

0

120

100

80

60

40

20

0

(b)

Mean ( ± standard error of the mean) percent mobility of control vehicle (single-housed, saline-injected mice) animals in the forced-swim test (FST)separated by (a) male and (b) female rats shows a significant difference between the four strains (each strain differs from the other). There was nodifference in relation to the duration of treatment. Acute injection = black bars, subchronic injection = white bars (n = 8 mice per group).

Antidepressants and early life stress Binder et al. 783

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

locomotion in an open arena with a different inbred strain

profile from the hole-board nose-poke measure.

Distance travelled

A factorial ANOVA comparing strain� environment in

vehicle-treated animals showed an expected difference

between the strains [F(3,1154) = 341.57, P < 0.01, Fig. 6]

with regard to environment [F(2,1154) = 14.121, P < 0.01]

but no significant strain� environment interaction. It

seems that DBA and 129 mice are less active than FVB

and C57 animals. The results further suggest an increased

activity of mice that underwent a UCMS treatment

compared with control and MS. A strain� drug comparison

shows a significant interaction [F(6,1192) = 5.3173,

P < 0.01], with an escitalopram-induced increase in

activity in C57 animals (P < 0.01). No differences in

activity were seen with nortriptyline (Fig. 5).

To understand whether our behavioural findings in the

FST in strain 129 are independent of any changes in

activity in the HBT, we compared:

(1) Vehicle animals of the control group with animals of

the MS group to investigate whether the environ-

mental manipulation induced a difference in activity.

(2) Control 129 male animals treated with nortriptyline

with vehicle mice to understand whether the drug

per se induces a change in activity.

(3) 129 male animals after MS treated with saline with

nortriptyline-treated animals to confirm that the

reversal of the depression-like behaviour shown in

the FST is not accompanied by a change in general

activity.

None of the three analyses showed a significant effect,

suggesting that the difference seen in the FST is not due

to drug or environment-induced sedation or arousal.

Exploration and anxiety

The number of holes explored by the animals in the HBT

is considered to be a measure of exploratory behaviour.

The time an animal spends at the centre of an open arena

is thought to reflect anxiety, with increasing time spent at

the centre indicative of reduced anxiety. A one-way

ANOVA considering the number of holes explored and

time spent at the centre showed a large strain difference

[number of holes: F(3,1200) = 245.44, P < 0.01; time

spent at the centre: F(3,1200) = 293.37, P < 0.01].

These behaviours had a similar distribution, with strains

that spent less time at the centre exploring fewer holes

(Fig. 6).

Exploration and anxiety are considered to be related

traits, with tests of spontaneous behaviour reflecting a

conflict between exploratory drive versus anxiety/defen-

sive behaviour [30–32]. In this case, there is also

interdependence because the measures are obtained in

the same arena.

A two-way ANOVA on anxiety-related behaviour and

exploration showed a significant drug� strain interaction

[F(6,1192) = 6.4, P < 0.01]. A Bonferroni post-hoc ana-

lysis revealed decreased anxiety-related behaviour in 129

mice after escitalopram treatment. Unexpectedly,

although escitalopram decreased anxiety-related beha-

viour, it also reduced the number of holes explored

(significant factor drug for number of holes explored

[F(2,1192) = 27.432, P < 0.01] compared with saline-

treated mice (Fig. 6).

Fig. 4

100

(a)FST

∗ ∗

80

60

40

20

Saline Nortriptyline Escitalopram

Drug

Saline Nortriptyline Escitalopram

Drug

0

Per

cent

mob

ile

100HBT

∗

∗

150

100

50

0

Tim

e sp

ent i

n th

e ce

nter

(b)

Mean ( ± standard error of the mean) percent mobility in the forced-swim test [FST (a)] and time spent at the centre of the [HBT (b)]. A total of 129male mice show a significant long-lasting depressogenic effect of MS (white bars), which was reversed with nortriptyline treatment in the FST.Nortriptyline had no effect on control (black bars) mice or on unpredictable chronic mild stress UCMS (grey bars) mice. In contrast, escitalopram hada strong depressogenic effect in the FST test (a) but induced a reduction in anxiety-related behaviour in the hole-board test [HBT (b)] independent ofthe stressor applied (n = 16 mice per group).

784 Pharmacogenetics and Genomics 2011, Vol 21 No 12

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

As anxiety-related behaviour is closely connected to

depression, we analysed the anxiety-related behaviour of

129 male mice, comparing environment (MS and control)

with drug (nortriptyline and saline). The results suggest

that MS animals were more anxious than control mice

[F(1,60) = 5.5159, P = 0.02; Fig. 4b].

Drug levels in the brain

A further investigation was conducted to compare drug

levels in the hippocampus of mice. The two drugs used in

this study have different mechanisms of action, which can

interact with strain differences. Drug measurements were

obtained at the time of killing 10 animals of each strain

treated chronically with either escitalopram or nortripty-

line. Measurements of drug levels in the hippocampus

can inform on the average levels of drug that infiltrate the

brain–blood barrier and can be more informative than

dosage quantity alone. The results show that there is no

significant variation in the level of escitalopram across the

four strains; however, there is a differential effect of

nortriptyline, with strains FVB and 129 showing a higher

concentration of drug compared with C57 and DBA

(Fig. 7).

DiscussionOur objective was to test for genetic background and

environmental effects on depression-related behaviour

and response to antidepressants. The behavioural readout

we used was the FST and the HBT in a large factorial

design that we believe is unprecedented. The FST was

developed in rats to model behavioural despair (passivity

in an aversive environment) and has been validated as a

test of antidepressant response [29]. A broad spectrum

of antidepressants reduces immobility in the FST, but for

SSRIs, in particular, this is dependent on careful choice of

experimental details – for example, the effect is seen

only in sufficiently deep water [33,34]. In the original

version of the FST, animals are subjected to the test

twice. The immobility behaviour of the animal could

consequently be interpreted as ‘learned immobility

response,’ and therefore be identified as an adaptive

coping strategy rather than a depression-related beha-

viour [35,36].

As with many other behaviour tests, the FST has more

recently been applied from rats to mice [37]. In mice, a

single-trial version of the test as opposed to the two-trial

paradigm for the rat resulted in an immobility-reducing

effect after administering a variety of antidepressants

Fig. 6

Hole board

140

(a)

120

100

80

Tim

e sp

ent i

n th

e ce

nter

(in

s)

60

40

20

129 C57 DBA FVB

Strain Strain

129 C57 DBA FVB0

50Hole board

40

Num

ber o

f hol

es e

xplo

red

30

20

10

0

(b)

Mean ( ± standard error of the mean) time (s) spent at the centre (a) and number of head dips (b) in the hole board (HB) of control vehicle (blackbars) and control escitalopram (white bars) mice. The anxiety-related measure and the explorative behaviour of vehicle animals showed a similardistribution. Animals that spent more time at the centre (a) explored more holes (b). This common behavioural pattern was disrupted in 129 mice afterescitalopram treatment (n = 32).

Fig. 5

3000

2000

2500 ∗

Activity in the hole board

1000

500

C57 DBA FVB

Strain

129

1500

0

Dis

tanc

e in

cm

Mean ( ± standard error of the mean) activity in the hole-board test(HBT) of control (single housed) mice. Apart from expected straindifferences, no drug-induced activity change was observed, except inC57 animals after escitalopram treatment (n = 32; saline = black bars,escitalopram = white bars and nortriptyline = grey bars).

Antidepressants and early life stress Binder et al. 785

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

[37,38]; however, once again, the effects of SSRIs are

harder to detect [39].

The application of the FST from rats to mice is

experimentally straightforward but requires separate

validation and careful consideration of the distinct biology

of the mouse. Numerous mutations of depression-

relevant genes, including 5HT1a [40], exhibit FST

phenotypes and offer one form of validation.

The most striking feature of the data overview presented

in Fig. 2 is the robust profile of strain differences. This

was expected as the strains were chosen on the basis of

the strain survey by Lucki et al., [41], and subsequent

studies have shown strain-dependent differences regard-

ing the response to stressors [7] and the response to

different antidepressants [42], as well as differences in

the behavioural phenotype of inbred mice [43,44]. Differ-

ent and equally striking strain profiles are seen for the

exploration and anxiety measures of our tests (Fig. 5).

Overall, we found reduced immobility in the notrypty-

line-treated groups, but increased immobility with

escitalopram (Fig. 8).

The escitalopram effect on FST is large and not due to

sedation, as evidenced by the lack of difference in

distance travelled or by exploration in the HBT. To date,

studies have produced contradictory data regarding the

action of escitalopram or citalopram in inbred mice [45].

A recent study reported a positive effect of one acute

dose of citalopram on strain 129/Sv, whereas it showed no

effect on strain DBA/2J [46]. A recent letter in

‘Neuroscience letters’ argued that contradictory results

are often due to the use of the wrong animal model. They

concluded that valid animal models of depression are

‘disease’ animals (in this case the selection bred Flinders

rats) and not stressed healthy animals [47].

Although there is some consensus on experimental

paradigms to measure the action of antidepressants such

as the FST [48], there is less agreement on environ-

mental manipulations leading to a depression-like beha-

vioural abnormality in mice [26,49,50]. How to induce

‘depressed mood’ in a rodent and generate a consistent

and replicable model of depression remains unclear. It is

also a formidable task to make a careful comparison of

available and potential protocols. Above all, we were

concerned that for our pharmacogenomic study we should

use a protocol that will most likely produce a measurable

effect, and so chose to test UCMS and a long MS

protocol.

Ellenbroek and Cools [51] characterized the phenotype

of rats after a 24-h MS as being schizophrenia-like on the

basis of a range of measures including prepulse inhibition.

Fabricius et al., [13] transported this manipulation to mice

and tested 10-week-old C57BL/6j mice in open-field,

elevated-plus and Barnes mazes. They detected a

reduction in anxiety-related behaviour in the elevated

plus maze in MS animals with an increased open arm

time. They also detected decreased behavioural flex-

ibility (perseveration) in the Barnes maze and an effect

on hippocampal neuron number, but did not report the

results of tests of depression-related behaviours. We

found a long-lasting, strain-specific effect of MS on the

Fig. 7

90

80

FST

∗

∗

70

60

50

Per

cent

mob

ile

40

30

20

10

Saline EscitalopramDrug

Nortriptyline0

Shows the mean amount of active drug found in the hippocampus ofanimals at the time of euthanization by strain. No difference wasobserved in the levels of escitalopram across strains but there was adifference in the levels of nortriptyline, with strain 129 and FVB differingfrom C57 and DBA.

Fig. 8

2000

1500

1000

500

0

Mea

n dr

ug le

vels

in b

rain

(mm

ol)

Escitalopram

Antidepressant drug

Nortriptyline

Strain129C57DBAFVB

Mean [ ± standard error (SE) of the mean] mobility in the forced-swimtest (FST) across environmental stressors, sex and treatment durationshows an antidepressant effect of nortriptyline and a depressogenicinfluence of escitalopram (n = 384).

786 Pharmacogenetics and Genomics 2011, Vol 21 No 12

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

FST, with an increased depression-like behaviour seen in

129S1/SvImJ mice. The therapeutic effect of nortripty-

line across strains and sex shows a particularly strong

impact on strain 129 where it was able to reverse the

depressogenic effect of MS (Fig. 3). Overall, nortriptyline

appeared to reduce depression-related behaviour across

strains, sexes and drug administration regimes (Fig. 5).

Resilience or vulnerability to stress is time, strain, sex and

stressor dependent. Recent literature shows that resi-

lience to stress and antidepressant action can be directly

linked to a genetic manipulation in mice. Expression

levels of the serotonin-1A autoreceptor [52], induction of

the transcription factor DFosB [53] and a genetic

variation of the 2-amino-3-(5-methyl-3-oxo-1,2-oxazol-

4-yl)propanoic acid receptor [54] result in changes in

resilience to stress in mice. Furthermore, the same

pathways were necessary to convert SSRI nonresponders

to responders. Susceptibility to stress has not just been

linked to genetic polymorphism but also to dynamic

epigenetic changes. DNA methylation can make mice

more or less responsive to stressors [55,56].

Our results suggest that the choice and length of our

stress protocol, together with the genetic predisposition

of the animals, did not show a long-lasting effect in three

of four strains.

In FVB and DBA, the baseline strain pattern is extreme

enough that there is a ceiling effect on mobility and the

test is uninformative. This may suggest that in strains

known to have high levels of activity, such as FVB, a

longer swimming interval or perhaps a different test

paradigm altogether may have been more appropriate. A

further consideration is with regard to the lack of a normal

distribution of strain 129. The use of survival analysis as a

distribution-free method yielded more rigorous P values,

reinforcing the significance thresholds detected by our

parametric test. The fact that anxiety-related measures

from the HBT confirmed the depression-related results

seen in the FSTadds to the assumption that male animals

of strain 129 after one single 24-h MS have potential to be

models for depression-related behaviour.

The 129 inbred mouse strain consists of 13 different

substrains known to differ in their behavioural pheno-

type [57]. The substrain 129S1/SvlmJ is a popular

background strain for transgenic mice, as embryonic stem

cell lines of that strain are commonly available. To date,

only a few studies have shown interest in the behavioural

phenotype of this strain as it is known to be passive,

hypoactive, to perform poorly in memory tasks and in

stress-related paradigms under certain conditions and

show an absence of the corpus callosum [58–60]. In

recent years, the importance of the choice of background

strain in transgenic animals has become a hot topic [61–63]

and many knockout mice already available have been

generated using 129 embryonic stem cells. In a study

comparing two different backgrounds of null mice, an

increased susceptibility to stress was found in 129/SvEv

compared with C57BL/N mice, which made a difference

in the outcome of the phenotype of the null mouse

models [64]. It is difficult to determine whether

a phenotype observed in a mutant strain is due to the

mutation or the background strain chosen, especially

if the background is a mixture of two strains – for example,

129 for the embryonic stem cells and C57 for the

backcrossing [65].

Although strain 129 has not been extensively used in

studies on depression, the results show that this strain is

more sensitive to both depressogenic protocols and

antidepressant action compared with other commonly

used strains. This strain may prove to be a valuable choice

in future animal studies on depression.

A gene expression study on the hippocampus of mice

prepared in parallel with these has been compared with

the results of a pharmacogenomic genome-wide associa-

tion study using the same drugs. This comparison

produced convergent evidence from mice and humans,

suggesting a role of the Ppm1a gene in response to

nortryptiline but not escitalopram [66].

AcknowledgementsThe Genome-based Therapeutic Drugs for Depression

study was funded by a European Commission Framework

6 Grant, EC Contract Ref.: LSHB-CT-2003–503428.

Lundbeck provided both nortriptyline and escitalopram

free of charge. K. Aitchison has received consultancy fees

and honoraria for participating in expert panels from

pharmaceutical companies including Lundbeck and

GlaxoSmithKline.

Conflicts of interest

There are no conflicts of interest.

References1 Thase ME, Entsuah AR, Rudolph RL. Remission rates during treatment with

venlafaxine or selective serotonin reuptake inhibitors. Br J Psychiatry 2001;178:234–241.

2 Uher R, Farmer A, Maier W, Rietschel M, Hauser J, Marusic A, et al.Measuring depression: comparison and integration of three scales in theGENDEP study. Psychol Med 2008; 38:289–300.

3 El YM, Vaugeois JM. Genetic rodent models of depression. Curr OpinPharmacol 2007; 7:3–7.

4 Overstreet DH, Pucilowski O, Rezvani AH, Janowsky DS. Administrationof antidepressants, diazepam and psychomotor stimulants further confirmsthe utility of flinders sensitive line rats as an animal model of depression.Psychopharmacology (Berl) 1995; 121:27–37.

5 Kromer SA, Kessler MS, Milfay D, Birg IN, Bunck M, Czibere L, et al.Identification of glyoxalase-I as a protein marker in a mouse model ofextremes in trait anxiety. J Neurosci 2005; 25:4375–4384.

6 Cryan JF, Holmes A. The ascent of mouse: advances in modelling humandepression and anxiety. Nat Rev Drug Discov 2005; 4:775–790.

7 Mineur YS, Belzung C, Crusio WE. Effects of unpredictable chronic mildstress on anxiety and depression-like behavior in mice. Behav Brain Res2006; 175:43–50.

8 Stansfeld SA, Fuhrer R, Head J, Ferrie J, Shipley M. Work and psychiatricdisorder in the Whitehall II study. J Psychosom Res 1997; 43:73–81.

9 Lyons DM, Parker KJ, Schatzberg AF. Animal models of early life stress:implications for understanding resilience. Dev Psychobiol 2010; 52:616–624.

Antidepressants and early life stress Binder et al. 787

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

10 Gubernick DJ, Alberts JR. Maternal licking of young: resource exchange andproximate controls. Physiol Behav 1983; 31:593–601.

11 de Kloet ER, Sibug RM, Helmerhorst FM, Schmidt MV. Stress, genes andthe mechanism of programming the brain for later life. Neurosci BiobehavRev 2005; 29:271–281.

12 Anisman H, Zaharia MD, Meaney MJ, Merali Z. Do early-life eventspermanently alter behavioral and hormonal responses to stressors?Int J Dev Neurosci 1998; 16:149–164.

13 Fabricius K, Wortwein G, Pakkenberg B. The impact of maternal separationon adult mouse behaviour and on the total neuron number in the mousehippocampus. Brain Struct Funct 2008; 212:403–416.

14 Franklin TB, Russig H, Weiss IC, Graff J, Linder N, Michalon A, et al.Epigenetic transmission of the impact of early stress across generations.Biol Psychiatry 2010; 68:408–415.

15 George ED, Bordner KA, Elwafi HM, Simen AA. Maternal separation withearly weaning: a novel mouse model of early life neglect. BMC Neurosci2010; 11:123.

16 Van Heerden JH, Russell V, Korff A, Stein DJ, Illing N. Evaluating thebehavioural consequences of early maternal separation in adult C57BL/6mice; the importance of time. Behav Brain Res 2010; 207:332–342.

17 Claessens SE, Daskalakis NP, Van der Veen R, Oitzl MS, de Kloet ER,Champagne DL. Development of individual differences in stressresponsiveness: an overview of factors mediating the outcome of earlylife experiences. Psychopharmacology 2011; 214:141–154.

18 Groenink L, Bijlsma EY, Van Bogaert MJ, Oosting RS, Olivier B. Serotonin1Areceptor deletion does not interact with maternal separation-inducedincreases in startle reactivity and prepulse inhibition deficits.Psychopharmacology (Berl) 2011; 214:353–365.

19 Kessler MS, Bosch OJ, Bunck M, Landgraf R, Neumann ID. Maternal carediffers in mice bred for high versus low trait anxiety: impact of brainvasopressin and cross-fostering. Soc Neurosci 2011; 6:156–168.

20 Mehta M, Schmauss C. Strain-specific cognitive deficits in adult miceexposed to early life stress. Behav Neurosci 2011; 125:29–36.

21 Niwa M, Matsumoto Y, Mouri A, Ozaki N, Nabeshima T. Vulnerability in earlylife to changes in the rearing environment plays a crucial role in theaetiopathology of psychiatric disorders. Int J Neuropsychopharmacol 2011;14:459–477.

22 Weiss IC, Franklin TB, Vizi S, Mansuy IM. Inheritable effect of unpredictablematernal separation on behavioral responses in mice. Front Behav Neurosci2011; 5:3.

23 Levine S. Influence of psychological variables on the activity of thehypothalamic-pituitary-adrenal axis. Eur J Pharmacol 2000; 405:149–160.

24 Pryce CR, Bettschen D, Nanz-Bahr NI, Feldon J. Comparison of theeffects of early handling and early deprivation on conditioned stimulus,context, and spatial learning and memory in adult rats. Behav Neurosci2003; 117:883–893.

25 Macri S, Granstrem O, Shumilina M, Antunes Gomes dos Santos FJ,Berry A, Saso L, Laviola G. Resilience and vulnerability are dose-dependently related to neonatal stressors in mice. Horm Behav 2009;56:391–398.

26 Millstein RA, Holmes A. Effects of repeated maternal separation on anxiety-and depression-related phenotypes in different mouse strains. NeurosciBiobehav Rev 2007; 31:3–17.

27 Neumann ID, Kromer SA, Bosch OJ. Effects of psycho-social stress duringpregnancy on neuroendocrine and behavioural parameters in lactationdepend on the genetically determined stress vulnerability.Psychoneuroendocrinology 2005; 30:791–806.

28 Savignac HM, Dinan TG, Cryan JF. Resistance to early-life stress in mice:effects of genetic background and stress duration. Front Behav Neurosci2011; 4:5–13.

29 Porsolt RD, Anton G, Blavet N, Jalfre M. Behavioural despair in rats: a newmodel sensitive to antidepressant treatments. Eur J Pharmacol 1978;47:379–391.

30 Crawley J, Goodwin FK. Preliminary report of a simple animal behaviormodel for the anxiolytic effects of benzodiazepines. Pharmacol BiochemBehav 1980; 13:167–170.

31 Handley SL, Mithani S. Effects of alpha-adrenoceptor agonists andantagonists in a maze-exploration model of ‘fear’-motivated behaviour.Naunyn Schmiedebergs Arch Pharmacol 1984; 327:1–5.

32 Pellow S, Chopin P, File SE, Briley M. Validation of open:closed arm entriesin an elevated plus-maze as a measure of anxiety in the rat. J NeurosciMethods 1985; 14:149–167.

33 Detke MJ, Lucki I. Detection of serotonergic and noradrenergicantidepressants in the rat forced swimming test: the effects of water depth.Behav Brain Res 1996; 73:43–46.

34 Cryan JF, Valentino RJ, Lucki I. Assessing substrates underlying thebehavioral effects of antidepressants using the modified rat forcedswimming test. Neurosci Biobehav Rev 2005; 29:547–569.

35 Hawkins J, Hicks RA, Phillips N, Moore JD. Swimming rats and humandepression. Nature 1978; 274:512–513.

36 De Pablo JM, Ortiz-Caro J, Sanchez-Santed F, Guillamon A. Effects ofdiazepam, pentobarbital, scopolamine and the timing of saline injectionon learned immobility in rats. Physiol Behav 1991; 50:895–899.

37 Petit-Demouliere B, Chenu F, Bourin M. Forced swimming test in mice:a review of antidepressant activity. Psychopharmacology (Berl) 2005;177:245–255.

38 Borsini F, Meli A. Is the forced swimming test a suitable model forrevealing antidepressant activity? Psychopharmacology (Berl) 1988;94:147–160.

39 Castagne V, Porsolt RD, Moser P. Use of latency to immobility improvesdetection of antidepressant-like activity in the behavioral despair test inthe mouse. Eur J Pharmacol 2009; 616:128–133.

40 Heisler LK, Chu HM, Brennan TJ, Danao JA, Bajwa P, Parsons LH,Tecott LH. Elevated anxiety and antidepressant-like responses inserotonin 5-HT1A receptor mutant mice. Proc Natl Acad Sci USA 1998;95:15049–15054.

41 Lucki I, Dalvi A, Mayorga AJ. Sensitivity to the effects of pharmacologicallyselective antidepressants in different strains of mice. Psychopharmacology(Berl) 2001; 155:315–322.

42 Ibarguen-Vargas Y, Surget A, Touma C, Palme R, Belzung C. Multifacetedstrain-specific effects in a mouse model of depression and of antidepressantreversal. Psychoneuroendocrinology 2008; 33:1357–1368.

43 Voikar V, Koks S, Vasar E, Rauvala H. Strain and gender differences in thebehavior of mouse lines commonly used in transgenic studies. PhysiolBehav 2001; 72:271–281.

44 Jacobson LH, Cryan JF. Feeling strained? Influence of genetic backgroundon depression-related behavior in mice: a review. Behav Genet 2007;37:171–213.

45 Borsini F. Role of the serotonergic system in the forced swimming test.Neurosci Biobehav Rev 1995; 19:377–395.

46 Cervo L, Canetta A, Calcagno E, Burbassi S, Sacchetti G, Caccia S, et al.Genotype-dependent activity of tryptophan hydroxylase-2 determines theresponse to citalopram in a mouse model of depression. J Neurosci 2005;25:8165–8172.

47 Petersen A, Wortwein G, Gruber SH, Mathe AA. Escitalopram reducesincreased hippocampal cytogenesis in a genetic rat depression model.Neurosci Lett 2008; 436:305–308.

48 Cryan JF, Mombereau C. In search of a depressed mouse: utility ofmodels for studying depression-related behavior in geneticallymodified mice. Mol Psychiatry 2004; 9:326–357.

49 Willner P. Chronic mild stress (CMS) revisited: consistency and behavioural-neurobiological concordance in the effects of CMS. Neuropsychobiology2005; 52:90–110.

50 Millstein RA, Ralph RJ, Yang RJ, Holmes A. Effects of repeated maternalseparation on prepulse inhibition of startle across inbred mouse strains.Genes Brain Behav 2006; 5:346–354.

51 Ellenbroek BA, Cools AR. Early maternal deprivation and prepulse inhibition:the role of the postdeprivation environment. Pharmacol Biochem Behav2002; 73:177–184.

52 Richardson-Jones JW, Craige CP, Guiard BP, Stephen A, Metzger KL,Kung HF, et al. 5-HT1A autoreceptor levels determine vulnerability tostress and response to antidepressants. Neuron 14 2010; 65:40–52.

53 Vialou V, Robison AJ, Laplant QC, Covington HE 3rd, Dietz DM, Ohnishi YN,et al. Delta FosB in brain reward circuits mediates resilience to stress andantidepressant responses. Nat Neurosci 2010; 13:745–752.

54 Schmidt MV, Trumbach D, Weber P, Wagner K, Scharf SH, Liebl C, et al.Individual stress vulnerability is predicted by short-term memory andAMPA receptor subunit ratio in the hippocampus. J Neurosci 2010;30:16949–16958.

55 Elliott E, Ezra-Nevo G, Regev L, Neufeld-Cohen A, Chen A. Resilience tosocial stress coincides with functional DNA methylation of the Crf gene inadult mice. Nat Neurosci 2010; 13:1351–1353.

56 Uchida S, Hara K, Kobayashi A, Otsuki K, Yamagata H, Hobara T, et al.Epigenetic status of Gdnf in the ventral striatum determines susceptibilityand adaptation to daily stressful events. Neuron 2011; 69:359–372.

57 Simpson EM, Linder CC, Sargent EE, Davisson MT, Mobraaten LE, Sharp JJ.Genetic variation among 129 substrains and its importance for targetedmutagenesis in mice. Nat Genet 1997; 16:19–27.

58 Crawley JN. Unusual behavioral phenotypes of inbred mouse strains. TrendsNeurosci 1996; 19:181–182.

788 Pharmacogenetics and Genomics 2011, Vol 21 No 12

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

59 Gerlai R. Gene-targeting studies of mammalian behavior: is it themutation or the background genotype? Trends Neurosci 1996;19:177–181.

60 Crawley JN, Belknap JK, Collins A, Crabbe JC, Frankel W, Henderson N,et al. Behavioral phenotypes of inbred mouse strains: implications andrecommendations for molecular studies. Psychopharmacology (Berl) 1997;132:107–124.

61 Montagutelli X. Effect of the genetic background on the phenotype of mousemutations. J Am Soc Nephrol 2000; 11 (Suppl 16):S101–S105.

62 Fernandes C, Paya-Cano JL, Sluyter F, D’Souza U, Plomin R, Schalkwyk LC.Hippocampal gene expression profiling across eight mouse inbred strains:towards understanding the molecular basis for behaviour. Eur J Neurosci2004; 19:2576–2582.

63 Smith CM, Lawrence AJ, Sutton SW, Gundlach AL. Behavioral phenotypingof mixed background (129S5:B6) relaxin-3 knockout mice. Ann NY AcadSci 2009; 1160:236–241.

64 Peinado JR, Laurent V, Lee SN, Peng BW, Pintar JE, Steiner DF, Lindberg I.Strain-dependent influences on the hypothalamo-pituitary-adrenal axisprofoundly affect the 7B2 and PC2 null phenotypes. Endocrinology 2005;146:3438–3444.

65 Schalkwyk LC, Fernandes C, Nash MW, Kurrikoff K, Vasar E, Koks S.Interpretation of knockout experiments: the congenic footprint. Genes,Brain and Behavior 2007; 6:299–303.

66 Malki K, Uher R, Paya-Cano J, Binder E, Rietschel M, Zobel A, et al.Convergent animal and human evidence suggests a role of PPM1A genein response to antidepressants. Biol Psychiatry 2011; 69:360–365.

Antidepressants and early life stress Binder et al. 789

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.