Molecular Psychiatry (1999) 4, 443–452
 1999 Stockton Press All rights reserved 1359–4184/99 $15.00
ORIGINAL RESEARCH ARTICLE
Genetic alteration of the ␣2-adrenoceptor subtype c in
mice affects the development of behavioral despair and
stress-induced increases in plasma corticosterone levels
J Sallinen1, A Haapalinna1, E MacDonald3, T Viitamaa1, J Lähdesmäki2, E Rybnikova4,
M Pelto-Huikko5, BK Kobilka6 and M Scheinin2
1
Orion Corporation, Orion Pharma, FIN-20101 Turku, Finland; 2Department of Pharmacology and Clinical Pharmacology,
University of Turku, FIN-20520 Turku, Finland; 3Department of Pharmacology and Toxicology, University of Kuopio, POB
1627, FIN-70211 Kuopio, Finland; 4Pavlov Institute of Physiology, 199034 St Petersburg, Russia; 5Department of
Anatomy, Tampere University Medical School and Department of Pathology, Tampere University Hospital, FIN-33101
Tampere, Finland; 6Department of Molecular and Cellular Physiology, Howard Hughes Medical Institute, and Division of
Cardiovascular Medicine, Stanford University, Stanford, CA 94305, USA
␣2-Adrenoceptors (␣2-AR) modulate many central nervous system functions, such as regulation of sympathetic tone, vigilance, attention, and reactivity to environmental stressors.
Three ␣2-AR subtypes (␣2A, ␣2B, and ␣2C) with distinct tissue-distribution patterns are known
to exist, but the functional significance of each subtype is not clear. Since specific, ␣2-AR
subtype-selective pharmacological probes are not available, mice with genetically altered ␣2CAR expression were studied in order to investigate the possible involvement of the ␣2C-AR in
physiological and behavioral responses to acute and repeated stress. A modified version of
Porsolt’s forced swimming test was used to assess the possible effects of altered ␣2C-AR
expression on the development of behavioral despair. ␣2C-Overexpression increased and the
lack of ␣2C-AR (␣2C-KO) decreased the immobility of mice in the forced swimming test, ie ␣2CAR expression appeared to promote the development of behavioral despair. In addition, ␣2CKO was associated with attenuated elevation of plasma corticosterone after different stressors, and overexpression of ␣2C-ARs was linked with increased corticosterone levels after
repeated stress. Moreover, the brain dopamine and serotonin balance, but not norepinephrine
turnover, was dependent on ␣2C-AR expression, and the expression of c-fos and junB mRNA
was increased in ␣2C-KO mice. Since ␣2C-KO produced stress-protective effects, and ␣2C-AR
overexpression seemed to promote the development of changes related to depression, it is
suggested that a yet-to-be developed subtype-selective ␣2C-AR antagonist might have therapeutic value in the treatment of stress-related neuropsychiatric disorders.
Keywords: adrenergic receptors; antidepressant; biogenic amines; forced swimming test; hippocampus; immediate-early gene; transgenic mice
Introduction
␣2-ARs inhibit neuronal firing and regulate the release
of norepinephrine (NE), serotonin (5-HT), dopamine
(DA), and other brain neurotransmitters, thus participating in a variety of physiological functions, including
modulation of the central nervous system (CNS) reactivity to environmental stressors.1–5 Recently, the heterogeneity of ␣2-ARs has been delineated.6 The three
cloned ␣2-AR subtypes, designated as ␣2A, ␣2B, and ␣2C,
are well conserved across species and have distinct
tissue and cellular distributions.7–12 All ␣2-AR subtypes couple to inhibitory G-proteins, but also differ-
Correspondence: Dr J Sallinen, Orion Corporation, Orion Pharma,
PO Box 425, FIN-20101 Turku, Finland. E-mail: jukka.sallinen@
orion.fi
Received 29 October 1998; revised 7 January and 16 February
1999; accepted 16 February 1999
ences have been demonstrated in their intracellular
coupling and trafficking.13,14 Specific ␣2-AR subtypeselective drugs are not available and mainly therefore
the functional significance of the ␣2-AR subtypes has
remained obscure.
Much of the current knowledge on the physiological
roles of the ␣2-AR subtypes is based on recent studies
on genetically engineered mice.15 These studies have
demonstrated that the most evident CNS effects of subtype non-selective ␣2-AR agonists, ie sedation, hypotension, hypothermia, and analgesia, appear to be
mediated via ␣2A-ARs.16–20 It has also been demonstrated that genetically altered ␣2C-AR expression in
mice does not significantly change the potent inhibitory effect of ␣2-AR agonists on brain monoamine turnover.21 However, according to a recent study
employing transgenic mice, ␣2C-AR may have a subtle
modulatory effect on brain DA and 5-HT balance.21
Furthermore, the ␣2C-AR gene mutations in mice have
The ␣2c-adrenoceptor gene altered mice and stress
J Sallinen et al
444
been associated with changes in aggressive behavior,
the startle reflex and its prepulse inhibition,22 damphetamine-stimulated locomotor activity, and l-5hydroxytryptophan-induced behaviors.23 These results
raise the possibility that yet-to-be developed drugs
selective for the ␣2C-AR would have therapeutic potential in neuropsychiatric disorders and that they would
perhaps not have the typical sedative and cardiovascular adverse effects associated with current subtypenonselective ␣2-AR drugs.
In the current study, we explored whether ␣2C-ARs,
which are mainly expressed in the CNS, contribute to
the mediation of behavioral and physiological stress
responses. Porsolt’s forced swimming test (FST)24 was
selected for the evaluation of behavioral stress reactivity in two distinct transgenic mouse strains; one had
a targeted inactivation of the gene encoding the ␣2CAR (␣2C-knockout, ␣2C-KO), and the other had tissuespecific overexpression of ␣2C-ARs (␣2C-OE).21,25 In the
FST, animals are exposed to swim stress in order to
produce a stress reaction that decreases the animals’
activity in a subsequent test swim session; this
phenomenon is called ‘behavioral despair’.26 Since the
FST has predictive validity in the development of antidepressants,27–29 we aimed to predict whether drugs
acting via ␣2C-AR would possibly have either stressprotective and/or antidepressant-like actions in the
FST. Instead of drug administrations, we made an
assumption that ␣2C-KO mice would mimic the actions
of a chronically administered ␣2C-AR antagonist, and
that overexpression of the ␣2C-AR would mimic longterm treatment with an ␣2C-AR agonist. It should be
acknowledged, however, that the mutations have a lifelong effect on the ␣2C-AR expression; it is therefore not
possible to rule out that some developmental consequences resulting from the altered ␣2C-AR expression
may contribute to the results.
Stress is known to increase the release of NE as well
as other monoamines.4,30 Therefore, and because our
previous study hinted at ␣2C-AR-dependent modulation of DA and 5-HT balance in the brain,21 we investigated whether altered ␣2C-AR gene expression affects
the brain monoaminergic response to swim and
restraint stress. To further characterize the brain
regions where ␣2C-AR gene mutations might have
effects on neuronal activity, mRNA levels of the stressresponsive immediate-early genes c-fos and junB were
determined. Finally, stress-induced plasma corticosterone responses were determined as a general marker
for stress responsivity of the adrenocortical system.
Materials and methods
Animals
The experimental animals represented two strains of
genetically engineered mice and two types of wild-type
control mice. The mutations were generated at Stanford University (Stanford, CA, USA). One strain had a
targeted disruption of the ␣2C-AR gene (␣2C-KO) and the
other had tissue-specific overexpression of ␣2C-ARs
(␣2C-OE). The wild-type controls were littermates or
genetically closely related to the mutants, but because
the wild-type mice represented markedly different
strain backgrounds, they are designated as ␣2C-KO-wt
and ␣2C-OE-wt, respectively. A total of 401 male mice
were studied at the age of 10–20 weeks.
The generation of both mutant strains has been
described previously.21,25 Briefly, the ␣2C-AR gene was
inactivated in 129/Sv embryonic stem cells31 which
were injected into C57BL/6J blastocysts, and the
resulting chimeric mice were bred to F1 (C57BL/6J ×
DBA/2J) animals. These animals were back-crossed for
several generations to C57BL/6J mice; then intercrossed, and the tested ␣2C-KO mice were offspring of
closely related F11–12 pairs, which were combinations
of mice with wild-type, heterozygous or homozygous
genotypes for the ␣2C-AR mutation. All tested ␣2C-KO
mice were homozygous for the mutation. The germline
transmission of the mutation was monitored from
mouse tail biopsies by Southern (DNA) analysis.
The ␣2C-OE mice were generated by pronuclear
microinjection to one-cell fertilized eggs from the
FVB/N strain. A tyrosinase minigene construct was coinjected for visual identification of the transgenic progeny on the basis of coat color.32 The tested ␣2C-OE
mice were heterozygous. According to in situ mRNA
hybridization results and receptor autoradiography, the
overexpression of ␣2C-ARs is 苲3-fold in ␣2C-OE mice in
brain regions which normally express the receptor.21
Correspondingly, ␣2C-AR binding is absent in the
brains of ␣2C-KO mice.25
The mice were housed in groups of 6–10 in standard
polypropylene cages (38 cm × 22 cm × 15 cm) at 22 ±
1°C and kept on a 12:12-h light-dark cycle with light
onset at 06.00 h. Experiments were conducted between
08.30 and 16.00 h. The animal care was in accordance
with the regulation of the ICLAS (International Council
for Laboratory Animal Science) and the experiments
had approval of the local committee for laboratory animal welfare. Each animal was tested only once.
Experiment 1: Forced swimming test
A modified forced swimming test was used to evaluate
the effects of altered ␣2C-AR expression on the
behavioral stress responsivity of mice, ie to measure
the development of behavioral despair after different
durations of swim stress. Since no drugs were used,
the test was a modified version of the original test for
rats described by Porsolt et al.24 The mice were individually immersed for either 1, 2.5, 5, 10, or 20 min
in a transparent glass cylinder (20 cm high, 8 cm in
diameter), which contained water at 25°C to a depth of
8 cm; this first swim exposure was designated as preswim. The water was changed between each animal,
and the mice were gently dried after the pre-swim and
put into a fresh cage under a warming lamp for 15 min.
Those mice which were randomized to a 0-min preswim (control) group were also put under the warming
lamp from their home cages for 15 min. Twenty-four
hours later the mice underwent a 5-min test swim in
which the duration of activity of the mice in the water
was measured. Clear attempts to escape as well as
The ␣2c-adrenoceptor gene altered mice and stress
J Sallinen et al
Experiment 2: Analysis of brain biogenic amines and
their metabolites, c-fos amd junB mRNA, and plasma
corticosterone after acute and repeated stress
Groups of mice from each genotype were exposed to a
mild stress, 2.5 min swim in 25°C water, or to a moderate stress, 20 min immobilization by taping the animal’s body and tail to a bench top using 5 cm wide
easily removable tape. Twenty minutes from the beginning of the stress exposure, the animals were killed by
decapitation and trunk blood and brains were collected. The brains were rapidly frozen in an isopentane
bath in dry ice. In a separate experiment, groups of
mice of each genotype were immobilized for 2 h on 7
consecutive days. On the 7th day, brain and blood
samples were collected immediately after the 2-h
immobilization. Samples from control groups of each
genotype without stress exposure were taken simultaneously in both acute and repeated stress tests.
Immediate-early gene mRNA levels were determined
only from mice after acute stress exposure and their
controls. Biogenic amines and their metabolites were
determined from whole brain homogenates in 0.1 M
perchloric acid using electrochemical detection (ESA
Coulochem 5011, Bedford, MA, USA) after separation
by high-performance liquid chromatography (HPLC) on
a reversed-phase C18 column (Ultrasphere ODS, 4.6 ×
250 mm, Beckman Instruments, Fullerton, CA, USA).
The buffer systems described by Mefford33 were used,
with separate assays for indoles and the DA metabolite
homovanillic acid (HVA) and for catecholamines after
their purification on activated alumina. The minor
modifications to the original published procedures
have been described elsewhere.34 The NE metabolite,
3-methoxy-4-hydroxyphenylglycol
(MHPG),
was
determined in a separate HPLC assay using a modified
procedure previously validated for plasma samples.35
The brain homogenates were centrifuged, and the
supernatants were neutralized before extraction with
Bond Elut PH-columns (Analytichem International,
Harbor City, CA, USA).
end with [33P]dATP (Dupont-NEN, Boston, MA, USA)
using terminal deoxynucleotidyltransferase (Amersham,
Bucks, UK). Several control probes with the same
length and similar GC content and specific activity
were used to determine the specificity of the hybridizations. Addition of 100-fold excess of respective unlabelled probe abolished all hybridization signals.
In situ hybridization was carried out as described
previously.38 The slides were incubated in humidified
boxes at 42°C for 18 h with 5 ng ml−1 of the labelled
probe in a hybridization mixture containing 50% formamide (GT Baker, Deventer, Holland), 4 × SSC (1 ×
SSC consists of 0.15 M NaCl and 0.015 M sodium
citrate), 1 × Denhardt’s (0.02% bovine serum albumin,
0.02% Ficoll, 0.02% polyvinylpyrrolidone), 1% sarkosyl (N-laurylsarcosine; Sigma Chemical, St Louis, MO,
USA), 0.02 M sodium phosphate, pH 7.0, and 10% dextran sulfate (Pharmacia, Uppsala, Sweden). The sections were subsequently rinsed in 1 × SSC at 55°C for
60 min with four changes of SSC and finally in 1 × SSC
starting at 55°C and slowly cooled to room temperature
(⬇1 h), transferred through distilled water, briefly
dehydrated in 60% and 95% ethanol for 30 s each, airdried, and covered with Kodak Biomax MR autoradiography film (Kodak, Rochester, NY, USA) for 30 days.
Films were then developed with LX 24 developer
(Kodak) for 2 min and fixed with Unifix (Agfa-Gevaert,
Leverkusen, Germany) for 15 min.
The image analysis system consisted of an IBM-PC,
SensiCam digital camera (PCO Computer Optics
GmbH, Kelheim, Germany), Nikon 55-mm lens and
Northern Light precision illuminator (Imaging
Research, St Catharines, Ontario, Canada). The
measurements were performed using the Image-Pro
Plus program (Media Cybernetics, Silver Spring, MD,
USA). The gray levels corresponding to 14C-plastic
standards (Amersham) within the exposure range of
the film were determined and used in a third order
polynomial approximation to construct an optical density to radioactivity transfer function. The areas of
interest were interactively defined, and the average
activity in each area was calculated. At least five sections from each animal in each group were measured,
and their mean values were used in statistical analysis.
Corticosterone was measured in trunk blood samples
with radioimmunoassay using the Corticosterone 125I
RIA kit (ICN Biomedicals, Costa Mesa, CA, USA).
C-fos and junB mRNA expression levels were
determined by in situ hybridization
The brains from stressed and non-stressed ␣2C-KO and
␣2C-OE animals and their controls were frozen together
on the same specimen holder. The tissues were sectioned in a Microm HM 500 cryostat at 14 ␮m and
thaw-mounted onto Polysine glass slides (Menzel,
Germany). The sections were stored at −20°C until
used.
Oligonucleotides targeted to c-fos (nucleotides corresponding to amino acids 137–152 of rat c-fos36) and
junB (amino acids 195–210 of mouse junB37) were labeled to specific activity of 1 × 109 cpm ␮g−1 at the 3⬘
Data analysis
STATISTICA 4.5 computer software (StatSoft, Tulsa,
OK, USA) was used to determine the statistical significance of differences between various treatment and
genotype groups. P values smaller than 0.05 were considered as statistically significant. Comparisons
between genotypes were performed only between the
appropriate mutant and wild-type genotype pairs.
Multivariate ANOVA was used to explore overall differences between various treatment and genotype
groups and their possible interactions. Results from
brain biogenic amine and mRNA determinations were
evaluated with Student’s t-tests.
swimming around the cylinder were scored as activity
but minor limb movements to maintain the floating
posture were regarded as immotility. The observer was
blind to the mouse genotypes and pre-swim durations,
but the ␣2C-OE mice could be identified from their ␣2COE-wt controls by their different coat color.
445
The ␣2c-adrenoceptor gene altered mice and stress
J Sallinen et al
446
Results
Forced swimming tests
The mean duration of activity of the ␣2C-KO mice in the
test swim was clearly longer than that of their controls,
when all pre-swim groups (except the 0-min group)
were taken into analysis (Figure 1; 2-way ANOVA for
genotype difference: F(1,56) = 10.1, P = 0.0024). In contrast, the ␣2C-OE mice tended to be less active in the
test swim than their controls (Figure 2a; F(1,74) = 3.27,
P = 0.074). This genotype difference was statistically
significant when also the 0-min pre-swim group was
included in the analysis (F(1,87) = 3.99, P = 0.049). In
order to confirm the activity difference between ␣2C-OE
and ␣2C-OE-wt mice in the test swim, a second experiment was conducted using 0- and 20-min pre-swim
times. The hypoactivity of ␣2C-OE mice compared to
␣2C-OE-wt mice now appeared to be more clear (Figure
2b; F(1,36) = 16.53; P = 0.00025). Interestingly, the ␣2COE mice were clearly less active than their wild-type
controls also in the 0-min pre-swim group (Tukey’s
test; P = 0.017). No significant difference was observed
between ␣2C-KO and ␣2C-KO-wt mice with 0-min preswim (Figure 1). The pre-swim had a clear reducing
effect on the duration of activity in the test swim in all
comparisons (P ⬍ 0.0001), but ANOVA did not indicate pre-swim time × genotype interactions.
Brain biogenic amines and their metabolites
Results from analyses of the biogenic amines and their
metabolite levels in whole brain homogenates after
Figure 1 Development of behavioral despair in the forced
swimming test, measured as the extent of immobility produced by a pre-swim. Bars represent mean ± SEM (n = 8–10)
duration of swimming activity after different preswim times
in ␣2C-AR-deficient mice (␣2C-KO) (hatched bars) and their
wild-type controls (␣2C-KO-wt) (open bars). The ␣2C-KO mice
were more active than ␣2C-KO-wt mice in the test swim after
different pre-swim times (ANOVA; P = 0.0024).
Figure 2 Development of behavioral despair in two separate
forced swimming tests: bars represent mean ± SEM (n = 10)
duration of swimming activity after different pre-swim times
in mice overexpressing the ␣2C-AR (␣2C-OE; closed bars) and
their wild-type controls (␣2C-OE-wt; open bars). (a) The ␣2COE mice were slightly less active than ␣2C-OE-wt mice in the
test swim (P = 0.049). (b) A subsequent test confirmed that
also those ␣2C-OE mice which did not undergo a pre-swim
(the 0-min group) were less active than their wild-type controls (P = 0.017).
The ␣2c-adrenoceptor gene altered mice and stress
J Sallinen et al
Table 1 Effect of acute swim and immobilization on brain monoamines and their metabolites in mice with altered ␣2C-AR
expression
n
NE
MHPG
DA
DOPAC
HVA
5-HT
5-HIAA
KO-wt mice
no stress
swim stress
restraint stress
7 2.95 ± 0.06
0.21 ± 0.017
10.1 ± 0.3
9 2.71 ± 0.08*
0.27 ± 0.019* 9.09 ± 0.8
7 2.60 ± 0.09*** 0.36 ± 0.009*** 9.17 ± 0.5
KO mice
no stress
swim stress
restraint stress
8 2.87 ± 0.07
0.21 ± 0.017
8.80 ± 0.6
10 2.64 ± 0.07*** 0.29 ± 0.020*** 8.98 ± 0.5
10 2.45 ± 0.05*** 0.33 ± 0.020*** 8.03 ± 0.6
1.05 ± 0.05 1.04 ± 0.03
4.16 ± 0.11
2.28 ± 0.10
1.09 ± 0.04 1.26 ± 0.04*** 3.89 ± 0.09
2.57 ± 0.06***
1.18 ± 0.04 1.21 ± 0.04*** 3.56 ± 0.17** 2.48 ± 0.06
OE-wt mice
no stress
swim stress
restraint stress
9 1.38 ± 0.13
10 1.40 ± 0.10
10 1.53 ± 0.14
0.15 ± 0.007
5.58 ± 0.5
0.20 ± 0.007* 6.07 ± 0.5
0.23 ± 0.015*** 5.98 ± 0.5
1.03 ± 0.06 0.65 ± 0.02
2.74 ± 0.23
1.16 ± 0.08 0.88 ± 0.04*** 2.93 ± 0.10
1.25 ± 0.07* 0.87 ± 0.04*** 2.87 ± 0.17
3.10 ± 0.10
3.71 ± 0.13***
3.68 ± 0.17***
OE mice
no stress
swim stress
restraint stress
6 1.45 ± 0.15
8 1.35 ± 0.13
6 1.57 ± 0.03
0.15 ± 0.011
6.33 ± 0.6
0.19 ± 0.011* 6.05 ± 0.7
0.26 ± 0.015*** 7.11 ± 0.3
1.11 ± 0.07 0.82 ± 0.07†
1.13 ± 0.08 0.90 ± 0.06
1.34 ± 0.04**1.01 ± 0.03*
3.43 ± 0.25
3.43 ± 0.20
3.54 ± 0.23
1.11 ± 0.07 1.13 ± 0.04
4.39 ± 0.12
1.21 ± 0.09 1.35 ± 0.05*** 4.02 ± 0.13
1.30 ± 0.04* 1.41 ± 0.04*** 3.93 ± 0.10**
2.89 ± 0.23
2.78 ± 0.21
3.00 ± 0.07
2.36 ± 0.10
2.84 ± 0.13*
2.93 ± 0.10**
Asterisks refer to significant differences in t-tests between no stress and stress groups of the same genotype (*P ⬍ 0.05; **P
⬍ 0.01; ***P ⬍ 0.001). + Indicates significant difference between genotypes in corresponding treatment groups (P ⬍ 0.05). The
values are in nmol g−1 brain tissue as means + SEM.
NE, norepinephrine; MHPG, 3-methoxyphenylglycol; DA, dopamine, DOPAC, 3-4-dihydroxyphenylacetic acid; HVA, homovanillic acid; 5-HT, 5-hydroxytryptamine ie serotonin; 5-HIAA, 5-hydroxyindoleacetic acid.
acute and repeated stress exposures are shown in
Tables 1 and 2. Compared to corresponding genotype
groups without stress, the acute swim-stress increased
the NE metabolite MHPG levels in all mice, with no
differences between mutant and respective wild-type
mice. Also the concentrations of the DA and 5-HT
metabolites, HVA and 5-HIAA, were clearly elevated
after swim-stress in ␣2C-KO, ␣2C-KO-wt, and ␣2C-OE-wt
mice, but only minor, if any, elevations in HVA and 5HIAA were observed in the ␣2C-OE group (Table 1). The
reductions of NE and 5-HT in the ␣2C-KO and ␣2C-KOTable 2
wt mice, and the increases of MHPG levels in all studied mice were more pronounced after acute immobilization than after swim-stress, indicating that the immobilization caused a more severe stress reaction than
the swim.
Repeated stress again resulted in similar MHPG
responses in mutant and wild-type mice. Attenuated
HVA and 5-HIAA responses, similar to those observed
after acute stress in ␣2C-OE mice, were again seen in the
␣2C-OE group, and the HVA and 5-HIAA levels were
distinctly elevated in the other genotype groups (Table
Effect of repeated immobilization on brain monoamines and their metabolites in mice with altered ␣2C-AR expression
n
NE
MHPG
DA
DOPAC
HVA
5-HT
5-HIAA
KO-wt mice
no stress
stress
7 3.01 ± 0.11
17 2.79 ± 0.07
0.21 ± 0.010
9.87 ± 0.30
0.35 ± 0.017*** 10.3 ± 0.24
1.11 ± 0.15 1.13 ± 0.12
1.58 ± 0.14 1.74 ± 0.07**
4.56 ± 0.27
4.07 ± 0.20
KO mice
no stress
stress
8 3.00 ± 0.06
0.23 ± 0.013
9.92 ± 0.21
16 2.62 ± 0.06*** 0.34 ± 0.013*** 9.67 ± 0.25
1.11 ± 0.13 1.05 ± 0.13
1.54 ± 0.14 1.64 ± 0.09**
4.82 ± 0.25
2.18 ± 0.06
3.84 ± 0.18** 2.94 ± 0.10***
2.41 ± 0.11
2.99 ± 0.09**
OE-wt mice
no stress
stress
6 1.32 ± 0.20
8 1.42 ± 0.14
0.16 ± 0.014
5.71 ± 0.77
0.26 ± 0.012*** 6.45 ± 0.57
1.05 ± 0.09 0.63 ± 0.07
1.21 ± 0.08 0.88 ± 0.07*
2.66 ± 0.20
3.23 ± 0.26
3.35 ± 0.15
3.95 ± 0.12**
OE mice
no stress
stress
7 1.37 ± 0.20
8 1.20 ± 0.11
0.17 ± 0.012
6.46 ± 0.71
0.23 ± 0.011*** 5.98 ± 0.45
1.10 ± 0.08 0.71 ± 0.05
1.16 ± 0.09 0.78 ± 0.03
2.69 ± 0.29
2.81 ± 0.17
3.39 ± 0.12
3.67 ± 0.09
Asterisks refer to significant differences in t-tests between no stress and stress groups of the same genotype (*P ⬍ 0.05; **P
⬍ 0.01; ***P ⬍ 0.001). The values are in nmol g−1 brain tissue as means + SEM.
Abbreviations as in Table 1.
447
The ␣2c-adrenoceptor gene altered mice and stress
J Sallinen et al
448
2). An overall ANOVA, performed over all acute stress
challenge groups, and including non-stressed animals,
indicated that ␣2C-KO-mice had lower brain 5-HT, 5HIAA, DOPAC and HVA concentrations (P ⬍ 0.05)
than ␣2C-KO-wt mice, and that ␣2C-OE mice had higher
brain HVA concentrations than ␣2C-OE-wt mice. In
addition, compared to ␣2C-OE-wt mice, trends without
statistical significance for higher levels of 5-HT, 5HIAA and DOPAC were observed in ␣2C-OE mice without stress, and after swim stress these measures were
lower than in the ␣2C-OE-wt mice. After repeated stress
the differences between the genotypes were smaller,
but the trends in the means were again similar.
Analysis of c-fos and jun-B mRNA expression in
brain by in situ hybridization
The control (non-stressed) ␣2C-KO mice had higher levels of c-fos and junB mRNAs in cortex and hippocampus than ␣2C-KO-wt (Figure 3). In the hippocampus, the
expression of junB mRNA was significantly elevated in
the CA1 and CA2 fields of Ammon’s horn of nonstressed ␣2C-KO animals when compared to nonstressed ␣2C-KO-wt mice, whereas in the rest of the hippocampal fields there were no significant differences
between these two groups (Figure 4). There were no
differences in the levels of c-fos and junB mRNA
expression in cortex, hippocampus or striatum
between non-stressed ␣2C-OE and ␣2C-OE-wt mice.
After swim and restraint stresses, clear increases were
observed in c-fos mRNA levels in cortex and hippocampus as well as junB mRNA levels in cortex in all
strains of mice (P ⬍ 0.001). In addition, the differences
in the mRNA levels between ␣2C-KO and ␣2C-KO-wt
mice, which could be seen in non-stressed animals,
almost completely disappeared after the stresses.
Plasma corticosterone
The plasma corticosterone levels were elevated 4–12fold after stress, as expected. No differences were
observed in the baseline corticosterone concentrations
between corresponding genotypes (Figure 5a, b). ␣2CKO-mice had constantly lower corticosterone concentrations than ␣2C-KO-wt mice after different stressors
(Figure 5a; 2-way ANOVA, genotype difference: F(1,96)
= 14.9, P = 0.00021). The difference was most clear after
repeated stress when the mean concentrations in ␣2CKO and ␣2C-KO-wt mice were 220 ± 22 and 298 ± 17
ng ml−1, respectively (Tukey’s test, P = 0.00078). Compared to ␣2C-KO and ␣2C-KO-wt mice, the corticosterone responses to acute stress were more intense in
␣2C-OE and ␣2C-OE-wt mice. The ␣2C-OE mice had
smaller corticosterone responses than their wild-type
controls after acute stressors (F(1,46) = 4.25; P = 0.045),
like the ␣2C-KO mice. Interestingly, after repeated
stress, the difference between ␣2C-OE and ␣2C-OE-wt
mice was opposite to that observed after acute stress
(Figure 5); ie ␣2C-OE mice had higher corticosterone
levels (230 ± 13 ng ml−1) than ␣2C-OE-wt mice (178 ±
9 ng ml−1) (Tukey’s test, P = 0.0013; interaction
between ␣2C-OE and ␣2C-OE-wt genotypes vs acute and
repeated stress: F(2, 66) = 4.86; P = 0.011).
Figure 3 Quantitative results showing the effects of swim
and restraint stresses on c-fos and junB mRNA levels in cortex, hippocampus and striatum of ␣2C-KO mice and their controls. The values represent means ± SEM from four animals.
*P ⬍ 0.001 compared to corresponding ␣2C-KO-wt group.
The ␣2c-adrenoceptor gene altered mice and stress
J Sallinen et al
449
Figure 4 Quantitative evaluation of junB mRNA expression
in different hippocampal fields of non-stressed ␣2C-KO
(closed bars) and ␣2C-KO-wt mice (open bars). The values represent means ± SEM from four animals. *P ⬍ 0.001 compared
to corresponding area of control mice. CA1, CA2, CA3, CA4,
fields of Ammon’s horn; DG, dentate gyrus.
Discussion
␣2C-AR gene disruption (␣2C-KO) was associated with
increased duration of activity in the FST and attenuated increases in plasma corticosterone after different
stressors. In addition, markedly enhanced c-fos and
junB mRNA expression levels were observed in several
brain regions in non-stressed ␣2C-KO mice. Compared
to ␣2C-KO mice, overexpression of ␣2C-ARs had
opposite effects on behavior in the FST and on the
corticosterone responses after repeated immobilization.
Stress-induced elevations in brain HVA and 5-HIAA
concentrations after both acute and repeated stress
were attenuated in ␣2C-OE mice, but HVA and 5-HIAA
were increased similarly in ␣2C-KO and ␣2C-KO-wt
mice. Overall analysis of the brain monoamine and
metabolite levels of both non-stressed and stressed animals indicated that ␣2C-KO mice had significantly
lower 5-HT, 5-HIAA, DOPAC and HVA concentrations
than their controls, whereas ␣2C-OE-mice showed
opposite trends in the corresponding measures; HVA
levels were also statistically significantly higher in ␣2COE than in ␣2C-OE-wt mice. In addition, the baseline
levels of the two genetically different wild-type strains
were different. These differences may mask some
stress-induced changes and complicate the interpretation of the stress effects. Interestingly, brain NE turnover was not altered by the ␣2C-AR gene manipulations
according to brain NE and MHPG content, and the
stress-induced changes in NE and MHPG concentrations were similar in both mutant strains and their
corresponding controls. The current brain monoamine
and metabolite results from non-stressed and stressexposed mice actually confirm our previous suggestion, based on studies employing a subtype non-selec-
Figure 5 The plasma corticosterone levels (ng ml−1) in (a)
␣2C-KO mice (hatched bars) and in (b) ␣2C-OE mice (closed
bars) and in their respective wild-type controls (open bars)
without stress and after different stressors. Compared to similarly treated wild-type mice, the stress-induced elevations in
plasma corticosterone were attenuated in ␣2C-KO mice after
acute and repeated stress (P ⬍ 0.001), whereas the repeated
stress increased the corticosterone levels in ␣2C-OE mice significantly more than in their controls (P ⬍ 0.001).
tive ␣2-AR agonist, that altered ␣2C-AR expression
modulates brain DA and 5-HT (but not NE) metabolism.21 The underlying mechanisms are, however,
not clear.
FST was used here to test the association between
␣2C-AR function and stress-related behavioral despair.
The observed statistically significant differences in the
response patterns of ␣2C-KO and ␣2C-OE mice and their
corresponding wild-type controls were opposite; this
strengthens the significance of the statistical comparisons, since the two mutant and wt strain pairs have
genetically altered ␣2C-AR expression but are otherwise
independent and have differences in their baseline
behaviors due to their markedly different strain backgrounds. In addition, the locomotor activity of the mice
in the repeated stress groups was determined repeatedly between the immobilizations and was found not
to differ from the activity of non-stressed control mice
representing the corresponding genotype groups (data
not shown). This suggests that swimming activity after
The ␣2c-adrenoceptor gene altered mice and stress
J Sallinen et al
450
stress could be distinguished from general locomotor
activity which was not altered by the mutations or by
repeated stress.
FST is extensively used to evaluate the potential
efficacy of prospective antidepressants because of its
predictive validity and sensitivity for most of the current antidepressants.28,29,39 Complex functional
relationships have been suggested in the FST between
␣2-ARs and the brain serotonergic system on the basis
that ␣2-agonists and -antagonists interact with antidepressants in the FST.40,41 However, selective ␣2-antagonists or -agonists have no clear independent effects in
the FST, and the ␣2-agonist-induced immobility in the
FST may result from the sedative effect of clonidine.28,40,41,42
The brain noradrenergic response to stress, mirrored
by increased MHPG concentrations, was not altered by
the gene manipulations. One explanation for this is
that ␣2A-AR is the predominant ␣2-AR subtype in the
locus coeruleus (LC), which is a crucial regulatory site
in the activation of the brain noradrenergic system in
the response to stress.30 Interestingly, ␣2-AR antagonists are known to increase and ␣2-AR agonists to
decrease arousal, by altering the firing rate of LC neurons and the release of NE in the CNS. Since the sedative effects of ␣2-AR agonists appear to be mediated by
␣2A-AR, it is possible that this subtype is also responsible for the effects of ␣2-AR antagonists on arousal. In
addition, the traditional ␣2-AR agonist clonidine and
the antagonist yohimbine have anxiolytic and anxiogenic effects, respectively. These properties, if ␣2A-ARmediated, would antagonise the proposed ␣2C-ARmediated effects in the FST, when subtype non-selective ␣2-AR agonists or antagonists are used. Furthermore, binding of many purportedly selective ␣2-AR
antagonists to other receptors, such as 5-HT1A-receptors, may be an additional confounding factor.43 We
speculate that subtype non-selective ␣2-AR antagonists
could induce excessive arousal via the LC, perhaps
leading to feelings of anxiety. The current results suggest that a selective ␣2C-AR antagonist might have antidepressant-like actions without prominent NE-releasing and anxiogenic properties.
Since the brain monoamine levels were determined
from whole brain homogenates, the current results do
not allow one to exactly determine any anatomical
structures that could be responsible for the observed
DA and 5-HT changes. Nevertheless, ␣2C-AR-dependent modulation of nigrostriatal and mesocortical
dopaminergic pathways may in part explain the
observed effects on brain DA metabolism and swimming activity, given that ␣2C-ARs are predominantly
expressed in the striatum.7–12 Another candidate structure is the hippocampus, which is a very stress-sensitive region and receives intense NE innervation. It may
be involved in the 5-HT changes, because a wealth of
evidence points to complex interactions between ␣2ARs, and brain NE and 5-HT systems, as well as with
the actions of antidepressants in the hippocampus, as
highlighted by Mongeau and colleagues.44 The ␣2C-ARs
are abundantly expressed in the CA1 field of Ammon’s
horn,7–9,21 which is in accordance with the increased
junB mRNA expression in ␣2C-KO mice, observed
selectively in the CA1 and CA2 subregions of the hippocampus (Figure 4). Mongeau and coworkers45 have
also demonstrated that hippocampal ␣2-ARs located in
the 5-HT terminals are tonically activated by endogenous NE; lack of this tonic neuronal inhibition due to a
lack of functional ␣2C-AR could also explain the 5-HT
differences and increased hippocampal c-fos and junB
mRNA levels in non-stressed ␣2C-KO mice. Further, the
␣2-AR antagonist yohimbine has been shown to
increase c-fos mRNA expression in brain,46 which is in
accordance with the current results. Finally, it is also
possible that the observed net effects in DA and 5-HT
metabolism are secondary and result from changes in
the balance of additional neurotransmitter systems,
such as glutamate or GABA, and/or changes in the tone
of various neural networks.
It is generally accepted that stress and depression are
linked, although their causal relationships are not
clear. It has been proposed that stress-induced
increases in glucocorticoid levels are detrimental to
some brain structures, such as the hippocampus,
which could lead to atrophy, or even death, of vulnerable neurons and thus lead to impaired function, and
depression.47 On the other hand, hypercortisolism,
consistently observed in depressed patients, has also
been recently related to an abnormal adrenocortical
feedback modulation occurring at a central site, such
as the hippocampus, rather than at the level of the pituitary.48,49 From this point of view, alterations in the
regulation of the corticosterone system by the ␣2C-AR
gene manipulations, either directly or indirectly
through changes in the brain monoamine balance, may
also account for the behavioral responses observed in
the FST. This is supported by findings demonstrating
that the corticosterone synthesis inhibitor, metyrapone,
increases the duration of activity in the FST.50 Moreover, glucocorticoid receptor antisense oligonucleotides have shown anti-immobility effects in FST when
injected into the dentate gyrus of the hippocampus,51
a region where the ␣2C-AR seems to be the only
expressed ␣2-AR subtype in mice.21
It is possible that overexpression of ␣2C-AR leads to
a sustained activation of ␣2C-AR mediated circuitries,
which could lead to an equivalent of a constitutively
stressful state in the ␣2C-OE mice. This could explain
why those ␣2C-OE mice which were forced to swim for
the first time (Figure 2) were already less active than
their wild-type controls (while no difference was seen
in their undisturbed locomotor activity).21 It could also
explain the slightly increased baseline concentrations
of brain HVA and 5-HIAA and the attenuated elevations of HVA, 5-HIAA and corticosterone levels after
acute stress as well as the decreased mRNA levels.
More speculatively, this hypothesis is in line with a
recent study demonstrating increased ␣2-AR radioligand binding to a non-␣2A-AR subtype in brains of
unmedicated depressed suicides.52
In conclusion, our results demonstrate how excess
function of a single adrenergic receptor subtype, or loss
The ␣2c-adrenoceptor gene altered mice and stress
J Sallinen et al
of its function, is capable of causing substantial alterations in the stress-responsivity of the corticosterone
system and the brain monoamine balance. The current
study therefore emphasizes the importance of singlegene mutations in responses to environmental stressors, which may contribute to the manifestation of various neuropsychiatric disorders. The FST results also
suggest that a selective ␣2C-AR antagonist would have
therapeutic potential as an antidepressant.
Acknowledgements
We thank Dr Richard Link for generating the strains of
mice used in these studies, and Ms Helena Hakala, Ms
Elina Kahra and Ms Päivi Saikkonen for skillful technical assistance.
References
1 Ruffolo RR, Nichols AJ, Hieble JP. Functions mediated by alpha-2
adrenergic receptors. In: Limbird LE (ed). The Alpha-2 Adrenergic
Receptors. Humana Press: New York, 1988, pp 187–280.
2 Ruffolo RR, Nichols AJ, Stadel JM, Hieble PJ. Pharmacologic and
therapeutic applications of alpha2-adrenoceptor subtypes. Annu
Rev Pharmacol Toxicol 1993; 32: 243–279.
3 Foote SL, Aston-Jones GS. Pharmacology and physiology of central
noradrenergic systems. In: Bloom FE, Kupfer DJ (eds). Psychopharmacology: The Fourth Generation of Progress. Raven Press: New
York, 1995, pp 335–345.
4 Robbins TW, Everitt BJ. Central norepinephrine neurons and
behavior. In: Bloom FE, Kupfer DJ (eds). Psychopharmacology. The
Fourth Generation of Progress. Raven Press: New York, 1995, pp
363–372.
5 Flügge G. Alterations in the central nervous ␣2-adrenoceptor system under chronic psychosocial stress. Neuroscience 1996; 75:
187–196.
6 Bylund DB, Eikenberg DC, Hieble JP. International union of pharmacology nomenclature of adrenoceptors. Pharmacol Rev 1994; 46:
121–136.
7 Nicholas AP, Pieribone V, Hökfelt T. Distributions of mRNAs for
alpha-2 adrenergic receptor subtypes in rat brain: an in situ
hybridization study. J Comp Neurology 1993; 328: 575–594.
8 Nicholas AP, Hökfelt T, Pieribone VA. The distribution and significance of CNS adrenoceptors examined with in situ hybridization. Trends Pharmacol Sci 1996; 17: 245–255.
9 Scheinin M, Lomasney JW, Hayden-Hixson DM, Schambra UB,
Caron MG, Lefkowitz RJ et al. Distribution of alpha2-adrenergic
receptor subtype gene expression in rat brain. Brain Res Mol Brain
Res 1994; 21: 133–149.
10 Rosin DL, Talley EM, Lee A, Stornetta RL, Gaylinn BD, Guyenet
PG et al. Distribution of ␣2C-adrenergic receptor-like immunoreactivity in the rat central nervous system. J Comp Neurology 1996;
372: 135–165.
11 Talley EM, Rosin DL, Lee A, Guyenet PG, Lynch KR. Distribution
of ␣2C-adrenergic receptor-like immunoreactivity in the rat central
nervous system. J Comp Neurology 1996; 372: 111–134.
12 Wang R, MacMillan LB, Fremeau Jr RT, Magnuson MA, Lindner J,
Limbird LE. Expression of ␣2-adrenergic receptor subtypes in the
mouse brain: evaluation of spatial and temporal information
imparted by 3 kb of 5⬘ regulatory sequence for the ␣2A-AR-receptor
gene in transgenic animals. Neuroscience 1996; 74: 199–218.
13 von Zastrow M, Link R, Daunt D, Barsh G, Kobilka B. Subtypespecific differences in the intracellular sorting of G protein-coupled
receptors. J Biol Chem 1993; 268: 763–766.
14 Daunt DA, Hurt C, Hein L, Kallio J, Feng F, Kobilka BK. Subtypespecific intracellular trafficking of ␣2-adrenergic receptors. Mol
Pharmacol 1997; 51: 711–720.
15 MacDonald E, Kobilka BK, Scheinin M. Gene targeting—homing in
on ␣2-adrenoceptor-subtype function. Trends Pharmacol Sci 1997;
18: 211–219.
16 MacMillan LB, Hein L, Smith MS, Piascik MT, Limbird LE. Central
hypotensive effects of the ␣2A adrenergic receptor subtype. Science
1996; 273: 801–803.
17 Link RE, Desai K, Hein L, Stevens ME, Chruscinski A, Bernstein D
et al. Cardiovascular regulation in mice lacking ␣2-adrenergic
receptor subtypes b and c. Science 1996; 273: 803–805.
18 Hunter JC, Fontana DJ, Hedley LR, Jasper JR, Lewis R, Link RE
et al. The relative contribution of ␣2-adrenoceptor subtypes to the
antinociceptive action of dexmedetomidine and clonidine in rodent models of acute and chronic pain. Br J Pharmacol 1997; 122:
1339–1344.
19 Lakhlani PP, MacMillan LB, Guo TZ, McCool BA, Lovinger DM,
Maze M et al. Substitution of a mutant ␣2a-adrenergic receptor via
‘hit and run’ gene targeting reveals the role of this subtype in sedative, analgesic, and anesthetic-sparing responses in vivo. Proc Natl
Acad Sci USA 1997; 94: 9950–9955.
20 Stone LS, MacMillan LB, Kitto KF, Limbird LE, Wilcox GL. The
␣2a adrenergic receptor subtype mediates spinal analgesia evoked
by ␣2 agonists and is necessary for spinal adrenergic-opioid synergy. J Neurosci 1997; 17: 7157–7165.
21 Sallinen J, Link RE, Haapalinna A, Viitamaa T, Kulatunga M,
Sjöholm B et al. Genetic alteration of ␣2C-adrenoceptor expression
in mice: influence on locomotor, hypothermic, and neurochemical
effects of dexmedetomidine, a subtype-nonselective ␣2-adrenoceptor agonist. Mol Pharmacol 1997; 51: 36–46.
22 Sallinen J, Haapalinna A, Viitamaa T, Kobilka BK, Scheinin M.
Adrenergic ␣2C-receptors modulate the acoustic startle reflex, prepulse inhibition, and aggression in mice. J Neurosci 1998; 18:
3035–3042.
23 Sallinen J, Haapalinna A, Viitamaa T, Kobilka BK, Scheinin M. dAmphetamine and l-5-hydroxytryptophan-induced behaviours in
mice with genetically-altered expression of the ␣2C-adrenergic
receptor subtype. Neuroscience 1998; 86: 959–965.
24 Porsolt RD, LePichon M, Jalfre M. Depression: a new animal model
sensitive to antidepressant treatments. Nature 1977; 266: 730–732.
25 Link RE, Stevens MS, Kulatunga M, Scheinin M, Barsh GS, Kobilka
BK. Targeted inactivation of the gene encoding the mouse ␣2C-adrenoceptor homolog. Mol Pharmacol 1995; 48: 48–55.
26 Porsolt RD. Behavioural despair. In: Enna SJ, Malick JB, Richelson
E (eds). Antidepressants: Neurochemical, Behavioural and Clinical
Perspectives. Raven Press: New York, 1981, pp 121–139.
27 Willner P. The validity of animal models of depression. Psychopharmacology 1984; 83: 1–16.
28 Borsini F, Meli A. Is the forced swimming test a suitable model for
revealing antidepressant activity? Psychopharmacology 1988; 94:
147–160.
29 Geyer MA, Markou A. Animal models of psychiatric disorders. In:
Bloom FE, Kupfer DJ (eds). Psychopharmacology: The Fourth Generation of Progress. Raven Press: New York, 1995, pp 787–798.
30 Valentino RJ, Curtis AL, Page ME, Pavcovich LA, Florin-Lechner
SM. Activation of the locus ceruleus brain noradrenergic system
during stress: circuitry, consequences, and regulation. In: Goldstein
DS, Eisenhofer G, McCarty R (eds). Advances in Pharmacology.
Academic Press: San Diego, 1998, pp 781–784.
31 Nagy A, Rossant J, Nagy R, Abramow-Newerly W, Roder JC. Derivation of completely cell culture-derived mice from early-passage
embryonic stem cells. Proc Natl Acad Sci USA 1993; 90: 8424–
8428.
32 Overbeek PA, Aguilar-Cordova E, Hanten G, Schaffner DL, Patel P,
Lebovitz RM et al. Coinjection strategy for visual identification of
transgenic mice. Transgenic Res 1991; 1: 31–37.
33 Mefford I. Applications of high performance liquid chromatography with electrochemical detection to neurochemical analysis:
measurement of catecholamines, serotonin and metabolites in rat
brain. J Neurosci Meth 1981; 3: 207–224.
34 MacDonald E, Scheinin H, Scheinin M. Behavioural and neurochemical effects of medetomidine, a novel veterinary sedative. Eur
J Pharmacol 1988; 158: 119–127.
35 Ojala-Karlsson P, Scheinin M. Improved high performance liquid
chromatographic method for the routine determination of unconjungated 3-methoxy-4-hydroxyphenylethyleneglycol in human
plasma using solid-phase extraction and electrochemical detection.
J Chromatogr 1991; 565: 131–139.
36 Curran T, Gordon MB, Rubino KL, Sambucetti LC. Isolation and
451
The ␣2c-adrenoceptor gene altered mice and stress
J Sallinen et al
452
37
38
39
40
41
42
43
44
characterization of the c-fos (rat) cDNA and analysis of post-translational modification in vitro. Oncogene 1987; 2: 79–84.
Kononen J, Pelto-Huikko M. Multiple consecutive hybridizations
in radioactive in situ hybridization. Trends Genetic Technical Tips
Online 1997; URL: http://tto.trends.com/
Ryseck R-P, Hirai SI, Yaniv M, Bravo R. Transcriptional activation
of c-jun during the G0/G1 transition in mouse fibroblasts. Nature
1988; 334: 535–537.
Detke MJ, Johnson I, Lucki I. Acute and chronic antidepressant
drug treatment in the rat forced swimming test model of
depression. Exp Clin Psychopharmacol 1997; 5: 107–112.
Cervo L, Grignaschi G, Samanin R. Alpha 2-adrenoceptor blockade
prevents the effect of desipramine in the forced swim test. Eur J
Pharmacol 1990; 175: 301–307.
Bourin M, Colomber M-C, Malinge M, Bradwejn J. Clonidine as a
sensitizing agent in the forced swimming test for revealing antidepressant activity. J Psych Neurosci 1991; 16: 199–203.
Kauppila T, Tanila H, Carlson S, Taira T. Effects of atipamezole,
a novel ␣2-adrenoceptor antagonist, in open-field, plus-maze, two
compartment exploratory, and forced swimming tests in the rat.
Eur J Pharmacol 1991; 205: 177–182.
Haapalinna A, Viitamaa T, MacDonald E, Savola J-M, Tuomisto L,
Virtanen R et al. Evaluation of the effects of a specific ␣2-adrenoceptor antagonist, atipamezole, on ␣1- and ␣2-adrenoceptor subtype
binding, brain neurochemistry and behaviour in comparison with
yohimbine. Naunyn-Schmedeberg’s Arch Pharmacol 1997; 356:
570–582.
Mongeau R, Blier P, de Montigny C. The serotonergic and norad-
45
46
47
48
49
50
51
52
renergic systems of the hippocampus: their interactions and the
effects of antidepressant treatments. Brain Res Rev 1997; 23:
145–195.
Mongeau R, Blier P, de Montigny C. In vivo electrophysiological
evidence for tonic activation by endogenous noradrenaline of ␣2adrenoceptors on 5-hydroxytryptamine terminals in the rat hippocampus. Naunyn-Schmedeberg’s Arch Pharmacol 1993; 347:
266–272.
Gubits RM, Smith TM, Fairhurst JL, Yu H. Adrenergic receptors
mediate changes in c-fos mRNA levels in brain. Mol Brain Res
1989; 6: 39–45.
Duman RS, Heninger GR, Nestler EJ. A molecular and cellular
theory of depression. Arch Gen Psychiatry 1997; 54: 597–608.
Young EA, Haskett RF, Murphy-Weinberg V, Watson SJ, Akil H.
Loss of glucocorticoid fast feedback in depression. Arch Gen Psychiatry 1991; 48: 693–699.
Akil HA, Morano MI. Stress. In: Bloom RE, Kupfer DJ (eds). Psychopharmacology: The Fourth Generation of Progress. Raven Press:
New York, 1995, pp 773–786.
Baez M, Volosin M. Corticosterone influences forced swim-induced
immobility. Pharmacol Biochem Behav 1994; 49: 729–736.
Korte SM, De Kloet ER, Buwalda B, Bouman SD, Bohus B. Antisense to the glucocorticoid receptor in hippocampal dentate gyrus
reduces immobility in forced swim test. Eur J Pharmacol 1996; 301:
19–25.
De Paermentier F, Mauger JM, Lowther S, Crompton MR, Katona
CLE, Horton RW. Brain ␣-adrenoceptors in depressed suicides.
Brain Res 1997; 757: 60–68.