BRAIN,

BEHAVIOR,
and IMMUNITY
Brain, Behavior, and Immunity 18 (2004) 81–90
www.elsevier.com/locate/ybrbi

Impaired T-cell dependent humoral response and its relationship

with T lymphocyte sensitivity to stress hormones in a chronic
mild stress model of depression
D.M. Silberman,a V. Ayelli-Edgar,a M. Zorrilla-Zubilete,b L.M. Zieher,b
and A.M. Genaroa,*
a
CEFYBO-CONICET, Serrano 669 3rd floor, CP 1414-Buenos Aires, Argentina
b
1ra C

atedra de Farmacolog
ıa, Facultad de Medicina-UBA, Paraguay 2155, CP 1121-Buenos Aires, Argentina
Received 22 January 2003; received in revised form 13 May 2003; accepted 13 June 2003

Abstract

The humoral response and the role of catecholamines and corticosterone were analyzed in a chronic mild stress (CMS) model of
depression. Mice subjected for more than 6 weeks to CMS showed a significant decrease in T-cell dependent antibody production.
However, T-cell independent humoral response was not altered. Serum corticosterone levels and splenic norepinephrine (NE)
contents showed an early increase but they were not altered after prolonged CMS exposure. Nevertheless, hormonal inhibitory effect
on T lymphocyte reactivity was higher in 6-week CMS mice compared to non-exposed animals. Thus, our results suggest that the
impaired T-cell dependent humoral response in a CMS model of depression is neither related to changes in glucocorticoids nor in
NE levels but is correlated with an increment of T-cell sensitivity to stress hormones. These findings would underlie the involvement
of catecholamines and glucocorticoid lymphocyte receptors in the immune alterations observed in stress and depression.
Ó 2003 Elsevier Inc. All rights reserved.

Keywords: Chronic mild stress; Depression; Corticosterone; Catecholamines; Lymphocyte receptors; Immune response

1. Introduction of depression in mice (Ayelli-Edgar et al., 2002). On the

Over the last years considerable evidence showed a
dynamic interaction between the brain and the immune
system (Miller, 1998). Many findings have suggested the
association between the clinical status of depression and
an immunological dysfunction (Irwin, 1995; Ravindran
et al., 1998). Major depressive disorder has been associ-
ated with a suppression of several functions of the
immune system, such as a reduction of lymphocyte
proliferation in response to mitogens (Herbert and
Cohen, 1993; Leonard, 1995), impaired neutrophil
phagocytosis and suppression of the cytotoxic activity of
natural-killer (NK) cells (Irwin, 1995). In fact, we have
previously described a decrease in mitogen-induced
T-cell proliferation in a chronic mild stress (CMS) model
*
Corresponding author. Fax: +54-11-4856-2751.
E-mail address: genaro@cefybo.edu.ar (A.M. Genaro).
other hand, other studies demonstrated some evidence of
immunocyte-mediated autoimmune reaction such as an
increment in the levels of antibodies to phospholipids,
nuclear and Epstein–Barr Virus; complement proteins
C3 and C4 and acute-phase proteins (Maes et al., 1995).
Immune abnormalities in depression are difficult to
interpret due to the heterogeneity of the patients sample
and to variables such as age, sex, and hospitalization
status (Maes et al., 1997). Studies performed in an ani-
mal model of depression allow getting an homogenous
and consistent array of results. Animal models of de-
pression involve the application of various stressors and
induce a state similar to the one observed in human
depression. Depression is a complex psychological dis-
order, which is defined by DSM-IV (American Psychi-
atric Association, 1994) by two core symptoms: loss of
interest or pleasure (anhedonia) and depressed mood.
Anhedonia can be modeled in animals, but not

0889-1591/$ - see front matter Ó 2003 Elsevier Inc. All rights reserved.
doi:10.1016/S0889-1591(03)00109-0
82 D.M. Silberman et al. / Brain, Behavior, and Immunity 18 (2004) 81–90
depressed mood. It follows that anhedonia is the es-
sential symptom to be reproduced in an attempt to
simulate a depression-like state. Studies in rats and in
mice have shown that chronic sequential exposure to
mild unpredictable stressors (CMS model) reduces the
consumption of, and the preference for, palatable sweet
solutions (Monleon et al., 1995; Willner et al. (1992)).
Moreover, the CMS model also offers a realistic simu-
lation of the etiology of depression, as chronic low-
grade stress is an important factor in the onset of the
illness (Overstreet and Steiner, 1998).
The adaptive response to stressors involves the acti-
vation of the hypothalamic–pituitary–adrenal (HPA)
axis and of the sympathetic nervous system (SNS)
(Servatius et al., 2000). Products of both of these sys-
tems (e.g., glucocorticoids and catecholamines) are able
to modulate the activity of several immune effector cells
in a directed way (Chrousos, 1995; Elenkov et al., 2000).
This occurs when neurohormones and neurotransmit-
ters bind to their receptors on lymphocytes (Elenkov et
al., 2000; Roszman and Brooks, 1997). Furthermore, it
is assumed that both pathways (HPA axis and the SNS)
probably act in a cooperative way in order to maintain
homeostasis.
The aim of the present work was to study the inter-
action between SNS–HPA-immune system in an animal
model of depression. For this purpose, BALB/c mice
were subjected to the above-mentioned CMS model
from 1 to 10 weeks. The effect of CMS exposure on
behavioral parameters, antibody formation to T-cell
dependent and independent antigens, catecholamines
and corticosterone levels and their effect on lymphocyte
reactivity were analyzed.
2. Materials and methods
2.1. Drugs
Phytohemagglutinin (PHA), epinephrine, norepi-
nephrine, and corticosterone were purchased from Sigma
Chemical. [3 H]Thymidine (20 Ci/mmol) was purchased
from New England Nuclear (NEN), Life Science prod-
ucts. Other materials were from standard commercial
sources.
2.2. Mice
Inbred female BALB/c mice were purchased from the
Instituto Nacional de Tecnolog
ıa Agropecuaria (INTA).
Inbred female C3H mice were obtained from the Insti-
tuto de Oncolog
ıa ‘‘A.H. Roffo.’’ All animals were be-
tween 60 and 100 days of age. The animals were cared in
accordance with the principles and guidelines of the
Guide for the Care and Use of Laboratory Animals, US
National Research Council, 1996.
2.3. Chronic mild stress model
Three experimental groups of 60 animals were used
through the experiments. Animals were singly housed
and maintained on a 12-h light/dark cycle under con-
trolled temperatures (18–22°). Except as described be-
low, food and water were freely available. First, animals
were trained to consume a 1% sucrose solution. After a
2-day period of adaptation, animals were distributed
into two subgroups ðn ¼ 30Þ and sucrose solution intake
baseline tests were performed (three tests per week) over
15 days for all subjects. These tests involved a 3-h period
of food and water deprivation (from 11:00 AM to 2:00
PM) followed by the offering of a sucrose solution for
1 h. Intake was determined by weighing the bottles
containing the sucrose solution at the beginning and at
the end of each test. Sucrose solution consumption per
gram of body weight was calculated for each subject.
After this phase (2 weeks) one group was housed in
normal conditions (non-exposed animals) and the other
group was subjected throughout the experiment to
chronic unpredictable mild stress (CMS-exposed ani-
mals). The stress scheme was slightly modified from that
previously used in rats by Willner et al. (1992) and in
mice by Monleon et al. (1995), and consisted of: 16-h
period of water deprivation; two periods of continuous
overnight illumination; two periods (7 and 17 h) of 45°
cage tilt; one 17-h period in a soiled cage (100 ml water
in sawdust bedding); one period (8 h) of food depriva-
tion; one 17-h period of paired housing (animals are
always housed in the same pairs, but the location al-
ternates between the home cages of each member of the
pair). All the individual stressors used had been classi-
fied as ‘‘mild’’ according to the Animals (Scientific
Procedures) Act of 1986 (UK legislation). The stressors
were scheduled throughout the week, in a similar man-
ner to that previously described (Monleon et al., 1995;
Willner et al., 1992) for 6 weeks (n ¼ 46) or for 1–10
weeks (n ¼ 44).
2.4. Open field behaviour
Open field test was performed between 4:00 and 6:00
PM. Each animal was placed in the centre of a uni-
formly lit open-field with rounded edges measuring
42
35
15 cm uniformly divided into 30 squares of
7
7 cm. The entire grid was divided into four equal
quadrants. Two observers, unaware of the group to
which each animal belonged, recorded the parameters
for 10 min. Exploratory behaviour involved sniffing and
general orientation to various parts of the test cage.
Exploration that occurred within a given quadrant for
five or more consecutive seconds was defined as ‘‘non-
ambulatory exploration.’’ Exploration that occurred in
two or more quadrants for five consecutive seconds was
defined as ‘‘ambulatory exploration.’’ We also assessed
 D.M. Silberman et al. / Brain, Behavior, and Immunity 18 (2004) 81–90
locomotion, defined as the ‘‘number of lines crossed,’’
and rearing.
2.5. Immunizations
Sheep red blood cells (SRBC) and allogeneic cells
(from C3H mice) were used as immunogens to eval-
uate T-cell dependent humoral response. Dextran
B512 (DxB512) and lipopolysaccharide (LPS) were
used to determinate T-cell independent humoral re-
sponse. For allogeneic response, animals were injected
intradermically (id) with 107 cells in saline. For SRBC
response, BALB/c mice were immunized intraperito-
neally (ip) with 2.5% SBRC in saline. For LPS or
DxB512 immunizations, each BALB/c mouse received
an ip injection of 10 lg of LPS or DxB512 in 0.1 ml
saline. Mice injected with vehicle alone were used as
controls.
2.6. Antibody titers
Animals were sacrificed 1 week after immunization
and blood was collected into plain tubes (without
anticoagulant). Serial dilutions of sera were made and
hemagglutination assays using C3H red cells, SRBC,
LPS-coated SRBC, or DxB512-coated SRBC were
performed according to the method described by
Welsh and Batchelor (1978). Briefly, a 2% cell sus-
pension is made up in 0.25% carboxymethylcellulose in
saline. Antisera are serially diluted with normal mouse
serum to ensure that there is a standard protein con-
centration in all tubes. One volume of 25 ll of red cell
suspension is added to one volume of antiserum di-
lution in a series of round bottom wells cut into
transparent plastic block. The reagents are incubated
for 90 min at 37 °C and results are read macroscopi-
cally: agglutinated red cells appear as a carpet spread
over the entire bottom of the wall, whereas non-
agglutinated cells form a small tight button. Antibody
titers were determined as the higher dilution with
positive reaction.
2.7. Cell suspensions and culture conditions
Lymphoid cell suspensions from control and CMS
mice spleens were obtained as previously described
(Ayelli-Edgar et al., 2002). Briefly, spleens were removed
and disrupted through a 1 mm metal mesh and cell
suspension was filtered through a 10 lm nylon mesh.
The suspension was depleted of non-lymphoid cells after
centrifugation over Ficoll/Hypaque. After three washes
in RPMI 1640, cells were re-suspended in RPMI 1640
supplemented with 10% of batched-tested non-stimula-
tory fetal calf serum, 2 mM glutamine, 100 U/ml of
penicillin, 100 lg/ml of streptomycin, and 50 lM
b-mercaptoethanol.
83
2.8. Mitogen assay
Proliferation was determined by culturing 2
105
cells per well in 96-well plates in 0.150 ml triplicate
aliquots in supplemented medium. Aliquots of 50 ll of
PHA were added to the microcultures to yield a final
concentration of 50 lg/ml. In control cultures, stimu-
lants were replaced by 50 ll of culture medium. Then
cells were cultured at 37 °C in a 5% CO2 atmosphere for
different periods. Mitogenic activity was measured by
adding 1 lCi [3 H]thymidine per well for the last 18-h
period of culture. The thymidine incorporation was
measured by scintillation counting after retention over
GF/C glass-fiber filters (Whatman) of the acid-insoluble
macromolecular fraction. The means of triplicate de-
terminations were calculated for each mitogen concen-
tration. Mitogen-stimulated cells displayed the expected
kinetic proliferation, with a peak of proliferation at the
third day of culture. To analyze the influence of cate-
cholamine and corticosterone on proliferative responses,
co-incubation was carried out with epinephrine, nor-
epinephrine or corticosterone, at concentrations ranging
of 1
109 –1
105 M.
2.9. Catecholamines assay
Catecholamine concentrations were determined in
spleen samples by the fluorometric assay described by
Laverty and Taylor (1968). Briefly, spleens were ho-
mogenized in 12.5% sodium sulfite, 10% EDTA in 0.4 N
percloric acid. After 24 h at 4 °C the homogenate was
centrifuged at 5000 rpm for 10 min. The supernatants
were brought to pH 8.2 and seeded in pre-washed alu-
mina columns. The eluate was oxidized with iodine in an
alkaline medium. The fluorescence was recorded at 375
nm in a spectrofluorometer using an excitation source of
325 nm.
2.10. Corticosterone determination
To avoid fluctuations on plasma corticosterone levels
due to circadian rhythmus, animals were bled at 12:00
PM on the day of sacrifice. Blood from animals under
different experimental conditions was collected on ice in
0.25 M EGTA and separated in a refrigerated centri-
fuge. Plasma was stored at )80 °C until assay was
performed. Corticosterone levels were determined using
highly sensitive double antibody radioimmunoassay kits
(ICN Biomedical).
2.11. Statistical analysis
StudentÕs t test for unpaired values was used to de-
termine the level of significance for normally distributed
data. Group differences were tested by one-way analysis
of variance (ANOVA). Proliferation data were analyzed
84 D.M. Silberman et al. / Brain, Behavior, and Immunity 18 (2004) 81–90
by two-way ANOVA with fixed factors for experimental
groups and hormone concentrations. When multiple
comparisons were necessary after ANOVA, the Stu-
dent–Newman–Keuls test was applied. When results
were not normally distributed the non-parametric sta-
tistic Mann–Whitney U test was performed. Differences
between means were consider significant if p < :05.
3. Results
3.1. Sucrose intake and body weight
In the sucrose solution training phase (baseline
phase), sucrose intake did not show significant differ-
ences between control and CMS group (F ½2; 177 ¼
2:787, NS) (data not shown). Mean sucrose intake in the
last baseline test  SEM was 10.9  0.31 ll/g. Similarly,
variations in mice weight were not observed during this
phase (F ½2; 177 ¼ 2:146, NS] (data not shown). Mean
weight in the last baseline test  SEM was 30.7  0.2 g.
As it is shown in Fig. 1, none of these parameters sig-
nificantly changed in control animals during this ex-
periment. On the other hand, in stressed animals, both
parameters significantly decreased after the second week
of stress exposure, reached their maximal peak during
the fourth week and remained at these levels in the
subsequent weeks (Fig. 1). After 6 weeks of CMS ex-
posure, sucrose intake was significantly lower in stressed
animals than in non-exposed mice (t½1; 58 ¼ 25:3;
p < :0001) (test 18). Body weight of stressed animals
was significantly lower respect to controls (t½1; 58 ¼
10:2; p < :0001).
3.2. Open-field behavior
As it is shown in Fig. 2, the CMS exposure signifi-
cantly altered the open-field behavior. Animals sub-
Fig. 1. Sucrose intake and body weight in control and CMS animals. Fluid intake (panel A) and body weight (panel B) were determined in non-
exposed () and CMS-exposed (j) animals. Results showed represent the mean  SEM of determinations performed in 30 animals of each subgroup
of one representative experimental group in a 1-h test (three per week).
jected to CMS for 6 weeks spent significantly more time
in ambulatory exploration than control animals
(t½1; 58 ¼ 17:1; p < :0001) (Fig. 2A). Conversely, the
time engaged in the non-ambulatory exploration was
significantly diminished in CMS animals compared with
controls (t½1; 58 ¼ 16:4; p < :0001) (Fig. 2B). More-
over, the locomotive activity in CMS animals was about
2.5-fold higher than the activity observed in controls
(t½1; 58 ¼ 13:0; p < :0001) (Fig. 2C). Similarly, CMS
exposure induced a significant increment in the number
of rearing compared to control animals (t½1; 58 ¼
10:45; p < :0001) (Fig. 2D). Although behavioral
changes were evidenced notoriously in all animals after 5
weeks of CMS exposure, some of them were observed
after 3 weeks of CMS exposure (data not shown).
3.3. Antibody response to T-cell dependent and T-cell
independent antigens
To investigate the effects of CMS exposure on the
humoral response, we examined antibody production
after immunization with T-cell dependent antigens (C3H
cells and SRBC) and T-cell independent antigens (LPS
and DxB512). We found that serum from animals ex-
posed 6 weeks to CMS had a positive reaction at smaller
dilutions than those from control animals when SRBC
(Fig. 3A) or C3H cell (Fig. 3B) were used as the im-
munogen (U ¼ 0:00, p < :0002 and U ¼ 9 p ¼ :0148,
respectively). However, similar titers were obtained
when the challenge was performed with LPS (Fig. 3C) or
DxB512 (Fig. 3D) (U ¼ 21:5; p ¼ :2786 and U ¼ 23,
p ¼ :3823, respectively). In order to study the relation-
ship between the time of CMS exposure and the de-
crease in T-cell dependent antibody response, we
evaluated the antibody production after SRBC chal-
lenge in animals exposed 2, 4, 6, and 10 weeks to the
CMS scheme. As depicted in Fig. 4, serum from animals
subjected for 2 weeks to CMS did not show differences
 D.M. Silberman et al. / Brain, Behavior, and Immunity 18 (2004) 81–90 85

Fig. 2. Effect of CMS exposure on the open-field test. Non-ambulatory exploration (panel A), ambulatory exploration (panel B), locomotion (panel
C), and rearing (panel D) were determined in non-exposed (open bars) and 6 weeks CMS-exposed (dark bars) animals in a 10 min open-field test.
Results showed represent the means  SEM of determinations performed in 30 animals of each subgroup of one representative experimental group.

Fig. 3. T-cell dependent and independent antibody production in control and CMS animals. Antibody production was determined in serum of non-
exposed and 6-week CMS-exposed (CMS) animals inoculated with T-cell dependent antigens: SRBC (panel A) and C3H cells (panel B) and T-cell
independent antigens: LPS (panel C) and DxB512 (panel D) by hemagglutination test. Results show the inverse of the highest dilution of serum that
had a positive reaction determined in eight animals of each group.
86 D.M. Silberman et al. / Brain, Behavior, and Immunity 18 (2004) 81–90
Fig. 4. Time course relationship between CMS exposure and T-cell
dependent antibody production. Animals non-exposed and exposed
for the indicated time to the CMS scheme were inoculated with SRBC.
After one week, serum was obtained and the antibody production was
determined by hemagglutination test. Results show the inverse of the
highest dilution of serum that had a positive reaction determined in six
animals of each group.
in the anti-SRBC titer respect to control (U ¼ 15:5;
p ¼ :6931). After 4 weeks of CMS exposure antibody
production was lower than the corresponding control
value (U ¼ 5:5; p ¼ :0411). Antibody titer was signifi-
cantly diminished in animals subjected for 6 or 10 weeks
to CMS compared with control animals (U ¼ 0;
p < :002). No differences were found in the antibody
production against DxB512 antigen after 2–10 weeks
of CMS exposure (data not shown).
3.4. Serum corticosterone levels
To analyze the correlation between serum cortico-
sterone levels and changes in humoral response, we de-
termined hormone levels after CMS exposure (Table 1).
Animals subjected to the CMS procedure showed a
significant increase in hormone levels during the first 3
weeks of exposure (CMS vs control; tð1; 10Þ ¼ 10:49,
p < :0001; t ¼ 10:255, p < :0001; and t ¼ 2:411, p ¼
:0366 for 1, 2, or 3 weeks of exposition, respectively).
After 3 weeks, serum corticosterone levels returned to
basal values and did not show any statistically signifi-
Table 1
Catecholamine concentrations in splenic samples and serum cortico-
sterone levels from control and CMS-exposed animals
Time of CMS [Norepinephrine] [Corticosterone]
exposure (ng/mg) (ng/ml)
Non-exposed 10.9  1.0 181  8
1 week 19.0  1.8 480  39
2 weeks 17.6  1.7 442  32
3 weeks 18.8  1.8 231  25
4 weeks 11.7  1.1 190  19
6 weeks 12.7  1.4 193  11
Catecholamines and corticosterone analyses were performed in
non-exposed and 1–6 weeks CMS-exposed animals. Results represent
the means  SEM of at least four animal of each group.
cant difference when compared to controls (for 6 weeks,
t ¼ ½1; 14 ¼ :9093, p ¼ :3786, NS).
3.5. Spleen NE contents
To investigate catecholamine participation in the ef-
fect of stress on humoral response, splenic NE content
was determined after CMS exposure. As described in
Table 1, splenic NE content showed a similar pattern to
serum corticosterone levels presenting an early increase,
which was lost after 3 weeks of CMS exposure [1-week
CMS vs control; tð1; 10Þ ¼ 4:308, p ¼ :0015; 2 weeks,
tð1; 10Þ ¼ 3:793, p ¼ :0035; 3 weeks, tð1; 10Þ ¼ 4; 616,
p ¼ :001; and 6 weeks t ¼ ½1; 14 ¼ 1:191, p ¼ :2535,
NS].
3.6. Catecholamines and corticosterone effects on lym-
phocyte reactivity after mitogen stimulation
To investigate whether CMS exposure was associated
with changes in lymphocyte sensitivity to catecholamine
and corticosterone, we evaluated the effect of these
hormones on mitogen-stimulated T-cell proliferation
(Fig. 5). Addition of epinephrine to the cultures resulted
in a biphasic effect on PHA-induced proliferation of
control lymphocytes. However, on CMS lymphocytes
the concentrations tested induced only an inhibitory
effect (Fig. 5A). Two-way ANOVA revealed that epi-
nephrine significantly altered mitogen-induced T-cell
proliferation (F ½4; 40 ¼ 23:19; p < :0001) dependent on
catecholamine concentrations. Significant chronic stress
effect (F ½1; 40 ¼ 88:93; p < :0001) was observed for
proliferation when cells were incubated with exogenous
epinephrine. However, no significant interaction be-
tween group and hormone concentration was observed
(F ½4; 40 ¼ 1:70; p ¼ :1689, NS). Post hoc analysis re-
vealed that lymphocytes from CMS group exhibited
altered sensitivity to epinephrine effect compared to cells
from control group. Thus, when added to normal cell
culture, 108 M of epinephrine-stimulated cell prolifer-
ation in 21.2  5.0%. In contrast, the effect of 108 M of
epinephrine in CMS cell culture was an inhibition of
20.6  5.4% on cell proliferation, which is significantly
different (t½1; 8 ¼ 5; 660; p ¼ :0005). The addition of
105 M of epinephrine to lymphocyte cultures resulted
in a 17.8  3.1% inhibition on the proliferation of nor-
mal lymphocytes and in a 39.0  3.2% inhibition on the
proliferation of CMS lymphocytes, which is significantly
higher (t½1; 8 ¼ 5:383; p ¼ :0007). Similar results were
obtained using norepinephrine (data not shown). On the
other hand, the addition of corticosterone resulted in a
dose-dependent modulation of both control and CMS
lymphocytes (Fig. 5B). Two-way ANOVA revealed that
corticosterone significantly influence PHA-induced T-
cell proliferation (F ½4; 40 ¼ 91:12; p < :0001) depen-
dent on the steroid concentration. Significant chronic
 D.M. Silberman et al. / Brain, Behavior, and Immunity 18 (2004) 81–90 87

Fig. 5. Effect of epinephrine and corticosterone on PHA-induced T-cell proliferative response. PHA-stimulated T cells from non-exposed (open bars)
or 6 weeks CMS-exposed animals (dark bars) were co-incubated with the indicated concentrations of epinephrine (panel A) and corticosterone (panel
B). After incubation as described in Section 2, [3 H]thymidine incorporation was determined. Results are expressed as a percentage of [3 H]thymidine
incorporation at the peak of proliferation in the absence of epinephrine or corticosterone (basal proliferation). Values for basal proliferation were:
38,543  5572 dpm for non-exposed T cells and 16,332  2671 for CMS-exposed T cells. Neither of both hormones affected the non-stimulated cells.
Data showed represent the means  SEM from five independent experiments of triplicate cultures performed with one animal of each group.

stress effect was observed for T-cell proliferation under
these conditions (F ½1; 40 ¼ 56:12; p < :0001). More-
over, there was a statistically significant interaction
between group and steroid concentration (F ½4; 40 ¼
4:38; p < :005). Post hoc analysis revealed that
lymphocytes from CMS group exhibited changes in
corticosterone effects compared to the control group.
The stimulatory effect was lower and the inhibitory ef-
fect was higher on CMS lymphocytes than on control
lymphocytes. Thus, 108 M of corticosterone induced a
111.3  14.1% of stimulation in control cells whereas the
effect on CMS lymphocytes was significantly lower,
32.0  7.0% of stimulation (t½1; 8 ¼ 5:253; p < :0008).
When 105 M of corticosterone was added an inhibitory
effect of 36.0  3.5% was observed on normal lympho-
cyte proliferation. The inhibition was significantly
higher (t½1; 8 ¼ 3:812; p ¼ :0052) on CMS-lymphocyte
proliferation (53.8  3.2% of inhibition). It is important
to note that, according to our previous results, basal
PHA-induced proliferation was decreased in lympho-
cytes from animals exposed 6 weeks to CMS.
4. Discussion
The present study shows that T-cell dependent hu-
moral response is diminished in mice subjected to a
chronic mild stress model of depression. Furthermore,
we found that T lymphocytes have an increased sensi-
tivity to the inhibitory effect of stress hormones (cate-
cholamine and corticosterone). The CMS model has
been proposed as a relatively valid model of endogenous
depression, which meets criteria for construct, face and
predictive validity (Willner, 1997; Willner et al., 1992).
The present data agree with Monleon et al. (1995) in
that stressors significantly reduced the intake of a 1%
sucrose solution. This finding is considered typical in the
depression-like state observed in rats. Similarly, de-
creased locomotive activity in a novel field test is usually
observed in experimental depression (Willner, 1997).
However, our open-field test results showed an incre-
ment in ambulatory behavior as well as in locomotive
activity and rearing in stressed mice. This discrepancy
could be due to many factors, i.e., sensitivity to CMS
can be different according to the strain used and to the
animal suppliers, differences in CMS procedure details,
etc. Likewise, the duration of the single housing period
previous to the beginning of the CMS experiment could
influence the intensity of social interaction occurring
during CMS which has been reported to be an impor-
tant element of this procedure (Willner, 1997). More-
over, there are data documenting increased locomotive
activity during the open-field test after CMS (Benelli
et al., 1999; Dun^cko et al., 2001). Besides, the psycho-
motor agitation is a symptom can be present in human
depression (American Psychiatric Association, 1994).
Concerning the humoral response, CMS had a clear
effect on T-cell dependent antibody production without
disturbing T-cell independent response. A decrease in
antibody titer to T-cell dependent antigen is observed
after 4 weeks of CMS exposure. This impairment is
pronounced after 6 weeks. These findings are in agree-
ment with our previous results showing an impaired T-
cell proliferative response (Ayelli-Edgar et al., 2002).
Although there are no reports on humoral response af-
ter antigenic challenge in depression, several studies
have indicated an alteration of the antibody response to
immunization in the context of stress. Thus, poorer
antibody response to hepatitis B or influenza vaccine has
been observed in family caregivers of Alzheimer patients
as well as in others experiencing severe stress (Glaser
et al., 1992, 1998; Jabaaij et al., 1996; Kiecolt-Glaser
88 D.M. Silberman et al. / Brain, Behavior, and Immunity 18 (2004) 81–90
et al., 1996). Evidences of a decreased antibody pro-
duction and impaired isotype switching in response to
immunization with a thymus-dependent antigen have
been reported in a genetic model of stress (Murray et al.,
2001). Likewise, it was shown that chronic restraint
stress induced severe disruption of the functional ability
of lymphocyte to proliferate and to produce cytokines,
and antibody titers against tetanus toxin (Tournier
et al., 2001).
It has been known for some time that stress and
emotions are associated with neurochemical and hor-
monal changes that in turn influence the reactivity of
cells of the immune system (Maier et al., 1994). Acti-
vation of both the hypothalamic–pituitary–adrenal
(HPA) axis and the sympathetic–adrenal–medullar axis
play a prominent role in the response to psychological
stress (Azpiroz et al., 1999; Maier et al., 1994). Cells of
the immune system, like cells in other organ systems,
express receptors for hormones and neurotransmitters
(Elenkov et al., 2000; Roszman and Brooks, 1997).
Triggering of these receptors results in the modulation
of immune reactivity. Glucocorticoids have profound
effects over the immune system development and func-
tion (Munck and Guyre, 1991). On the other hand,
strong evidence has been accumulated indicating the
participation of the autonomic sympathetic system in
the modulation of lymphocyte activity via specific re-
ceptors that in turn regulate intracellular signals, such as
cyclic nucleotides levels (Elenkov et al., 2000). More-
over, it has been suggested that an inadequate commu-
nication between the neuroendocrine and the immune
system contribute to the pathophysiology of disorders
with immune alteration (Kavelaars et al., 2000). Thus,
we investigated the participation of catecholamines and
corticosterone in the disruption of T-cell dependent
antibody response. The impaired antibody production
observed after 4 weeks of CMS exposure was not cor-
related with alteration of these stress hormone levels.
This is in accordance with our previous results (Ayelli-
Edgar et al., 2003) reporting no significant variations in
corticosterone and catecholamine levels between control
and 8-week CMS exposed animals. Ayensu et al. (1995)
reported high corticosterone levels after 4 weeks of CMS
exposure in rats. However, other authors have not
found elevated levels of this hormone after prolonged
stress situations (Azpiroz et al., 1999). In addition,
Azpiroz et al. (1999), did not observe differences neither
in hypothalamic nor in hippocampal NE levels in ani-
mals under CMS conditions. These data suggest that
there was not an adrenal (cortical or medullar) activa-
tion in CMS animals. It has been proposed that there is
an adaptive response of HPA axis in the presence of
high-prolonged glucocorticoid concentrations (Azpiroz
et al., 1999; Pignatelli et al., 2000). Moreover, it was
demonstrated that chronic stress induces an hypersup-
pressive state for corticosterone secretion in response to
acute stress, which is caused by partial habituation,
coping and adaptation to the stressors (Mizoguchi et al.,
2001). Something similar could happen with NE secre-
tion. Indeed, we have observed an increase in serum
corticosterone and catecholamine levels during the first
weeks of CMS exposure, which return to basal values
after three weeks of stress exposure. On the other hand,
differences in corticosterone and catecholamines effects
on lymphocyte reactivity were observed in animals that
showed a significant impairment of antibody produc-
tion. Catecholamines showed a biphasic effect on
non-exposed lymphocytes depending on the tested con-
centration. While the lowest catecholamine concentra-
tion induced a slight increment of proliferation, the
highest concentration produced a strong inhibitory
effect on lymphocytes of non-exposed mice. Only an
inhibitory effect was observed on CMS lymphocytes
exposed to different concentration of catecholamines.
These results are in accordance with our previous find-
ings showing that lymphocytes from animals subjected
to CMS have an altered lymphocyte catecholamine re-
activity due to a significant increase in b2 adrenergic
receptors density (Ayelli-Edgar et al., 2003). On the
other hand, the addition of corticosterone to the cul-
tures induced a biphasic effect on both normal and CMS
lymphocytes. However, the stimulatory effect obtained
with low concentrations of corticosterone was higher in
normal lymphocytes and the inhibitory effect observed
for high concentrations was greater on CMS lympho-
cytes. In humans, chronic psychosocial stress of care-
givers of dementia patients has been associated with
reduced lymphocyte sensitivity to glucocorticoids when
stimulated in vitro with a T-cell mitogen (Bauer et al.,
2001). However, Stark et al. (2001) working in a mouse
model found that a psychosocial stressor induced glu-
cocorticoid resistance in mouse splenic macrophages but
not in splenic T-cells. On the other hand, Bauer et al.
(2001) showed that splenocytes collected from chroni-
cally stressed animals with stressor re-exposure are more
sensitive to exogenous application of dexamethasone
and corticosterone than animals chronically stressed
without stressor re-exposure. In addition, this group
(Bauer et al., 2003) reported that alterations in immune
function and steroid regulation associated with depres-
sion are not related to elevated basal levels of cortisol
and suggests that lymphocyte steroid resistance may be
associated with treatment resistant depression. It is
reasonable to think that complex interactions occur
during chronic stress exposure and depression and it is
very difficult to understand which is the particular al-
teration in the immune–neuroendocrine communication
associated to a specific condition. In our model it is
possible that early increments of catecholamine and
corticosterone levels may induce long-lasting changes
in lymphocyte sensitivity to catecholamines and
corticosterone.
 D.M. Silberman et al. / Brain, Behavior, and Immunity 18 (2004) 81–90
In summary, the present study would be the first ex-
perimental evidence showing that chronic stress may
increase lymphocyte sensitivity to stress hormones in-
hibitory effects. This leads to an impairment in T-cell
dependent immune response, underlying the functional
significance of the stress hormone receptors on lym-
phocytes. Additional studies should be performed to
understand deeply the implication of these findings in
the immunological alteration observed in stress and
depression.
Acknowledgments
The authors thank Daniel Gonz
alez for his invaluable
help in the animal stress model and Mrs. Patricia
Fern
andez and Mrs. Ana I. Casella for the secretarial
assistance. Also, Dr. Juan Jos
e L
opez for his critical
revision of the manuscript. This work was supported by
grant Ram
on Carrillo-Arturo O~nativia from Ministerio
de Salud de la Naci
on and PIP 0543/98 from CONI-
CET.
References
American Psychiatric Association, 1994. DSM IV- Diagnostic and
Statistical Manual of Psychiatric Disorders, fourth ed. American
Psychiatric Association, Washington, DC.
Ayelli-Edgar, V., Cremaschi, G., Sterin-Borda, L., Genaro, A.M.,
2002. Altered expression of autonomic neurotransmitter receptors
and proliferative responses in lymphocytes from a chronic mild
stress model of depression: effects of fluoxeine. Brain Behav.
Immun. 16, 333–350.
Ayelli-Edgar, V., Silberman, D.M., Cremaschi, G.A., Zieher, L.M.,
Genaro, A.M., 2003. Altered lymphocyte catecholamine reactivity
in mice subjected to chronic mild stress. Biochem. Pharmacol. 65,
15–23.
Ayensu, W.K., Pucilowski, O., Mason, G.A., Overstreet, D.H.,
Rezvani, A.H., Janowski, D.S., 1995. Effects of chronic mild stress
on serum complement activity, saccharin preference and cortico-
sterone levels in Flinders lines of rats. Physiol. Behav. 57, 165–169.
Azpiroz, A., Fano, E., Garmendia, L., Arregi, A., Cacho, R., Beitia,
G., Brain, P.F., 1999. Effects of chronic mild stress (CMS) and
imipramine administration, on spleen mononuclear cell prolifera-
tive response, serum corticosterone level and brain norepinephrine
content in male mice. Psychoneuroendocrinology 24, 345–361.
Bauer, M.E., Perks, P., Lightman, S.L., Shanks, N., 2001. Restraint
stress is associated with changes in glucocorticoid immunoregula-
tion. Physiol. Behav. 73, 525–532.
Bauer, M.E., Papadopoulos, A., Poon, L., Perks, P., Lightman, S.L.,
Checkley, S., Shanks, N., 2003. Altered glucocorticoid immuno-
regulation in treatment resistant depression. Psychoneuroendocri-
nology 28, 49–65.
Benelli, A., Filaferro, M., Bertolini, A., Genedani, S., 1999. Influence
of S-adenosyl-L -methionine on chronic mild stress-induced anhe-
donia in castrated rats. Br. J. Pharmacol. 127, 645–654.
Chrousos, G.P., 1995. The hypothalamic–pituitary–adrenal axis and
immune-mediated inflammation. N. Engl. J. Med. 332, 1351–1362.
Dun^cko, R., Kiss, A., S  kult
eyov
a, I., Rusnak, M., Jeov
a, D., 2001.
Corticotropin-releasing hormone mRNA levels in response to
chronic mild stress rise in male but not in female rats while tyrosine
89
hydroxylase mRNA levels decreased in both sexes. Psychoneu-
roendocrinology 26, 77–89.
Elenkov, I.J., Wilder, R.L., Chrousos, G.P., Vizi, E.S., 2000. The
symphathetic nerve—an integrative interface between two super-
system: the brain and the immune system. Pharmacol. Rev. 52,
595–638.
Glaser, R., Kiecolt-Glaser, J.K., Malarkey, W.B., Sheridan, J.F., 1998.
The influence of psychological stress on the immune response to
vaccines. Ann. N. Y. Acad. Sci. 840, 649–655.
Glaser, R., Kiecolt-Glaser, J.K., Bonneau, R.H., Malarkey, W.,
Kennedy, S., Hughes, J., 1992. Stress-induced modulation of the
immune response to recombinant hepatitis B vaccine. Psychosom.
Med. 54, 22–29.
Herbert, T.B., Cohen, S., 1993. Depression and immunity: a meta-
analytic review. Psychol. Bull. 113, 472–486.
Irwin, M., 1995. Psychoneuroimmunology of depression. In: Bloom,
F.E., Kupfer, D.J. (Eds.), Psychopharmacology: The Fourth
Generation of Progress. Raven Press, New York, pp. 983–988.
Jabaaij, L., van Hattum, J., Vingerhoets, J.J., Oostveen, F.G.,
Duivenvoorden, H.J., Baillieux, R.E., 1996. Modulation of im-
mune response to rDNA hepatitis B vaccination by psychological
stress. J. Psychosom. Res. 41, 129–137.
Kavelaars, A., Kuis, W., Knook, L., Sinnema, G., Heijen, C.J., 2000.
Disturbed neuroendocrine–immune interactions in chronic fatigue
syndrome. J. Clin. Endocrinol. Metab. 85, 692–696.
Kiecolt-Glaser, J.K., Glaser, R., Gravenstein, S., Malarkey, W.B.,
Sheridan, J., 1996. Chronic stress alters the immune response to
influenza virus vaccine in older adults. Proc. Natl. Acad. Sci. USA
93, 3043–3047.
Laverty, R., Taylor, K., 1968. The fluorometric assay of catechola-
mines and related compounds: improvement and extension to the
hydroxyindole technique. Ann. Biochem. 22, 269–279.
Leonard, B.E., 1995. Stress and immune system-immunological
aspects of depressive illness. In: Leonard, B., Miller, K. (Eds.),
Stress, the Immune system and Psychiatry. Wiley, New York, pp.
113–136.
Maes, M., Smith, R., Scharpe, S., 1995. The monocyte-T-lymphocyte
hypothesis of major depression. Psychoneuroendocrinology 20,
111–116.
Maes, M., Vandoolaeghe, E., van Hunsel, F., Brill, T., Demendts, P.,
Wauters, A., Neels, H., 1997. Immune disturbances in treatment-
resistant depression: modulation by antidepressive treatments.
Hum. Psychopharm. Clin. 12, 153–162.
Maier, S.F., Watkins, L.R., Fleshner, M., 1994. Psychoneuroimmun-
ology. The interface between behavior, brain and immunity. Am.
Psychol. 49, 1004–1017.
Miller, A.H., 1998. Neuroendocrine and immune system interactions
in stress and depression. Psychol. Clin. Nor. Am. 21, 443–463.
Mizoguchi, K., Yuzurihara, M., Ishige, A., Sasaki, H., Chui, H.,
Tabira, T., 2001. Chronic stress differentially regulates glucocor-
ticoid negative feedback in rats. Psychoneuroendocrinology 26,
443–459.
Monleon, S., DÕAquila, P., Parra, A., Simon, V.M., Brain, P.F.,
Willner, P., 1995. Attenuation of sucrose consumption in mice by
chronic mild stress and its restoration by imipramine. Psychophar-
macology 117, 453–457.
Munck, A., Guyre, P.M., 1991. Glucocorticoids and immune function.
In: Ader, R., Felten, D.L., Cohen, N. (Eds.), Psychoneuroimmun-
ology. Academic Press, Rochester, pp. 447–475.
Murray, S.E., Lallman, H.R., Heard, A.D., Rittenberg, M.B., Stenzel-
Poore, M.P., 2001. A genetic model of stress displays decreased
lymphocytes and impaired antibody responses without altered
susceptibility to Streptococcus pneumoniae. J. Immunol. 167, 691–
698.
Overstreet, D.H., Steiner, M., 1998. Genetic and environmental
models of stress-induced depression in rats. Stress Med. 14, 261–
268.
90 D.M. Silberman et al. / Brain, Behavior, and Immunity 18 (2004) 81–90
Pignatelli, D., Maia, M., Castro, R., da Coceicao Magalhaes, M.,
Vivier, J., Defaye, G., 2000. Chronic stress effects on the rat adrenal
cortex. Endocr. Res. 26, 537–544.
Ravindran, A.V., Griffiths, J., Merali, Z., Anisman, H., 1998.
Circulating lymphocyte subsets in major depression and
dysthymia with typical or atypical features. Psychosom. Med. 60,
283–289.
Roszman, T.L, Brooks, W.H., 1997. Interactive signaling pathways of
the neuroendocrine–immune network. Chem. Immunol. 69, 203–
222.
Servatius, R.J., Natelson, B.H., Moldow, R., Pogach, L., Brennan,
F.X., Ottenweller, J.E., 2000. Persistent neuroendocrine changes in
multiple hormonal axes after a single o repeated stressor exposures.
Stress 3, 263–274.
Stark, J.L., Avitsur, R., Padgett, D.A., Campbell, K.A., Beck, F.M.,
Sheridan, J.F, 2001. Social stress induces glucocorticoid resistance
in macrophages. Am. J. Physiol. Regulatory Integrative Comp.
Physiol. 280, R1799–R1805.
Tournier, J.N., Mathieu, J., Mailfert, Y., Multon, E., Drouet, C.,
Jouan, A., Drouet, E., 2001. Chronic restraint stress induces severe
disruption of the T-cell specific response to tetanus toxin vaccine.
Immunology 102, 87–93.
Welsh, K.I., Batchelor, J.R., 1978. Assays for antibodies against
histocompatibility antigens. In: Weir, D.M. (Ed.), Handbook of
Experimental Immunology, vol. 2. Blackwell Scientific Publications
DN, Oxford, pp. 35.1–35.20.
Willner, P., 1997. The chronic mild stress procedure as an animal
model of depression: valid, reasonably reliable, and useful.
Psychopharmacology 134, 371–377.
Willner, P., Muscat, R., Papp, M., 1992. Chronic mild stress-induced
anhedonia: a realistic animal model of depression. Neurosci.
Biobehav. R16, 525–534.