Species differences in anxiety-related responses in male prairie and meadow voles: the effects of social isolation.

Abstract

Prairie (Microtus ochrogaster) and meadow voles (M. pennsylvanicus) are closely related species that differ in life strategy and social behaviors, and thus provide an excellent comparative model for the study of neuronal and hormonal mechanisms underlying behavior. In the present study using the elevated plus maze (EPM) test, we found that male prairie voles entered the open arms of the EPM more and remained there longer, and showed a higher level of overall locomotor activity than did male meadow voles. In addition, two weeks of social isolation induced an increase in open arm entries in prairie, but not meadow, voles. Prairie voles also had a higher level of circulating corticosterone compared to meadow voles, and the EPM test increased circulating corticosterone in prairie voles. Finally, social isolation coupled with the EPM test influenced Fos-immunoreactive expression in several brain areas, including the medial preoptic area, ventromedial hypothalamus, amygdala, and prefrontal cortex differently between the two species. Together, these data indicate a neural circuit involved in mediating anxiety-associated behavior in voles, and that the functioning of this circuit is influenced by social environment differently between social and non-social species.

Full text

Page 1

Species differences in anxiety-related responses in male prairie and
meadow voles: The effects of social isolation
Jennifer R. Stowea, Yan Liua, J. Thomas Curtisa, Marc E. Freemanb, Zuoxin Wanga,*
aDepartment of Psychology and Program in Neuroscience, Florida State University, Tallahassee, FL 32306, USA
bDepartment of Biological Sciences and Program in Neuroscience, Florida State University, Tallahassee, FL 32306, USA
Abstract
Prairie (Microtus ochrogaster) and meadow voles (M. pennsylvanicus) are closely related species that differ in life strategy and social
behaviors, and thus provide an excellent comparative model for the study of neuronal and hormonal mechanisms underlying behavior. In the
present study using the elevated plus maze (EPM) test, we found that male prairie voles entered the open arms of the EPM more and
remained there longer, and showed a higher level of overall locomotor activity than did male meadow voles. In addition, two weeks of social
isolation induced an increase in open arm entries in prairie, but not meadow, voles. Prairie voles also had a higher level of circulating
corticosterone compared to meadow voles, and the EPM test increased circulating corticosterone in prairie voles. Finally, social isolation
coupled with the EPM test influenced Fos-immunoreactive expression in several brain areas, including the medial preoptic area, ventromedial
hypothalamus, amygdala, and prefrontal cortex differently between the two species. Together, these data indicate a neural circuit involved in
mediating anxiety-associated behavior in voles, and that the functioning of this circuit is influenced by social environment differently
between social and non-social species.
D 2005 Elsevier Inc. All rights reserved.
Keywords: Elevated plus maze; Corticosterone; C-fos; Amygdala; VMH; MPOA
1. Introduction
Pathological anxiety is defined as fear without a relevant
corresponding object or event. Animal models of anxiety are
widely sought in an attempt to analyze pathological anxiety
states based on the assumption that some anxiety mecha-
nisms are essential for survival and are a feature of all
mammals [1]. The elevated plus maze (EPM) is one of the
most commonly used behavioral paradigms for testing
animal anxiety; it relies on a rodent’s relative aversion to
venture onto the open arms of the maze versus the safer
closed arms [2,3]. Animals that spend more time in the open
arms are thought to be less anxious. It has been suggested
that the EPM test is sensitive to alterations in anxiety
produced by ecologically relevant stimuli [4].
Activation of the hypothalamic–pituitary–adrenal
(HPA) axis is a physiological marker of stress and is
typically measured by an increased level of plasma
corticosterone (CORT) in rodents [5]. The EPM test is
known to increase CORT levels in rats, particularly if the
animals are confined to the open arms of the maze [3].
Fluctuations in CORT levels are also seen in response to
changes in social interactions. Social isolation, for
example, is considered stressful or anxiety producing
and plays a role in influencing behavior on the EPM in a
strain- and/or gender-specific manner [6,7]. Socially
isolated rats show an increase in anxiety-like behavior
on the EPM that is associated with increases in
circulating CORT [8].
Microtine rodents have proven to be a useful model for
studying the effects of social interactions on physiology
and behavior. For example, prairie voles (Microtus
ochrogaster) are highly social, demonstrating the charac-
teristics of monogamy [9]. Prairie voles display high levels
of social affiliation and mating induces pair bond
0031-9384/$ - see front matter D 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.physbeh.2005.08.007
* Corresponding author. Department of Psychology, Florida State
University, Tallahassee, FL 32306-1270. Tel.: +1 850 644 5057; fax: +1
850 644 7739.
E-mail address: zwang@psy.fsu.edu (Z. Wang).
Physiology & Behavior 86 (2005) 369 – 378

Page 2

formation between the mates [10]. Social isolation has
been found to increase CORT levels in female prairie voles
[11], and increased CORT seems to facilitate pair bonding
in male, but not female, prairie voles [12,13]. In contrast,
meadow voles (Microtus pennsylvanicus) are promiscuous,
asocial, display low levels of affiliative behavior, and
mating does not induce pair bond formation [14,15].
Together, these data suggest that social environment may
differentially influence endocrine responses and behavior
of closely related vole species with different life strategies.
Therefore, comparing these species provides an excellent
opportunity for the study of hormonal and neurochemical
mechanisms underlying behavior [16,17].
As social environment has significant impacts on anxiety
[6,7,11], we hypothesized that social isolation may affect
anxiety-related behaviors and the underlying hormonal and
neuronal mechanisms differently in monogamous and
promiscuous vole species. To test this hypothesis, the
present study was conducted 1) to compare anxiety behavior
and associated circulating CORT and regional brain
activation between male prairie and meadow voles, and 2)
to examine the effects of social environment, in particular
acute or chronic social isolation, on those measures. We
used the EPM test to study anxiety behavior and radio-
immunoassay (RIA) to measure plasma levels of CORT. Fos
is a protein product of the immediate early gene c-fos that is
rapidly induced following sensory stimulation and can be
used as a marker of neuronal activation [18,19]. Therefore,
we used Fos immunoreactivity (Fos-ir) to label brain areas
activated following EPM testing.
2. Materials and methods
2.1. General methods
2.1.1. Animals
Subjects were sexually naive male prairie and meadow
voles that were the offspring from a laboratory breeding
colony. All voles were weaned at 21 days of age and placed
in same sex sibling pairs in plastic cages (29?18?13 cm)
that contained cedar chip bedding. Food and water were
provided ad libitum. All cages were maintained in a
14L:10D photoperiod with lights on at 0700 h while
temperature was maintained at about 21 -C. All subjects
were 3–4 months of age at the beginning of experiments
and each subject was randomly assigned to either treatment
or control group. All groups were placed in the testing room
24 h or 2 weeks prior to the onset of behavioral testing to
allow for habituation (see below).
2.1.2. Social isolation
In the acute social isolation situation, voles were
isolated from their cage mates for a period of 24 h [11]
whereas in the chronic social isolation situation, voles
were housed individually for a period of 2 weeks prior to
behavioral testing. During these two periods of social
isolation, animals were placed in the testing room, and an
opaque divider was placed between cages to eliminate
visual cues. All animals were maintained under the same
food, photoperiod, and temperature conditions as in the
colony room.
2.1.3. Elevated plus maze test
The elevated plus maze (EPM) has been validated in the
study of both rats [3] and mice [2] and has been successfully
employed in work with voles [20]. Briefly, the EPM
(Columbus Instruments, Columbus, OH) is comprised of
two open arms (35 cm (L)?6.5 cm (W)) and two closed
arms (35 cm (L)?5 cm (W)?15 cm (H)) that cross in the
middle, and is elevated 45 cm off the ground. The EPM test
was carried out between 0900 and 1100 h under standard
lighting conditions.
The subjects were individually placed on the intersection
of the EPM facing an open arm and then observed for 10
min. Number of entries into the open or closed arms, time
spent in each arm or in the center, rears, and falls were
recorded by an experimenter. A rear was defined as standing
on the hind paws with the front 2 paws placed on the walls
of the maze. An entry into an arm was counted only when
all four paws of an animal crossed from the center panel
onto the arm. After the behavioral test, each subject was
returned to a clean cage for 2 h without any further
disturbance. The maze was cleaned thoroughly with soapy
water between animals.
2.1.4. CORT radioimmunoassay
Two hours following the EPM testing, subjects were
anesthetized with sodium pentobarbitol (1mg/10 g body
weight) and decapitated. Pilot experiments were conducted
to compare plasma samples obtained by rapid decapi-
tation, sodium pentobarbitol followed by decapitation, or
sodium pentobarbitol followed by heart puncture, and the
data indicated no significant group differences in CORT
levels. Therefore, sodium pentobartitol administration
followed by decapitation was used in all experiments.
Approximately 1.0 ml of trunk blood was collected in a
tube containing 50 Al of EDTA to prevent clotting. Blood
samples were centrifuged for 20 min at 4 -C at 3000
RPM, and plasma was extracted for use in the radio-
immunoassay. Plasma samples were analyzed for CORT
using a Coat-A-Count\assay kit (Diagnostic Products
Corp., Los Angeles, CA).
Because prairie voles are known to have relatively high
glucocorticoid levels, plasma was diluted 1:10 in assay
buffer (0.1 M phophate-buffered saline (PBS), pH 7.4)
with 1% sodium azide) to insure that results would
reliably fall within the standard curve fit by linear
regression [38]. Meadow vole plasma was also diluted
in the same fashion to obtain an accurate comparison. All
plasma samples were run in duplicate. Interassay variance
was less than 10%.
J.R. Stowe et al. / Physiology & Behavior 86 (2005) 369–378
370

Page 3

2.1.5. C-fos immunocytochemistry
Following anesthesia and blood extraction, brains were
rapidly removed and placed into 4% paraformaldehyde in
0.1 M PBS) at 4 -C for fixation, with the solution changed at
6 h. After 24 h of fixation, brains were stored in 30%
sucrose in PBS, and then sliced into 40 Am coronal sections
using a microtome, and stored in 0.1 M PBS containing
1.0% sodium azide.
Floating brain sections at 120 Am intervals were
processed for c-fos immunocytochemistry using established
procedures [22,23]. Briefly, sections were rinsed in 0.05M
Tris–NaCl buffer (pH 7.6); incubated in 3.0% methanol and
0.5% H2O2 in the same buffer for 30 min; and then
incubated in 10% normal goat serum (NGS) in 0.05 M
Tris–NaCl with 0.5% Triton X-100 (pH 7.6) for 1 h.
Sections were then incubated in rabbit polyclonal IgG
antibody for c-fos (1:30,000, Jackson Immuno) in 0.05 M
Tris–NaCl with 0.5% Triton X-100 and 2.0% NGS (pH 7.6)
for 72 h at 4 -C and then one hour at room temperature,
followed by incubation with the biotinylated goat anti-rabbit
secondary antibody (1:300, Jackson Immuno) for 2 h.
Sections were incubated in Vector Elite ABC complex in the
Tris–NaCl buffer for 90 min, stained with Nickel-DAB
(Vector Laboratories, Inc. Burlingtone, CA), mounted on
microscope slides, and then cover slipped. To control for
variability all sections within each experiment were
processed simultaneously.
2.1.6. Data quantification and analysis
For all animals, plasma samples and brain slides were
coded to conceal group identity until all samples were
analyzed. The c-fos labeled cells were examined in the
medial (MeA), anterior cortical (ACo), and central (CeA)
subnuclei of the amygdala (AMYG); bed nucleus of the stria
terminalis (BST; including the anterior dorsal and anterior
ventral parts); lateral septum (LS; intermediate); paraven-
tricular nucleus (PVN); ventromedial hypothalamus
(VMH); medial preoptic area (MPOA); anterior hypothal-
amus (aHYP); and prefrontal cortex (PFC). These areas
were chosen because they have been implicated in anxiety
and social behaviors [23–27], activated following an EPM
test [26,28], and/or involved in regulation of the HPA axis
activity in rodents [29].
Brain sections were matched between animals with 3–4
sections per brain area being examined. Fos-ir cells were
visualized under 10? magnification using a Zeiss Axioskop
II microscope, and the images were captured using a
computerized image program (NIH Image 1.60). All Fos-ir
cells in the sampling area were quantified and analyzed
bilaterally. The average number of cells containing Fos-ir
staining from both sides of each brain area was used to
provide individual means for data analysis. All behavioral,
radioimmunoassay, and c-fos immunostaining data were
analyzed by either one-way or two-way analysis of variance
(ANOVA), and significant effects were further evaluated
using a Student Newman–Keul’s (SNK) post-hoc test.
2.2. Experiment 1: does social isolation influence anxiety
behavior on the EPM and subsequent CORT and neuronal
activation differentially between male prairie and meadow
voles?
Prairie and meadow voles show remarkable differences
in life strategies and social behaviors [9,15,30]. As social
environment has significant impacts on anxiety [6,7,11],
we hypothesized that in response to social isolation these
two species may differ in anxiety-related behaviors and the
underlying hormonal and neuronal mechanisms. In the
present experiment, we studied anxiety behaviors in the
EPM and examined the subsequent circulating CORT and
neuronal activation in male prairie and meadow voles that
were either housed in pairs or socially isolated for 24 h or
for 2 weeks.
2.2.1. Design
Male prairie and meadow voles were randomly
assigned into one of 4 experimental groups. In the no
social isolation group (no isolation; n=9 prairie, n=8
meadow), subjects were housed with a same-sex sibling
and transferred to the testing room for 24 h prior to being
tested. In the 24 h social isolation group (24 h; n=9
prairie, n=6 meadow), subjects were housed individually
for 24 h in the testing room. In the 2 weeks social
isolation group (2 wks; n=9 prairie, n=6 meadow),
subjects were housed individually for a period of 2 weeks
in the testing room. At the end of treatment, subjects from
all above three groups were individually tested on the
EPM for 10 min. A fourth group of animals served as
controls for handling (Control; n =8 prairie, n =6
meadow). For this group, subjects were housed with a
sibling and handled such that they were picked up from
the home cage, placed alone into a clean cage and picked
up again 10 min later although no EPM test was
conducted. After the EPM test or being handled, subjects
in all groups were placed individually into clean cages and
left undisturbed for 2 h. The 2 h time period was chosen
based on our pilot data showing clear Fos-ir induction in
the prairie vole brains following social stimulation.
Thereafter, subjects were anesthetized, trunk blood was
collected for CORT radioimmunoassay, brains were
harvested, placed into 4% paraformaldehyde for 24 h of
fixation, and brain sections were processed for c-fos
immunocytochemistry. Behavioral and CORT data were
analyzed by a 2-way ANOVA followed by an SNK post-
hoc test. Group differences in the number of Fos-ir cells in
each species were examined by a one-way ANOVA
followed by an SNK post-hoc test.
2.3. Experiment 2: does the EPM test elevate CORT time-
dependently in male prairie voles?
Social isolation is anxiety producing for prairie voles as
it elevates plasma CORT in both pups [31] and adult
J.R. Stowe et al. / Physiology & Behavior 86 (2005) 369–378
371

Page 4

females [11]. However, in our experiment socially isolated
voles showed more entries on the open arms of the EPM,
which may indicate less anxiety in lab rodents [2,3] but not
in wild animals [32]. Unfortunately, our CORT data 2 h
after the EPM test were not conclusive. In rats, the EPM
test induces a significant increase in circulating CORT
approximately 20 min post-test [33]. Therefore, it remained
possible that the EPM test could also induce an increase in
plasma CORT in male prairie voles, but the 2-h waiting
period, which was needed for Fos-ir induction, may have
been too long to detect such a change in plasma CORT. To
test this hypothesis, a time course experiment was
performed to examine CORT elevation following the
EPM test.
2.3.1. Design
As above described, male prairie voles were tested on
the EPM for 10 min, put in a clean cage, and then
sacrificed at 10 (n=8), 20 (n=8), 30 (n=7), or 60 min
(n=8) following the EPM test. All subjects were sacrificed
between 1100 and 1300 h as in Experiment 1. Trunk blood
was collected and processed for CORT measurement using
RIA. Handled control animals, without the EPM test, were
also sacrificed at different time points. As no differences
were found in their CORT levels, they were pooled into
one control group (n=19) for statistical analysis. CORT
levels were analyzed using a one-way ANOVA followed
by an SNK post-hoc test.
2.4. Experiment 3: does social isolation alone alter basal
levels of neuronal activation in male prairie voles?
Data from Experiment 1 demonstrated that socially
isolated prairie voles had significant increases in the number
of Fos-ir cells in several brain regions following the EPM
test, but this increased Fos-ir expression was not found in
non-isolated animals. It was not clear, however, if increased
Fos-ir staining was due to increased brain responsiveness to
the EPM test or to an increased basal level of neuronal
activation resulting from social isolation. Therefore, several
additional groups of animals were processed in the present
experiment to test the hypothesis that social isolation alone
might elevate the basal level of neuronal activation in the
brain of prairie voles.
2.4.1. Design
Male prairie voles were randomly assigned into one of 3
experimental groups. Two groups were established based on
length of social isolation: socially isolated for 24 h (n=8) or
for 2 weeks (n=8). Control animals were housed with a
male cage mate (n=8). After each treatment, subjects were
anesthetized and sacrificed between 1100 and 1300 h, and
their brain sections were processed for Fos-ir staining as
above described. Group differences in Fos-ir staining in
selected brain areas were analyzed by a one-way ANOVA
followed by an SNK post-hoc test.
3. Results
3.1. Elevated plus maze behavior
Species differences were found in behaviors on the EPM
between male prairie and meadow voles (Table 1). Prairie
voles entered the open arms more, and remained there
longer, than did meadow voles. Prairie voles also entered the
closed arms more and demonstrated greater overall activity
as indicated by the total arm entries than did meadow voles.
Meadow voles, on the other hand, spent more time in the
closed arms and displayed rearing behavior more frequently
than prairie voles.
Significant species by treatment effects were also found:
generally prairie, but not meadow, voles’ behaviors were
affected by social isolation (Fig. 1). There was a significant
species by treatment interaction (F(2, 42)=5.66, p<0.05) in
which prairie voles that were isolated for 2 weeks entered to
the open arms of the EPM more frequently than prairie voles
that were not isolated or were isolated for 24 h. Prairie voles
that were isolated for 2 weeks also had a greater number of
total arm entries on the EPM than those that were isolated
for 24 h (F(2, 42)=3.23, p<0.05). Social isolation experi-
ence did not alter male meadow vole behavior on the EPM.
3.2. Plasma CORT concentration
A significant species difference was found in circulating
levels of plasma CORT: prairie voles had higher levels of
CORT than did meadow voles (F(1, 53)=171.0, p<0.001;
Table 2). Further, prairie, but not meadow, voles showed
approximately a 25% increase in the level of plasma CORT
2 h after the EPM test but this increase did not reach
statistical significance. Acute or chronic social isolation
experience did not alter plasma CORT levels in either
species 2 h post EPM.
3.3. Fos-ir induction following the EPM test
When analyzed by a two-way ANOVA, prairie voles
showed persistent higher levels of Fos-ir labeling than
meadow voles in all brain areas measured. To focus on
the effects of the EPM test and social isolation on
neuronal activation, we then analyzed Fos-ir data using a
Table 1
Species differences in behavior on the elevated plus maze (meanTSEM)
Behavior on the EPMPrairie voleMeadow vole
p
Frequency Open arms
Closed arms
Total entries
Rears
Falls
Open arms
Closed arms
Center
7.9T1.0
20.3T2.8
28.2T3.5
14.9T1.3
1.1T0.2
136.9T16.7
370.5T20.7
42.5T8.3
1.4T0.5
10.2T1.7
11.6T1.9
23.5T2.9
0.1T0.1
29.4T13.7
503.0T31.9
36.5T18.8
0.001
0.01
0.001
0.01
0.001
0.001
0.001
ns
Duration
(sec)
J.R. Stowe et al. / Physiology & Behavior 86 (2005) 369–378
372

Page 5

one-way ANOVA within each species. Cells expressing
Fos-ir were found in several regions of the hypothalamus,
and social isolation influenced the number of Fos-ir cells
following the EPM test in a region-specific manner. In the
MPOA, the EPM test increased the number of Fos-ir cells
in prairie voles that experienced social isolation
(F(3,29)=7.23, p<0.001), but this effect was not found
in meadow voles (Figs. 2 and 3A). In VMH, the EPM
test increased the number of Fos-ir cells in socially
isolated prairie (F(3, 30)=5.65, p<0.01) and meadow voles
(F(3,24)=6.95, p<0.01; Fig. 3B). No treatment effects
were found in the PVN (Fig. 3C), BST (Fig. 3D), or CeA
(Fig. 3E) of either species. However, in the MeA (Fig. 4),
prairie voles isolated for 2 weeks had more Fos-ir cells
than the control group (F(3, 30)=3.13, p<0.05), whereas
meadow voles isolated for either 24 h or 2 weeks had
increased numbers of Fos-ir cells following the EPM test
(F(3, 24)=14.46, p<0.001; Fig. 3F).
Treatment did not influence Fos-ir labeling in the ACo or
aHYP of prairie voles. However, meadow voles that were
socially isolated for 24 h showed increased numbers of Fos-
ir cells following the EPM test in both areas (for ACo: F(3,
24)=9.14, p<0.001, Fig. 3G; For aHYP: F(3, 24)=6.91,
p<0.01, Fig. 3H). In the lateral septum, meadow voles that
were socially isolated for either 24 h or 2 weeks had more
Fos-ir cells following the EPM test than did control animals
(F(3, 24)=5.09, p<0.01; Fig. 4I). Finally, animals that
experienced 24 h of social isolation showed an increase in
the number of Fos-ir cells in the prefrontal cortex (Fig. 3J)
in both prairie (F(3, 29)=3.99, p<0.05) and meadow voles
(F(3, 24)=6.15, p<0.01).
3.4. Time dependent increase in plasma CORT following the
EMP test
In Experiment 2, a significant group difference was
found in the levels of plasma CORT in male prairie voles
(F(4, 45)=5.5, p<0.001; Fig. 5). Twenty minutes after the
EPM test, males had a level of plasma CORT significantly
higher than that of control males. At 1 h post EPM test,
plasma CORT had returned to the baseline level.
3.5. Basal level Fos-ir expression associated with social
isolation
In Experiment 3, in comparison to control animals that
were housed in pairs, increased Fos-ir staining was observed
in several brain regions of male prairie voles following 24 h
or 2 weeks of social isolation in the absence of the EPM test
(Table 3). Animals that experienced 24 h of social isolation
had higher basal levels of Fos-ir expression in the MPOA,
VMH and aHYP relative to controls and to those animals
isolated for 2 weeks. In addition, animals that experienced 2
weeks of social isolation had fewer Fos-ir cells in LS
compared to the control animals and animals isolated for 24
h. No group differences were found in any other brain areas.
4. Discussion
Voles are a group of rodents that show remarkable
diversity in social organizations and behaviors, and thus
provide a useful comparative model for studying the
neuronal and hormonal bases of behavior [15–17,34]. Here,
we compared highly social, monogamous prairie voles with
non-social, promiscuous meadow voles to examine anxiety-
related behaviors, circulating CORT concentrations, and
neuronal activation in the brain, and to study the effects of
social isolation experience on these markers. Our data
demonstrate that social isolation is more stressful for social
prairie voles than for non-social meadow voles, and exerts
stimulus- and species-specific effects on anxiety-related
Table 2
Plasma levels of corticosterone (ng/ml, meanTSEM) in prairie and meadow
voles
Prairie vole Meadow vole
Control
No isolation- EPM
24-h isolation- EPM
2-wk isolation- EPM
1517.1T190.9
2023.6T180.8
2002.4T147.8
2000.5T221.0
462.9T441.1
401.8T51.1
416.3T43.4
418.6T38.0
no isolation
24 hrs
Time in open arms (sec/10 min)
# open arm entries
b
a
a
a
b
a
ab
a
a
a
ab
c
2 wks
C.B. A.
# total arm entries
PrairieMeadow
0
4
8
12
16
20
PrairieMeadow
0
10
20
30
40
50
60
Prairie Meadow
0
50
100
150
200
250
Fig. 1. Prairie voles spent more time on the open arms of the EPM (A), entered the open arms more frequently (B) and showed a higher level of overall
locomotor activity indicated by the total arm entries (C) than did meadow voles. Further, prairie voles that were socially isolated for 2 weeks entered the open
arms more than did the other two groups (B) and showed a higher level of locomotor activity compared to the prairie voles socially isolated for 24 h (C). Social
isolation had no significant effects on any of the behaviors measured. Error bars indicate SEM.
J.R. Stowe et al. / Physiology & Behavior 86 (2005) 369–378
373