Corticosterone response to the plus-maze: high correlation with risk assessment in rats and mice. Physiol Behav

Ethopharmacology Laboratory, School of Psychology, University of Leeds, UK.
Physiology & Behavior (Impact Factor: 2.98). 12/1999; 68(1-2):47-53. DOI: 10.1016/S0031-9384(99)00140-7
Source: PubMed
Exposure to the elevated plus-maze induces behavioural and physiological effects in rodents consistent with fear/anxiety. Maze-naive animals display high levels of risk assessment towards the open arms, and explore these areas less extensively than other parts of the maze while, immediately following the test, pain latencies, skin conductance levels, and plasma corticosterone titres (CORT) are significantly elevated. Although previous research has suggested a link between the plasma CORT response and open-arm exploration, significant elevations in CORT have also been found with restricted exposure to the closed arms. The present study employed ethological measures in an attempt to further characterise the relationship between behavioural and CORT responses to this widely used animal model of anxiety. Our results confirm that, relative to home-cage controls, 5-min exposure to the plus-maze significantly increases plasma CORT levels in test-naive male Wistar rats and male Swiss-Webster mice. Furthermore, in both species, the CORT response was found to be highly correlated with measures of risk assessment (mice: rs = +0.87; rats: rs = +0.58), but not with measures of open-arm activity (entries, time), general locomotor activity, rearing, or head dipping. Findings are discussed in relation to the functional significance of risk assessment in potentially dangerous situations and the potential involvement of glucocorticoids in this process. All rights reserved.


Available from: Paul F Brain, Jan 30, 2016
Physiology & Behavior 68 (1999) 47–53
0031-9384/99/$ – see front matter © 1999 Elsevier Science Inc. All rights reserved.
PII: S0031-9384(99)00140-7
Corticosterone response to the plus-maze: High correlation with
risk assessment in rats and mice
R.J. Rodgers
, J. Haller
, A. Holmes
, J. Halasz
, T.J. Walton
, P.F. Brain
Ethopharmacology Laboratory, School of Psychology, University of Leeds, Leeds LS2 9JT, UK
Institute of Experimental Medicine, H-1450 Budapest, P.O. Box 67, Hungary
School of Biological Sciences, University of Wales Swansea, Singleton Park, Swansea SA2 8PP, UK
Received 31 March 1999; accepted 14 July 1999
Exposure to the elevated plus-maze induces behavioural and physiological effects in rodents consistent with fear/anxiety. Maze-naive
animals display high levels of risk assessment towards the open arms, and explore these areas less extensively than other parts of the maze
while, immediately following the test, pain latencies, skin conductance levels, and plasma corticosterone titres (CORT) are significantly
elevated. Although previous research has suggested a link between the plasma CORT response and open-arm exploration, significant el-
evations in CORT have also been found with restricted exposure to the closed arms. The present study employed ethological measures in
an attempt to further characterise the relationship between behavioural and CORT responses to this widely used animal model of anxi-
ety. Our results confirm that, relative to home-cage controls, 5-min exposure to the plus-maze significantly increases plasma CORT levels
in test-naive male Wistar rats and male Swiss–Webster mice. Furthermore, in both species, the CORT response was found to be highly
correlated with measures of risk assessment (mice:
0.87; rats:
0.58), but not with measures of open-arm activity (entries,
time), general locomotor activity, rearing, or head dipping. Findings are discussed in relation to the functional significance of risk assess-
ment in potentially dangerous situations and the potential involvement of glucocorticoids in this process. © 1999 Elsevier Science Inc.
All rights reserved.
Elevated plus-maze; Plasma corticosterone; Stretched-attend postures; Risk assessment; Rats; Mice
1. Introduction
The elevated plus-maze is one of the most widely used
models for the study of anxiety-related processes in animals
[1–3], and has been validated for both rats [4] and mice [5].
In this test, anxiety is typically measured by indices of
open-arm avoidance and general locomotor activity by the
frequency of closed-arm entries (e.g., [5–7]). More recently,
behavioural scoring has been extended to include a variety
of specific acts and postures which, incorporated into more
comprehensive factor analyses, have revealed additional di-
mensions to plus-maze behaviour patterns, for example,
vertical activity, directed exploration, decision making and
risk assessment [6,8–13]. Inclusion of these ethological
measures permits comprehensive drug profiling in the
maze, thereby facilitating the identification of behaviourally
selective actions [2]. Furthermore, measures of risk assess-
ment (primarily, stretched-attend postures) have proved ex-
tremely valuable in identifying anxiolytic-like actions of
drugs (e.g., 5-HT
receptor ligands) not detected by con-
ventional scoring methods (e.g., [14–16]).
Physiological confirmation of the aversive or stressful
effects of plus-maze exposure is provided by evidence of
posttest elevations in pain latencies (e.g., [3]) and skin con-
ductance levels [17]. Furthermore, consistent with the ef-
fects of various physical and psychosocial challenges (for
review, see [18,19]), exposure to the plus-maze produces a
significant increase in plasma corticosterone (CORT) titres
[4,20–24]. Although this response is greater following con-
finement to an open arm [4], closed-arm confinement also
increases CORT levels relative to home-cage controls
[4,21]. Although these data suggest that a novelty-induced
CORT response to the plus-maze is exacerbated by forced
exposure to the more aversive open arms, they do not clarify
the relationship between CORT and spontaneous behaviour
in this test. The aim of the present study was to address this
* Corresponding author. Tel:
44-(0)113-233-5745; Fax:
E-mail address:
Page 1
R.J. Rodgers et al. / Physiology & Behavior 68 (1999) 47–53
issue by exploiting the behavioural detail provided by etho-
logical analysis [2,3]. Furthermore, in view of possible spe-
cies differences, the work involved parallel studies on mice
and rats conducted, respectively, at the University of Wales
Swansea (UK), and the Institute of Experimental Medicine
(Budapest, Hungary).
2. Materials and methods
2.1. Animals
Subjects were adult male Swiss–Webster mice (25–33 g,
supplied by Bantin & Kingman, Hull, UK), and adult male
Wistar rats (350–400 g, supplied by Charles River, Buda-
pest, Hungary). Animals of both species were group housed
(mice: 10/cage, 25
13 cm; rats: 6/cage, 50
20 cm), and maintained in temperature (mice: 21
rats: 23
C) and humidity (mice: 52
2%, rats: 60
10%)-controlled environments under 12 h reversed-light cy-
cles (mice: lights off, 0700 h; rats: lights off, 1100 h). Food
and water were freely available, with the exception of the
brief test periods.
2.2. Apparatus and procedure
Rat and mouse elevated plus-mazes, as well as the asso-
ciated test procedures, were based on those previously vali-
dated and now routinely used in the collaborating laboratories
(mice: [11], rats: [18]). The mouse plus-maze comprised
two open arms (30
0.25 cm) and two enclosed arms
15 cm) that extended from a common central
platform (5
5 cm). The apparatus was constructed from
black (floor) and clear (side and end walls of closed arms)
Plexiglas, and was elevated on a wooden pedestal to a
height of 60 cm above floor level. Open-arm exploration
was encouraged by the inclusion of a slight raised edge
around their perimeter (0.25 cm) and by testing under dim
red light (4
60 W indirect). The rat plus-maze comprised
two open arms (50
20 cm) and two enclosed arms (50
30 cm) that extended from a common central platform
20 cm). The apparatus was made from wood (painted
dark grey), and elevated to a height of 80 cm above floor
level. Testing was conducted under dim red illumination (1
40 W, indirect).
All testing was conducted during the dark phase of the
light/dark cycle, and commenced with the placement of sub-
jects (mice:
9; rats:
17) on the central platform of
the maze facing either an open (mice) or closed (rats) arm.
A conventional test duration of 5 min was employed for
both species and, in each series, the maze was thoroughly
cleaned between subjects. Mouse sessions were recorded on
videotape for subsequent analysis, while rat sessions were
scored live from a videosignal relayed to a monitor in a
nearby room. Immediately following testing, subjects were
transported to an adjacent laboratory where they were killed
by rapid decapitation, and trunk blood was collected. Con-
trol blood samples were collected from subjects of the same
age/housing conditions (mice:
12; rats:
10) taken
directly from their home cages.
2.3. Corticosterone assay
2.3.1. Mice
Trunk blood was collected in polypropylene tubes con-
taining citrate anticoagulant, centrifuged for 5 min at room
temperature, and stored at
C until analysed. Prior to
assay, CORT was displaced from corticosterone binding
globulin (CBG) by heating samples to 60
C for 30 min. Re-
agents (Amersham Life Science Limited, UK) were placed
into tubes in duplicate: standards (0.078–200 ng corticoster-
one in 100
L) contained 100
L rabbit anticorticosterone serum, and test samples
contained 100
L total serum volume
L rabbit anti-
corticosterone serum
I-corticosterone. All tubes were
thoroughly vortexed, covered, and left to incubate for 2 h at
room temperature, following which 400
L Amerlex-M
second antibody reagent was added and left to incubate for a
further 10 min. The antibody bound fraction was magneti-
cally separated for 15 min, supernatant discarded, and tubes
allowed to drain for 5 min; radioactivity in each tube was
determined by
-scintillation counting for 60 s. The sensi-
tivity of the assay was 0.2 pmol, and CORT levels were ex-
pressed as ng/mL plasma.
2.3.2. Rats
Trunk blood was sampled on ice-cold EDTA-containing
plastic tubes, and stored at
C until analysed. The assay
was performed in duplicate, and plasma CORT levels mea-
sured by radioimmunoassay [25,26]. Prior to assay, CORT
was separated from CBG at low pH (200
L 0.1 mol citric
acid; 1-h incubation at room temperature). Antiserum was
raised in rabbits against corticosterone carboxymethyl-
oxime bovine serum albumin (prepared at the Institute of
Experimental Medicine, Budapest), and
carboxymethyloxime-tyrosine-methyl ester (Isotope Insti-
tute Ltd, Budapest) was used as tracer. Eight CORT solu-
tions were used as standards (0.2 pmol/10
L–40 pmol/10
L), and test samples contained 10
L total serum volume
of 10
L CORT standard, 100
L antiserum, and 200
tracer. Following the addition of cold polyethyleneglycol
and incubation for 3 h at 4
C, samples were centrifuged at
C and radioactivity measured in the deposits. The sensi-
tivity of the assay was 1 pmol, and CORT levels were ex-
pressed as ng/mL plasma.
2.4. Behavioural analysis
Behaviours were scored by highly trained observers (in-
trarater reliability
0.90) using ethological analysis soft-
ware (mice: ‘Hindsight’ [27]; rats: ‘EthoVision,’ Noldus,
The Netherlands). Scoring was based on measures validated
and routinely used for the two species in the UK (mice:
[11]) and Hungarian (rats: [18]) laboratories.
For mice, behavioural parameters comprised both con-
ventional spatiotemporal and more recently developed etho-
Page 2
R.J. Rodgers et al. / Physiology & Behavior 68 (1999) 47–53
logical measures [11]. Conventional measures were the fre-
quencies of total, open and closed entries (arm entry
four paws into an arm), % open entries [(open/total)
and % time spent in open, closed, and central parts of the
maze [e.g., (time open/session duration)
100]. Ethologi-
cal measures comprised frequency scores for rearing (verti-
cal movement against the side and/or end of the walls; N.B.
mice very rarely exhibit unsupported rearing), head dipping
(exploratory movement of head/shoulders over the side of
the maze), and stretched-attend postures (SAP: exploratory
posture in which the body is stretched forward then re-
tracted to the original position without any forward locomo-
tion), as well as total duration scores in seconds for rearing
and grooming (licking, scratching, and washing of the head
and body). In view of the importance of thigmotactic cues to
rodent exploration in the plus-maze (e.g., [3]), head dipping
and SAP were also differentiated as a function of their oc-
currence in different parts of the maze. Thus, the closed arms
and centre platform were designated as “protected” areas
(i.e., offering relative security) and the “percent protected”
scores for head dipping and SAP calculated as the percent-
age of these behaviours displayed in or from the protected
areas (e.g., [(protected SAP/total SAP)
For rats [18], conventional behavioural parameters (re-
corded automatically) were the frequency of open-, closed-,
and total arm entries (arm entry
middle of the animal in
an arm), the ratio of open-arm entries (i.e., open/total), and
the percentage of time spent in the open arms [(time open/
session duration)
100]. Ethological measures, recorded
manually using the event-recorder facilities of Ethovision,
were protected head dipping (investigating over the sides of
the maze) and protected stretched-attend postures (SAP; in-
vestigating open arms from the central platform or a closed
arm). Both behaviours were recorded as frequencies and du-
rations, with the latter expressed in terms of percent time
spent engaged in their display [(time in seconds/session du-
2.5. Statistics
As appropriate for the characteristics of the respective
datasets, comparisons between plasma CORT levels in
home-cage controls and plus-maze–exposed animals were
performed by the Mann–Whitney
-test (mice) and inde-
-test (rats). Relationships between plasma CORT
levels and behavioural measures were examined using
Spearman’s Rank-Order correlations.
2.6. Ethics
Experiments conducted in the UK were performed under
licence in accordance with the Animals (Scientific Proce-
dures) Act 1986, while those conducted in Hungary were
performed in accordance with the Helsinki agreement on
animal experimentation.
3. Results
The behavioural profiles of Swiss–Webster mice and
Wistar rats are summarised in Table 1. In both species, 5-min
exposure to the elevated plus-maze produced significant in-
creases in plasma CORT levels in comparison to home-cage
controls: mice: control median
8.5 (range 2.6–18.6) ng/
mL, exposed
21.3 (12.8–23.7) ng/mL,
0.05; and rats:
control mean
12.3 mg/mL, exposed 227.2
0.01. Correlations between plasma CORT lev-
els and individual plus-maze parameters are summarised in
Table 1 and Fig. 1. For mice, significant correlations were
obtained only for percent protected SAP (
and % closed-arm time (
0.03) while, for rats,
the only significant correlation observed was for % time
spent displaying protected SAP (
4. Discussion
Present results confirm that test-naive Swiss–Webster
mice and Wistar rats not only avoid the open arms of an ele-
vated plus-maze but also display a range of other behav-
iours including rearing, head dipping, and stretched-attend
postures (e.g., [11,18]). However, despite these basic simi-
larities, there are some notable differences, for example, al-
Table 1
Correlations between plasma corticosterone levels and plus-maze
behaviours in male Swiss–Webster mice (
n 5 9) and male Wistar rats
(n 5 17)
Spearman’s R
Mouse plus-maze
Total arm entries 17.67 6 1.67 0.07, n.s.
Open-arm entries 6.89 6 0.82 0.37, n.s.
Closed-arm entries 10.78 6 1.04 0.02, n.s.
% Open entries 38.64 6 2.84 0.52, n.s.
% Open-arm time 20.19 6 2.39 20.32, n.s.
% Closed-arm time 19.55 6 2.03 20.72, p 5 0.03
% Centre time 60.26 6 3.69 0.62, n.s.
Rear (frequency) 7.89 6 2.12 20.38, n.s.
Rear (duration) 3.65 6 1.18 20.40, n.s.
Head dips 12.00 6 1.85 0.42, n.s.
% Protected head dips 76.54 6 4.72 0.57, n.s.
SAP 14.56 6 1.68 0.28, n.s.
% Protected SAP 70.79 6 4.83 0.87, p , 0.01
Grooming (duration) 1.56 6 1.03 20.52, n.s.
Rat plus-maze
Total arm entries 10.58 6 1.86 20.18, n.s.
Open-arm entries 3.11 6 0.84 20.18, n.s.
Closed-arm entries 7.47 6 1.12 20.11, n.s.
Ratio open entries 0.23 6 0.04 20.14, n.s.
% open-arm time 3.95 6 1.63 20.20, n.s.
% closed-arm time 82.60 6 3.29 0.18, n.s.
% centre time 13.41 6 2.53 20.12, n.s.
Head dips (frequency) 4.29 6 0.76 20.12, n.s.
% Time head dipping 2.45 6 0.55 20.20, n.s.
SAP (frequency) 12.70 6 1.24 0.33, n.s.
% Time protected SAP 9.77 6 1.05 0.58, p 5 0.02
Behavioral data are expressed as mean values 6 SEM. SAP 5
stretched attend posture.
Page 3
50 R.J. Rodgers et al. / Physiology & Behavior 68 (1999) 47–53
though displaying similar absolute levels of SAP, rats
showed a stronger aversion to the open arms (% time open)
as well as lower levels of locomotor activity (closed and to-
tal arm entries) and directed exploration (head dipping).
Furthermore, although both species clearly avoided the
open arms, mice spent the greater proportion of their time
on the centre platform, whereas rats spent most of their time
in the enclosed arms. These rat–mouse differences are
rather similar to those previously reported for Long–Evans
rats and Swiss–Webster mice tested in a visible burrow situ-
ation [28,29]. Thus, when briefly exposed to a cat, rats re-
treated to the tunnels (where they showed high levels of
freezing) whereas, following initial flight, mice persistently
returned to tunnel openings where they displayed high lev-
els of risk assessment. Although consistent with a possible
species difference in defensive strategy, marked strain dif-
ferences in the plus-maze profiles of both species [3] would
urge caution in interpreting present data in terms of a simple
species difference.
In agreement with previous findings [4,20–24], and rela-
tive to home-cage controls, brief exposure to the plus-maze
resulted in a significant elevation of plasma CORT levels in
both species. Furthermore, the CORT response of both spe-
cies was remarkably similar in magnitude, i.e., relative to
conspecific home-cage control values, the percent CORT
increase was approximately 151% for mice and 153% for
rats. Nevertheless, it is obvious that control CORT levels
were markedly different for the two species (8.5 ng/mL for
mice versus 89.9 ng/mL for rats). This discrepancy cannot
readily be attributed to differences in housing conditions
(group housing for both species), light cycle (reversed for
both species), time of sampling (early–middark phase for
both species), sampling method (rapid decapitation for both
species), handling (minimal for both species), or exposure
to distress vocalisations/odours (in both cases, animals were
sacrificed in a location entirely separate from holding rooms
and behavioural laboratories). However, while the CORT
assays employed were very similar in most respects, differ-
ences in the method used to separate CORT from CBG (for
mice, heating; for rats, lowering of pH) may have at least
partially contributed to the observed discrepancy in basal
values. In this context, it is important to note that the exist-
ing literature confirms wide variation in control CORT val-
ues for both species, but offers no clear explanation for such
differences. Thus, control values ranging from ,10 mg/mL
to .300 ng/mL (e.g., [4,20,22,23,30,31,32,61]) and from ,5
ng/mL to .40 ng/mL (e.g., [7,33–35]) have been reported
for rats and mice, respectively. As such, the home-cage con-
trol values recorded in the present study are not in any way
atypical of previous findings and, although very different
for the two species, clearly permitted the detection of statis-
tically significant increases in steroid levels in response to
plus-maze exposure.
In their now classic article on the validation of the plus-
maze as a model of anxiety in rats, Pellow and colleagues
[4] reported that confinement to either an open or an en-
closed arm of the maze produced a significant increase in
plasma CORT relative to home-cage controls. However, in
support of behavioural observations indicating a stronger
emotional response to the open arms, the CORT response to
the open arm was significantly greater than that to the en-
closed arm. Nevertheless, plasma CORT levels were signif-
icantly increased in response to the closed arms per se (see
also [21]), while, in the present study, a similar response
was evident in Wistar rats following a free exploration trial
characterised by very low levels of open-arm exploration.
Such observations suggest a relationship between this endo-
crine stress marker and a behavioural response other than
open arm activity per se. In agreement with this inference,
present results failed to show (for either species) a signifi-
cant correlation between plasma CORT levels and any mea-
sure of open-arm exploration. It should perhaps be noted
that although a correlation of 10.52 was observed between
CORT and % open-arm entries (mice), this was neither sta-
tistically significant nor supported by a low negative corre-
lation (20.32) for % open-arm time. Furthermore, the
CORT response did not correlate with indices of general lo-
Fig. 1. Correlations between risk assessment and plasma corticosterone
response to the elevated plus-maze. (A) Percent protected SAPs and
plasma corticosterone levels in maze-naive male Swiss–Webster mice (n 5
9); and (B) percent time in protected SAPs and plasma corticosterone
levels in maze-naive male Wistar rats (n 5 17). SAP 5 stretched attend
Page 4
R.J. Rodgers et al. / Physiology & Behavior 68 (1999) 47–53 51
comotor activity (closed entries), vertical activity (rearing),
or directed exploration (head dipping). Rather, for both spe-
cies, CORT levels correlated highly and positively with
stretched-attend postures (SAP). In neither instance, how-
ever, did this correlation involve the absolute frequency of
SAP; for mice, it specifically involved the proportion of
SAP displayed in or from protected areas of the maze and,
for rats, the proportion of time spent in protected SAP di-
rected specifically towards the open arms. As such, they key
behavioural association with the CORT response to the
maze appears to be aborted attempts to enter the open arms
rather than actual exploration of these potentially more dan-
gerous/threatening parts of the apparatus. The additional
(though negative) correlation observed in mice (but not rats)
between CORT and % time spent in the closed arms (rela-
tive safety vs centre platform and open arms) may relate to
the already noted difference in the spatiotemporal prefer-
ences displayed by the two species.
The stretched attend posture (SAP) was first documented
by Grant and Mackintosh [36], who referred to “stretched
attention” as an ambivalent behaviour (i.e., reflecting ap-
proach–avoid conflict). It has subsequently been identified
and further characterised in a variety of threatening situa-
tions, including social, predatory, and nonsocial contexts
[28,29,37,38]. On the basis of their elegant studies on anti-
predator behaviour in rats, Blanchard and Blanchard [28]
interpreted SAP in terms of “risk assessment,” i.e., informa-
tion-gathering behaviours displayed in potentially threaten-
ing situations, the function of which is to optimize the most
adaptive behavioural strategy (see also [39]). Consistent
with this view, subsequent research has shown that risk as-
sessment measures, and SAP in particular, are very sensi-
tive to the effects of anxiolytic and anxiogenic drugs (e.g.,
[15,16,40–44]). As such, present data would be consistent
with a strong positive relationship between CORT and risk
assessment in rats and mice exposed to the plus-maze.
Recently, McNaughton [45] has proposed a hippocampal
hyperactivity model of anxiety-related disorders in which
the hippocampus is seen as playing a critical role in risk as-
sessment. As this structure is a major CNS target of CORT
(for review: [46]), the strong association between CORT
and risk assessment in the plus-maze may reflect steroid fa-
cilitation of information-processing (and, thus, learning) in
novel and potentially dangerous environments. In this con-
text, it is already known that spatial learning occurs very
quickly in the plus-maze. Thus, an initially high level of risk
assessment and an absence of open/closed preference rap-
idly (i.e., within 2–3 min) gives way to much lower levels of
risk assessment and unambiguous open-arm avoidance
[11,47,48,49]. In view of this temporal pattern, however, it
seems somewhat improbable that the currently observed
correlations between CORT and risk assessment reflect an
activating influence of the steroid on information gathering.
First, CORT measurements in the present study were taken
from samples collected 2–5 min after plus-maze exposure,
whereas SAP scores reflect behaviour during the actual test.
Second, while the catecholamine response to stress is fast,
the glucocorticoid response is much slower (e.g. [50]) and
seems an unlikely mediator of SAP, the highest levels of
which are seen at the beginning of the test. Third, although
exogenously-administered (pharmacological) doses of CORT
have been reported to increase [20,51] or decrease [52,53]
plus-maze anxiety, metyrapone-induced inhibition of CORT
synthesis does not affect open-arm avoidance, suggesting
that endogenous CORT may not be directly involved in me-
diating spontaneous behavioural responses in this test [20].
Although not completely negating a role for CORT in the
mediation of risk assessment (information gathering), the
temporal dynamics of behaviour and steroid response would
be more consistent with a reverse causal link, i.e., the higher
the level of risk assessment (i.e., the greater the number of
aborted entries onto the open arms), the larger the CORT re-
sponse. In this context, it is pertinent to note that anticipa-
tion of threatening events in humans produces as marked an
elevation in cortisol as the event itself [54], leading to spec-
ulation that it may be cognitive differences in risk assess-
ment that may distingush subjects who score highly on trait
anxiety [55]. Although the present findings may simply re-
flect the acute homeostatic role of CORT in the face of a
perceived threat [46], they clearly indicate that this percep-
tion is based not on actual exploration of a dangerous envi-
ronment [4], but rather, the detection of its existence. Fur-
thermore, in view of growing evidence from animal and
human studies for the involvement of glucocorticoids in
cognitive function [31,56], it is tempting to speculate on the
potential involvement of CORT (and especially its action at
hippocampal GR and MR receptors) in the within- and be-
tween-trials “emotional” learning that is characteristic of the
plus-maze paradigm [11]. Research directed specifically at
this issue may help to resolve some of the already noted incon-
sistencies in the effects of exogenous CORT on plus-maze be-
haviour, the rather puzzling effects of GR and MR antagonists
in different animal models of anxiety (e.g. [57–59]), and the
apparently contradictory behavioural profile in such tests of
transgenic mice with impaired GR-mediated feedback inhibi-
tion of hypothalamic–pituitary–adrenal activity [60].
In summary, the present study has identified in two spe-
cies a high correlation between the plasma CORT response
to the elevated plus-maze and measures of risk assessment
in this test. The behavioural specificity and robustness of
this relationship is all the more remarkable, given the intrin-
sic methodological differences between the rat and mouse
components of the study. In view of the theoretical impor-
tance of the main finding, further work appears warranted
including 1) replication using fully identical methodologies,
and 2) clarification of the dynamics of the CORT response.
The authors express thanks to the University of Wales
Swansea (School of Biological Sciences) and the University
of Leeds (School of Psychology) for financial support of
Page 5
52 R.J. Rodgers et al. / Physiology & Behavior 68 (1999) 47–53
this research. Work in Hungary was supported by OTKA
Grant No. T 025844.
[1] Hogg S. A review of the validity and variability of the elevated plus-
maze as an animal model of anxiety. Pharmacol Biochem Behav
[2] Rodgers RJ. Animal models of anxiety: Where next? Behav Pharma-
col 1997;8:477–496.
[3] Rodgers RJ, Cole JC. The elevated plus-maze: Pharmacology, meth-
odology and ethology. In: Cooper SJ, Hendrie CA, editors. Ethology
and Psychopharmacology. Chichester: John Wiley & Sons Ltd, 1994.
pp. 9–44.
[4] Pellow S, Chopin P, File SE, Briley M. Validation of open:closed arm
entries in an elevated plus-maze as a measure of anxiety in the rat. J
Neurosci Methods 1985;14:149–167.
[5] Lister RG. The use of a plus-maze to measure anxiety in the mouse.
Psychopharmacology (Berlin) 1987;92:180–185.
[6] File SE. Behavioural detection of anxiolytic action. In: Elliott JM,
Heal DJ, Marsden CA, editors. Experimental Approaches to Anxiety
and Depression. London: John Wiley & Sons Ltd; 1992. pp. 25–44.
[7] Trullas R, Skolnick P. Difference in fear motivated behaviors among
inbred mouse strains. Psychopharmacology (Berlin) 1993;111:323–331.
[8] Cruz APM, Frei F, Graeff FG. Ethopharmacological analysis of rat
behavior on the elevated plus-maze. Pharmacol Biochem Behav
[9] Espejo EF. Structure of mouse behavior in the elevated plus-maze
test of anxiety. Behav Brain Res 1997;86:105–12.
[10] Fernandes C, File SE. The influence of open arm ledges and maze ex-
perience in the elevated plus-maze. Pharmacol Biochem Behav 1996;
[11] Holmes A, Rodgers RJ. Responses of Swiss–Webster mice to re-
peated plus-maze experience: Further evidence for a qualitative shift
in emotional state? Pharmacol Biochem Behav 1998;60:473–488.
[12] McCreary AC, McBlane JW, Spooner A, Handley SL. 5-HT systems
and anxiety: Multiple mechanisms in the elevated X-maze. Pol J
Pharmacol 1996;48:1–12.
[13] Rodgers RJ, Johnson NJT. Factor analysis of spatiotemporal and
ethological measures in the murine elevated plus-maze test of anxi-
ety. Pharmacol Biochem Behav 1995;52:297–303.
[14] Cole JC, Rodgers RJ. Ethological evaluation of the effects of acute
and chronic buspirone treatment in the murine elevated plus-maze
test: Comparison with haloperidol. Psychopharmacology (Berlin)
[15] Griebel G, Rodgers RJ, Perrault G, Sanger DJ. Risk assessment be-
haviour: evaluation of utility in the study of 5-HT-related drugs in the rat
elevated plus-maze test. Pharmacol Biochem Behav 1997;57:817–27.
[16] Setem J, Pinheiro AP, Motta VA, Morato S, Cruz APM. Ethopharma-
cological analysis of 5-HT ligands on the rat elevated plus-maze.
Pharmacol Biochem Behav 1999;62:515–521.
[17] Suer C, Dolu N, Ozesmi C, Sahin O, Ulgen A. The relation between
skin conductance level and plus-maze behaviour in male mice. Phys-
iol Behav 1998;64:573–576.
[18] Haller J, Halasz J, Makara GB, Kruk MR. Acute effects of glucocorti-
coids: Behavioral and pharmacological perspectives. Neurosci Biobe-
hav Rev 1998;23:337–44.
[19] Koolhaas JM, de Boer SF, de Rutter AJH, Meerlo P, Sgoifo A. Social
stress in rats and mice. Acta Physiol Scand 1997;161:69–72.
[20] Calvo N, Martijena ID, Molina VA, Volosin M. Metyrapone pretreat-
ment prevents the behavioral and neurochemical sequelae induced by
stress. Brain Res 1998;800:227–35.
[21] Copland AM, Balfour DJK. Spontaneous activity and brain 5-hydroxy-
indole levels measured in rats tested in two designs of elevated X-maze.
Life Sci 1987;41:57–64.
[22] File SE, Johnston AL, Baldwin HA. Anxiolytic and anxiogenic
drugs: Changes in behaviour and endocrine responses. Stress Med
[23] File SE, Zangrossi H, Sanders FL, Mabbutt PS. Raised corticosterone
after exposure to the elevated plus-maze. Psychopharmacology (Ber-
lin) 1994;113:543–6.
[24] Holmes A, Diffley EP, Walton TJ, Brain PF, Rodgers RJ. Lack of ha-
bituation of corticosterone response in mice repeatedly exposed to the
plus-maze. J Psychopharmac 1998;12S:32.
[25] Elenkov IL, Kovacs K, Kiss J, Bertok L, Vizi ES. Lipopolysaccharide
is able to bypass corticotropin-releasing factor in affecting plasma
ACTH and corticosterone levels: Evidence from rats with lesions of
the paraventricular nucleus. J Endocrinol 1992;133:231–6.
[26] Gomez-Sanchez C, Milewich L, Holland OB. Radioiodinated deriva-
tives for steroid immunoassay. Application to the radioimmunoassay
of cortisol. J Lab Clin Med 1977;80:902–9.
[27] Weiss SM. Pharmacological and behavioural examination of the de-
fensive reactions of laboratory mice to the calls of the Tawny Owl
(Strix aluco). PhD Thesis, University of Leeds, 1995.
[28] Blanchard RJ, Blanchard DC. Antipredator defensive behaviors in a
visible burrow system. J Comp Psychol 1989;103:70–82.
[29] Blanchard RJ, Parmigiani S, Bjornson C, Masuda C, Weiss SM,
Blanchard DC. Antipredator behavior of Swiss–Webster mice in a
visible burrow system. Aggress Behav 1995;21:123–36.
[30] Blanchard RJ, Nikulina JN, Sakai RR, McKittrick C, McEwen BS,
Blanchard DC. Behavioral and endocrine change following chronic
predatory stress. Physiol Behav 1998;63:561–9.
[31] de Quervain DJ-F, Roozendaal B, McGaugh JL. Stress and glucocor-
ticoids impair retrieval of long-term spatial memory. Nature 1998;
[32] Turner SW, Wen C, Li M, Fraser TB, Whitworth JA. Adrenocorti-
cotropin dose–response relationships in the rat: haemodynamic, met-
abolic and hormonal effects. J Hypertens 1998;16:593–600.
[33] Azpiroz A, Fano E, Garmendia L, Arregi A, Cacho R, Beitia G, Brain
PF. Effects of chronic mild stress (CMS) and imipramine administra-
tion, on spleen mononuclear cell proliferative response, serum corti-
costerone level and brain norepinephrine content in male mice. Psy-
choneuroendocrinology 1999;24:345–61.
[34] Chapman JC, Christian JJ, Pawlikowski MA, Michael SD. Analysis
of steroid hormone levels in female mice at high population density.
Physiol Behav 1998;64:529–33.
[35] Coleman MA, Garland T, Marler CA, Newton SS, Swallow JG,
Carter PA. Glucocorticoid response to forced exercise in laboratory
house mice (Mus domesticus). Physiol Behav 1998;63:157–64.
[36] Grant EC, Mackintosh JH. A comparsion of the social postures of
some common laboratory rodents. Behaviour 1963;21:246–59.
[37] Kaesermann H-P. Streteched attend posture, a non-social form of am-
bivalence, is sensitive to a conflict-reducing drug action. Psychophar-
macology (Berlin) 1986;89:31–37.
[38] van der Poel AM. A note on “stretched attention,” a behavioural ele-
ment indicative of an approach–avoidance conflict in rats. Anim Be-
hav 1979;27:446–450.
[39] Pinel JPJ, Mana MJ. Adaptive interactions of rats with dangerous in-
animate objects: Support for a cognitive theory of defensive behavior.
In: Blanchard RJ, Brain PF, Blanchard DC, Parmigiani S, editors.
Ethoexperimental Approaches to the Study of Behavior. Dordrecht:
Kluwer Aacdemic Publications, 1993. pp. 137–150.
[40] Blanchard RJ, Yudko EB, Rodgers RJ, Blanchard DC. Defense sys-
tem psychopharmacology: An ethological approach to the pharmacol-
ogy of fear and anxiety. Behav Brain Res 1993;58:155–65.
[41] Grewal SS, Shepherd JK, Bill DJ, Fletcher A, Dourish CT. Behav-
ioural and pharmacological characterisation of the canopy stretched
attend posture test as a model of anxiety in mice and rats. Psycho-
pharmacology (Berlin) 1997;133:29–38.
[42] Molewijk HE, van der Poel AM, Olivier B. The ambivalent behaviour
“stretched approach posture” in the rat as a paradigm to characterize
anxiiolytic drugs. Psychopharmacology (Berlin) 1995;121:81–90.
[43] Rodgers RJ, Cole JC, Cobain MR, Daly P, Doran PJ, Eells JR, Wallis
Page 6
R.J. Rodgers et al. / Physiology & Behavior 68 (1999) 47–53 53
P. Anxiogenic-like effects of fluprazine and eltoprazine in the mouse
elevated plus-maze: Profile comparisons with 8-OH-DPAT,
CGS12066B, TFMPP and mCPP. Behav Pharmacol 1992;3:621–634.
[44] Shepherd JK, Grewal SS, Fletcher A, Bill DJ, Dourish CT. Behav-
ioural and pharmacological characterisation of the elevated zero-
maze as an animal model of anxiety. Psychopharmacology (Berlin)
[45] McNaugthon N. Cognitive dysfunction resulting from hippocampal
hyperactivity—A possible cause of anxiety disorder? Pharmacol Bio-
chem Behav 1997;56:603–611.
[46] Meijer OC, de Kloet ER. Corticosterone and serotonergic neurotrans-
mission in the hippocampus: Functional implications of central corti-
costeroid receptor diversity. Crit Rev Neurobiol 1998;12:1–20.
[47] Rodgers RJ, Johnson NJT, Carr J, Hodgson TP. Resistance of experi-
entially-induced changes in murine plus-maze behaviour to altered
retest conditions. Behav Brain Res 1997;86:71–77.
[48] Rodgers RJ, Johnson NJT, Cole JC, Dewar CV, Kidd GR, Kimpson
PH. Plus-maze retest profile: Importance of initial stages of trial 1
and response to post-trial cholinergic receptor blockade in mice.
Pharmacol Biochem Behav 1996;54:41–50.
[49] Shepherd JK. Preliminary analysis of an elevated “zero-maze” as a
model of anxiety in laboratory rats. Paper delivered to the British As-
sociation for Psychopharmacology, Cambridge, July 1992.
[50] De Boer SF, Koopmans SL, Slangen JL, van der Gugten J. Plasma
catecholamine, corticosterone and glucose responses to repeated
stress in rats: Effects of interstressor interval. Physiol Behav 1990;47:
[51] Smythe JW, Murphy D, Timothy C, Costall B. Exogenous cortico-
sterone administration increases anxiety in rats. Br J Pharmacol 1996;
[52] Andreatini R, Leite JR. The effect of corticosterone in rats submitted
to the elevated plus-maze and pentylenetetrazol-induced convulsions.
Prog Neuropsychopharmacol Biol Psychiatry 1994;18:1333–47.
[53] McBlane JW, Handley SL. Effects of two stressors on behaviour in
the elevated X-maze: Preliminary investigation of their interaction
with 8-OH-DPAT. Psychopharmacology (Berlin) 1994;116:173–182.
[54] Mason JW. “Overall” hormonal balance as a key to endocrine organi-
sation. Psychosom Med 1968;30:791–808.
[55] File SE. Interactions of anxiolytic and antidepressant drugs with hor-
mones of the hypothalamic–pituitary–adrenal axis. In: File SE, editor.
Psychopharmacology of Anxiolytics and Antidepressants. New York:
Pergamon Press; 1991. pp. 29–55.
[56] Roozendaal B, Bohus B, McGaugh JL. Dose-dependent suppression
of adrenocortical activity with metyrapone: Effects on emotion and
memory. Psychoneuroendcocrinology 1996;21:681–693.
[57] Korte SM, de Boer SF, de Kloet ER, Bohus B. Anxiolytic-like effects
of selective mineralocorticoid and glucocorticoid antagonists on fear-
enhanced behaviour in the elevated plus-maze. Psychoneuroendocri-
nology 1995;20:385–394.
[58] Korte SM, Korte-Bouws GAH, Koob GF, de Kloet ER, Bohus B.
Mineralocorticoid and glucocorticoid receptor antagonists in animal
models of anxiety. Pharmacol Biochem Behav 1996;54:261.
[59] Smythe JW, Murphy D, Timothy C, Costall B. Hippocampal mineral-
ocorticoid, but not glucocorticoid, receptors modulate anxiety-like
behaviour in rats. Pharmacol Biochem Behav 1997;56:507–513.
[60] Rochford J, Beaulieu S, Rousse I, Glowa JR, Barden N. Behavioral
reactivity to aversive stimuli in a transgenic mouse model of impaired
glucocorticoid (type II) receptor function: Effects of diazepam and
FG-7142. Psychopharmacology (Berlin) 1997;132:145–152.
[61] Weinstock M, Poltyrev T, Schorer-Apelabum D, Men D, McCarty R.
Effects of prenatal stress on plasma corticosterone and catechola-
mines in response to footshock in rats. Physiol Behav 1998;64:439–44.
Page 7
  • Source
    • "Conventional measures such as the percentage of time spent in the open arms [calculated and analysed as an index for generalized-anxiety. In addition, we also measured the animal " s general locomotor activity by analysing the frequency of closed arms entries [32] and other ethological measures (which can be considered " risk assessment behaviours " ), including frequency of head dips (downward movement of rat " s head toward the floor from the open arms), rearing (vertical standing of rat on two hind legs against the walls of the closed arms) and stretchattend posture (when the rat stretched forward, without moving its paws, from the closed arms toward the centre and open arms). These parameters are a more sensitive index of anxiety-related behaviours than conventional measures [33,34]. "
    [Show abstract] [Hide abstract] ABSTRACT: The beneficial effects of Environmental Enrichment (EE) applied immediately after weaning or even in adulthood have been widely demonstrated. Less is known about the possible changes in behaviour and brain development of the progeny following the exposure of dams to EE. In order to further investigate this matter, female rats were reared in EE for 12 weeks, from weaning until delivery. After having confirmed the presence of relevant behavioural effects of EE, both control and EE females underwent mating. Maternal behaviour was observed and male and female offspring were then administered a battery of behavioural test at different ages. EE mothers showed a decreased frequency of total nursing and, during the first 2 days of lactation, an increase in licking/grooming behaviour. Maternal exposure to EE affected offspring behaviour in a sex-specific manner: social play behaviour and anxiety-like behaviour were increased in males but not in females and learning ability was improved only in females. As a general trend, maternal EE had a marked influence on motility in male and female offspring in both locomotor activity and swimming speed. Overall, this study highlights the importance of environmental stimulation, not only in the animals directly experiencing EE, but for their progeny too, opening the way to new hypothesis on the heritability mechanisms of behavioural traits.
    Full-text · Article · May 2016 · Physiology & Behavior
    • "For instance, exposure to unconditioned fear tests was found to increase levels of corticosterone (for a review see [1]). However, while some studies questioned the relationship between corticosterone levels and open field behaviors (for a review see: [3]), others found significant relationships between corticosterone levels and the expression of risk assessment behaviors [16] or interindividual differences in anxiety [17]. Few studies used simultaneous recording of autonomic responses (i.e. "
    [Show abstract] [Hide abstract] ABSTRACT: Due to their direct inaccessibility, affective states are classically assessed by gathering concomitant physiological and behavioral measures. Although such a dual approach to assess emotional states is frequently used in different species including humans, the invasiveness of procedures for physiological recordings particularly in smaller-sized animals strongly restricts their application. We used infrared thermography, a non-invasive method, to assess physiological arousal during open field and elevated plus maze tests in mice. By measuring changes in surface temperature indicative of the animals' emotional response, we aimed to improve the inherently limited and still controversial information provided by behavioral parameters commonly used in these tests. Our results showed significant and consistent thermal responses during both tests, in accordance with classical physiological responses occurring in stressful situations. Besides, we found correlations between these thermal responses and the occurrence of anxiety-related behaviors. Furthermore, initial temperatures measured at the start of each procedure (open field, elevated plus maze), which can be interpreted as a measure of the animals' initial physiological arousal, predicted the levels of activity and of anxiety-related behaviors displayed during the tests. Our results stress the strong link between physiological correlates of emotions and behaviors expressed during unconditioned fear tests.
    No preview · Article · Feb 2016 · Physiology & Behavior
  • Source
    • "Thus, in order to reduce the chances that our results could be affected by differences in the motility of the animals (due to exercise-induced fatigue), we used ethological measures (stretch attend postures and unprotected head dips) as well as the conventional ratio " time in the open arms vs. all arms " . Stretch attend postures are ethologically relevant risk assessment behaviours displayed in potentially threatening situations [46, 47] and are correlated with corticosterone response [48]. Two-way ANOVA revealed that single-housed mice spent significantly more time stretching [F (1,28) = 111.4 "