Tail-suspension induced hyperthermia: a new measure of stress reactivity

Journal of Psychiatric Research 37 (2003) 249–259 www.elsevier.com/locate/jpsychires

Tail-suspension induced hyperthermia: a new measure of stress reactivity Xiaoqing Liu, Dorothy Peprah, Howard K. Gershenfeld* Department of Psychiatry, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd., Dallas, TX 75390-9070, USA Received 18 June 2002; received in revised form 30 December 2002; accepted 2 January 2003

Abstract The tail suspension test (TST), an antidepressant screening paradigm, uses the uncontrollable, inescapable stressor of tail suspension to elicit immobility. As hyperthermia occurs following numerous stressors, hyperthermia might exist following the TST. We tested whether tail suspension induced hyperthermia (TSIH) was a distinct variable for TST. Hyperthermia was measured by two methods: a rectal probe and a subcutaneously implanted microchip (ELAMSTM). In outbred ICR male mice, TSIH was robustly demonstrated compared to control (No-TST) mice. TSIH peaked after TST and remained elevated at 120 min. Among five (129/SvEvTac, A/J, C57BL/6J, NMRI and ICR) strains examined for TSIH, significant strain variations were detected. NMRI showed the highest temperature rise (2.3  C) and A/J mice showed the lowest (0.6  C). Sex differences were found for the C57BL/6J and NMRI strains on TSIH. TSIH and duration of immobility were not significantly correlated (r=0.22, P=0.17) in outbred mice. Both duration of TST immobility and TSIH were measured when ICR male mice were administered diazepam, imipramine (a TCA antidepressant), venlafaxine (a SNRI antidepressant), sertraline and paroxetine (SSRI antidepressants), propranolol and nadolol (b-adrenergic receptor blockers), CP-154,526 (a CRF1 receptor antagonist), and indomethacin (a cyclo-oxygenase inhibitor). Diazepam dose-dependently increased immobility and decreased TSIH. Propranolol blocked TSIH, but nadolol had no effect. Antidepressants showed more complex patterns of effects with venlafaxine, sertraline, and paroxetine inhibiting TSIH. TSIH demonstrated inter-strain variability, sex differences and a distinct pharmacology, suggesting that TSIH provides an independent, robust physiologic parameter to supplement the TST paradigm. This TSIH method may prove useful for pharmacologic, transgenic, and mechanistic studies. # 2003 Elsevier Science Ltd. All rights reserved. Keywords: Hyperthermia; Stress; Sex differences; Strain differences; Tail suspension test; Emotion

1. Introduction The regulation of body temperature is a highly evolved, physiologically regulated, and easily measurable trait with a complex network of homeostatic regulators (Boulant, 1997). While studies of fever have documented the roles of thermosensitive neurons in the hypothalamus, pyrogenic cytokines, and the pharmacology of anti-pyretic drugs (Mackowiak, 1997), the phenomena and etiology of ‘‘psychogenic fever’’ or emotional hyperthermia remain less well understood [reviewed (Oka et al., 2001)]. * Corresponding author. Tel.: +1-214-648-7380; fax: +1-214-6485599. E-mail address: [email protected] (H.K. Gershenfeld).

Emotional or stress induced hyperthermia (SIH) is the rise of body temperature following exposure to psychological stress and has been demonstrated across species (mice, rats, pigs and humans) (Borsini et al., 1989; Briese, 1995; Briese and De Quijada, 1970; Briese et al., 1991b; Groenink et al., 1994, 1995; Hajos and Engberg, 1986; Hasan and White, 1979; Kleitman, 1945; Lecci et al., 1990b; Renburn, 1960; Van der Heyden et al., 1997; Zethof et al., 1995). In humans, anticipatory anxiety seems sufficient to induce hyperthermia (Briese, 1995; Renburn, 1960). This rise in temperature provides one dimension of an acute stress response that varies among individuals. In rodents, mild psychological stressors inducing hyperthermia include placing animals in novel environments (e.g. an open field arena), restricting an animal’s activity (e.g. restraint tube), noise, and handling animals in a variety of ways (e.g. cage change). In

0022-3956/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0022-3956(03)00004-9

250

X. Liu et al. / Journal of Psychiatric Research 37 (2003) 249–259

the ‘‘stress-induced hyperthermia’’ (SIH) paradigm, group-housed mice (e.g. 10–15 mice per cage) are sequentially removed at 1-min intervals and rectal temperatures are measured immediately upon removal (Lecci et al., 1990b; Zethof et al., 1994). While the first three mice show little change in temperature, by the 10th animal (10 min), there is a robust increase in temperature, which is maintained for 60-min after the procedure. Also, a singly housed mouse version of this technique has been developed with repeated disturbance as the stressor, offering an advantage when mice are a limiting resource (Van der Heyden et al., 1997). These robust SIH phenomenon in mice have been validated pharmacologically and interpreted as a stress reaction. While studying the tail suspension test (TST), a wellvalidated antidepressant screening test (Porsolt et al., 1987; Steru et al., 1987), we hypothesized that hyperthermia might be induced in response to this uncontrollable, inescapable stressor. The rise in rectal temperature following tail suspension (i.e. tail suspension induced hyperthermia, TSIH) might also provide a simple method for assessing individual differences in acute stress responses in mice. The TST paradigm hangs a mouse by its tail for 6-min. A typical response in this paradigm is struggling alternating with passive immobility. The duration of immobility is accumulated throughout the 6-min period and temperature is taken after tail suspension. This duration of immobility has been the principal measure in the TST and this immobility is interpreted as a measure of ‘‘behavioral despair’’. Traditionally, antidepressants decrease the duration of immobility. These initial experiments tested the validity and utility of an acute rise of body temperature following the TST as a measure of stress reactivity. Further experiments attempted to validate the paradigm by (1) demonstrating genetic and sex variation in TSIH responses and (2) characterizing TSIH pharmacologically, distinguishing this hyperthermic measure as a unique aspect of the TST response.

accessible (available in the USA), Swiss derived, general purpose outbred strain was selected to reduce cost and to provide generalizability. ICR and 129S6 strains were obtained from Taconic (Germantown, NY). A/J and C57BL/6J strains were obtained from the Jackson Laboratory (Bar Harbor, ME). Inbred NMRI mice were obtained from B & K Universal Ltd. (E. Yorkshire, UK). Mice arrived at 7 weeks of age. F1 mice were bred from 129S6 dams and NMRI sires at UTSW. These F1(NMRIx129) mice were part of an ongoing genetic experiments on the inheritance of the TSIH and provided a readily available isogenic strain. All mice were ear-notched or implanted with ELAMSTM transponders (microchips) for identification, and permitted 1 week for acclimation to their new housing conditions prior to testing. All mice were housed in groups of four with food and water freely available. This housing density (four /cage) was empirically determined from preliminary experiments, showing that this cage density minimized the hyperthermic effect due to sequential removal. The animals were maintained under a 12 h:12 h light:dark cycle with lights on at 06:00 h. For the pharmacological experiments, all mice were naı¨ve and used just one time. All experiments followed the NIH Guide for Care and Use of Laboratory Animals and were approved by the local animal institutional committee along with their suggestions for reduction of animal usage.

2. Materials and methods

2.3. Body temperature

2.1. Animals

A digital thermometer (Ret-3 rectal probe and a Thermalert TH-5 thermometer, Physitemp, Clifton, NJ) was used in these experiments. Rectal temperature was measured by inserting the probe 2 cm into the rectum of the mouse, while gently restrained manually. The probe was dipped into silicon oil before insertion and was held in the rectum for about 20 s, permitting readings to stabilize. Temperatures were determined to the nearest 0.1  C. In order to minimize the handling stress associated with measuring core rectal temperature, the ELAMSTM (Electronic Laboratory Animal Monitoring System by BioMedic Data Systems, Inc. Seaford, DE) was used to

One outbred strain (ICR) and four inbred strains (129/SvEvTac=129S6, A/J, C57BL/6J, and NMRI) of mice were used in these experiments. Inbred strains were selected to demonstrate strain differences as a way to validate the paradigm (Porsolt, 2000). The choice of particular inbred strains was based on the strains commonly used for transgenic analyses (129 and C57BL6) and those with sequenced DNA genomes (129, A/J, C57BL/6J). The Swiss-derived NMRI strain was included as the most commonly used and robust responding strain in the literature for the TST paradigm. A readily

2.2. Tail suspension Automated TST devices (Med Associates Inc., St. Albans, VT) were used to measure the duration (sec) of immobility in the tail suspension test as described in detail (Liu and Gershenfeld, 2001). Mice were suspended by the tail with tape to the apparatus. The apparatus automatically calculated the total duration of immobility during a 6-min TST period. After each trial, the testing cubicles were cleaned to remove olfactory cues with a diluted solution of a deodorizing detergent.

X. Liu et al. / Journal of Psychiatric Research 37 (2003) 249–259

measure subcutaneous body temperature. This system consists of a notebook (DAS-5002, a portable data acquisition system), a probe attached to the notebook, and programmable and implantable microchips (IPTT-200). The transponders were programmed with identification numbers (ID) prior to implantation. Upon arrival (7 weeks of age), mice were subcutaneously implanted with microchips using the manufacturers’ insertor as described (Kort et al., 1998). Implanted transponders were read by placing the probe within a distance of 5 cm. The probe read both body temperature and ID, while the animal was presented to the probe in a plastic tray (10 cm wide13 cm long3 cm high). This plastic tray was used to support all animal transfers during experiments. 2.4. Drugs Drug doses were judiciously selected upon review of the existing literature, including the effective dosages while avoiding sedative effects. Diazepam (RBI/ Sigma, Natick, MA) was suspended in 1PBS with 2% ethanol and 2% propylene glycol. Imipramine HCl, propranolol HCl and nadolol (RBI/ Sigma, Natick, MA) were dissolved in pyrogen-free saline (0.9% NaCl). Venlafaxine HCl (Wyeth-Ayerst, Princeton, NJ) and sertraline HCl (Zoloft, 20 mg/ml, Pfizer, New York, NY) were diluted in sterile water. Paroxetine HCl oral suspension (2 mg/ ml) (SmithKline Beecham, Philadelphia, PA) was diluted in sterile water for gavage. CP-154,526 (Pfizer, New York, NY) was suspended in 0.1% methylcellulose solution immediately before injection. Indomethacin (RBI/ Sigma, Natick, MA) was suspended in sterile water with 1% ethanol and 1% Tween 20. 2.5. Experimental procedures All animals were brought to the testing room at least 1 h prior to testing, and remained in the same room throughout the test. The testing room’s ambient temperature was between 21.3 and 23.4  C. Mice were tested individually between 13:00 and 17:00 h at 8 weeks of age. 2.5.1. Time course of TSIH Only male ICR mice were used. In the ELAMSTM experiment, pre-TST (baseline) temperature was measured before the experiment for all mice. The mice cages were randomly divided into two groups: (1) TST group (n=12) and (2) No-TST (control) group (n=11). The TST group underwent 6-min tail suspension in groups of four (a cage), followed by temperature measurement (time interval=0 min). The No-TST (control) group was not given 6-min of tail suspension, while temperature was measured (time interval=0 min) at the same time points as the TST group. Since the procedure of ELAMSTM measurement is non-invasive, a within-subject

251

design was used, permitting collection of temperatures from individuals at all time intervals. Hence, temperatures were repeatedly measured at time intervals of 15, 30, 45, 60 and 120 min after TST or No-TST. For those investigators without access to ELAMSTM microchip instrumentation and for correspondence between methods, the time course of TSIH was examined using a rectal probe. In the rectal probe experiment, male ICR mice (n=40) were tested. Before TST, pre-TST (baseline) temperature was taken for the first mouse in each cage (n=10). Then, all mice (total n=30) except for the first in each cage were given 6-min of tail suspension, followed by rectal temperature measurement (post-TST; time interval=0 min). A between-group design subsequently divided these mice by cage into five groups (n=6 mice per group), measuring separate groups’ rectal temperature at 15, 30, 45, 60 or 120 min. To examine the correspondence between rectal and ELAMSTM measurements of the
T induced by TST, isogenic F1 (129NMRI) mice (n=61) were measured at pre-TST (baseline) and post-TST using both rectal and ELAMSTM measures. Rectal measurement was preceded by ELAMSTM measurement. 2.5.2. Variation by strain and sex in TSIH Temperature measurements were by rectal probe only. Inbred 129S6 (n=36), A/J (n=54), C57BL/6J (n=40), NMRI (n=41) and outbred ICR (n=36) strains were included in this experiment. Both sexes were tested. Mice were divided into two groups (TST and No-TST) for each sex and each strain. The TST mice were tested with 6-min tail suspension, followed by measurement of rectal temperature. The No-TST mice were measured by rectal temperature without being tested on TST to provide baseline temperatures for that sex and strain. In order to avoid the confound of elevated baseline temperatures due to sequential removal of mice from each cage (i.e., stress induced hyperthermia), only the 1st and the 2nd mice of each cage in a separate group (N=8–14 mice; 4–8 cages) were measured for baseline temperatures. 2.5.3. Pharmacologic testing and validation All mice were naı¨ve and used just one time. All drugs were administered i.p. 30 min prior to testing, except for paroxetine, which was gavaged 60 min prior to testing. Pre-TST or pre-No-TST rectal temperatures were taken before the administration of drugs as baseline temperatures. In order to test the specific hypothesis of drugs inhibiting TSIH and to reduce animal use, drugs were initially tested on TST with groups of 8–12 mice per dose. If the drug was found to block TSIH, groups (N=8–12) of No-TST mice with more mice for TST groups were included to examine the hypothermic effect of those doses of the drug. The TST mice were tested on the TST apparatus for 6 min, and then rectal temperatures

252

X. Liu et al. / Journal of Psychiatric Research 37 (2003) 249–259

were measured immediately following TST. The No-TST groups’ temperatures were taken exactly as the TST mice without testing by TST. 2.6. Statistics For the TSIH time course experiment, change in temperature (
T) was obtained by subtracting the preTST (baseline) temperature from temperatures at other time intervals. Mixed two-way ANOVA tested the pattern of time course data collected via ELAMSTM, using TST/No-TST as a between-subject factor and time interval as a within-subject factor (pre-TST, 0, 15, 30, 45, 60 and 120 min) with
T as the dependent variable. One-way repeated measures ANOVA tested the time course data from TST or No-TST groups collected using ELAMSTM and simple contrasts were performed between pre-TST (baseline) and other time intervals for each data series to observe the recovery of TSIH of ELAMSTM. One-way ANOVA tested data collected using rectal probe, using time intervals as an independent variable. For the F1 mice experiment, the rectal vs. ELAMs methods of temperature measurements were compared in an isogenic population.
T was defined as the difference between pre-TST and post-TST temperatures, measured by ELAMSTM or rectal probe. A Pearson correlation was performed of
T obtained via ELAMSTM vs
T obtained via rectal temperatures. We tested for differences in TSIH by varying strain (genotype) and sex. A No-TST (baseline) group controls for handling. In these experiments,
T was obtained by subtracting the appropriate No-TST group mean for each sex and strain from the TST temperature of each mouse of the same sex and strain. One-way ANOVA tested for between-strain differences, using strain as an independent variable and mean
T as a dependent variable. For each strain, Student t-test examined sex difference. To test for strain differences in No-TST (baseline) temperatures, strain was used as the independent variable and NoTST (baseline) temperature as the dependent variable in one-way ANOVAs. One-way ANOVAs were used to test the order effect of sequential removal of mice from their home cages after TST for each strain. A Pearson correlation was performed between TSIH and duration of immobility in 40 ICR mice to observe the association of these two measures taken in this TST paradigm. In the pharmacological experiments, the duration of immobility was the total time duration of being immobile during the 6-min tail suspension testing period. Using dose as an independent variable, one-way ANOVA tested dose effects on the duration of immobility for each drug. The Bonferroni post hoc test was used to compare the duration of immobility at 0 mg/kg to immobility at other doses. Percentage change in immobility from 0 mg/kg(vehicle) was calculated by dividing the mean duration of immobility at other doses

by that of vehicle (0 mg/kg), multiplied by 100.
T was obtained by subtracting the baseline temperature from post-TST or post-No-TST temperature. Two-way ANOVA, using No-TST/TST and dose as independent variables, tested interaction effects on
T for the drug series in which both No-TST and TST groups were included. Simple contrasts compared
T at 0 mg/kg (vehicle) to
T at other doses in the No-TST and TST groups. One-way ANOVA, using dose as an independent variable, tested dose effect on
T in the TST group where only a TST group was included, followed by Bonferroni post hoc tests. Due to bimodal dose response on
T for the TST group in the venlafaxine experiment, one-way ANOVA was performed separately for the No-TST and TST groups.

3. Results The time course of TSIH (expressed as
T above preTST temperature) was measured with ELAMSTM microchips in outbred ICR male mice (Fig. 1). TST markedly increased body temperature (about 2  C) in the TST group, shown as 0-min interval. During the subsequent 120-min observation window, the TST group showed a decline in temperature but did not fully recover to pre-TST baseline. One way repeated measures ANOVA revealed a significant main effect of time interval [F (6, 66)=33.09, P < 0.001]. Simple contrasts indicated
T at 0, 15, 30 45, 60 and 120 min were significantly higher than pre-TST (baseline) (from P < 0.001 at 0 min to P=0.013 at 120 min). The minor handling stress of the No-TST group increased tem-

Fig. 1. Time course of tail suspension induced hyperthermia (TSIH) in outbred ICR male mice.
T was obtained by subtracting the pre-TST (baseline) temperature from temperatures at other time intervals (0, 15, 30, 45, 60 and 120 min). Pre-TST temperature was taken immediately before TST (TST group) or No-TST (No-TST group). Temperatures at subsequent time intervals (min) were taken at 0, 15, 30, 45, 60 and 120 min after TST / No-TST. TST-ELAMSTM (N=12) and No TSTELAMSTM (N=11) groups were measured with ELAMSTM microchips, while TST-Rectal (total N=40) were measured with a rectal probe.

X. Liu et al. / Journal of Psychiatric Research 37 (2003) 249–259

perature (about 0.9  C) at 0 min [F (6, 60)=7.32, P < 0.001], maintained this mild hyperthermia (simple comparisons did not show differences between 0 min and other intervals: 15, 30, 45, 60 min) with recovery to baseline 120 min later. The time course patterns for TST and No-TST groups were significantly different as indicated by a mixed two-way ANOVA, revealing a significant interaction effect between group and time interval [F (6, 126)=3.78, P < 0.01]. The time course of TSIH measured by rectal probe showed a similar pattern as that measured by ELAMSTM [F (6, 63)=18.47, P < 0.001], though
T was lower (Fig. 1). F1(129NMRI) mice (n=61) were simultaneously measured via ELAMSTM and rectal probes. The two methods demonstrated excellent reliability with no significant difference in means (
Trectal=1.73  C, S.D.=0.92  C and
TELAMS=1.80  C, S.D.=0.98  C) and with a correlation coefficient of 0.88 (P< 0.0001) between methods. To validate the paradigm’s sensitivity in detecting variation by strain and gender, we examined baseline temperature and TSIH of five strains divided by sex (Appendix). Baseline temperatures did not show variations among strains tested by one-way ANOVAs. Fig. 2 displays
T (the differences between TSIH and baseline temperature) for five strains split by sex. TSIH demonstrated marked variation across strains [F (4, 109) =73.24, P < 0.001]. The Swiss derived NMRI and ICR strains showed the greatest increases in
T, while A/J strain displayed the least. A significant sex difference was observed in C57BL/6J (t (20)=3.75, P < 0.01) and NMRI (t (18)=2.78, P=0.01) strains only. In housing four mice per cage, the potential confound of SIH in sequentially removing mice was examined for TSIH. In the strains tested, the effect of removal order on TSIH was not significant by ANOVAs. Among the outbred

Fig. 2. Strain and sex variations of tail suspension induced hyperthermia (TSIH).
T was obtained by subtracting the mean temperature of the No-TST group from the mean temperature of the TST mice. Asterisks indicate significant differences between genders. **P< 0.01, *P <0.05.

253

ICR male mice (n=40), the correlation coefficient between
T (difference between pre-TST and post-TST temperatures) and the duration of immobility was not significant (r=0.22, P=0.17). Distinct classes of drugs probed the mechanism of tail suspension induced hyperthermia (TSIH) and interrelationships between TST immobility and TSIH in ICR male mice (Table 1). Specifically, we tested the ability of drugs to inhibit the
T rise of TSIH. Diazepam, an anxiolytic agent via GABAA receptors, significantly lengthened the duration of TST immobility, producing a significant dose effect on immobility in a one-way ANOVA [F (3, 42)=14.24, P < 0.01]. Post hoc testing showed immobility at three doses (2.0, 3.5 and 5.0 mg/kg) were significantly longer than that for vehicle (0 mg/kg; P < 0.01). In contrast, diazepam decreased
T dosedependently for the TST group, while it did not affect
T for the No-TST control group, indicated by a significant two-way interaction effect between TST/No-TST and dose [F (2, 65)=6.31, P < 0.01]. Simple contrasts showed
T at 3.5 and 5.0 mg/kg of diazepam were significantly lower than that at 0 mg/kg in the TST group (P < 0.01). As expected for a tricyclic antidepressant, imipramine significantly shortened the duration of TST immobility [F (5, 42)=11.69, P < 0.01]. Imipramine had similar hypothermic effects in both No-TST and TST mice, shown by a main dose effect [F (5, 62)=9.75, P < 0.01]. The interaction between dose and
T was not significant [F (2, 62)=0.17, P=0.84]. Simple contrasts indicated
T at 30 and 60 mg/kg in both No-TST and TST groups were lower than those of controls (0 mg/kg; P < 0.01). Venlafaxine, a serotonin and norepinephrine reuptake inhibitor (SNRI), devoid of histaminergic and cholinergic antagonism, dose-dependently shortened immobility [F (5, 55)=11.45, P < 0.01] at doses of 8 mg/kg (P < 0.05), and 16, 24 and 32 mg/kg (P < 0.01). Venlafaxine was not hypothermic as it did not affect
T in the No-TST group. Venlafaxine did decrease
T in the TST group [F (5, 46)=3.38, P < 0.05] at 4 mg/kg (P < 0.01), 8 mg/kg (P < 0.05), 24 mg/kg (P < 0.05) and 32 mg/kg (P < 0.01), but not at 16 mg/kg. Sertraline, a selective serotonin reuptake inhibitor (SSRI), did not affect immobility. Nonetheless, sertraline decreased
T in the TST group, but not in the No-TST group, shown by a significant two-way interaction effect [F (2, 56)=4.46, P < 0.05). Simple contrasts indicated
T at 8, 16 and 24 mg/kg were lower than that at 0 mg/kg (P < 0.05 or P < 0.01). Surprisingly, paroxetine, another SSRI, affected neither immobility nor
T in ICR male mice. However, in inbred NMRI male mice, paroxetine (16 mg/kg p.o.) significantly blocked TSIH [F (1,18)=7.36, P=0.01) without significant effect on TST mobility (data not shown). Propranolol, a lipophilic, b1- and b2-adrenergic receptor antagonist, did not affect immobility. Propranolol decreased
T in the TST mice, but did not affect

254

X. Liu et al. / Journal of Psychiatric Research 37 (2003) 249–259

Table 1 Drug effects on TST immobility and tail-suspension induced hyperthermia (TSIH) in ICR male micea Drug

Dose (mg/kg)


T c( C)

TST immobility Immobility (s)

% from 0 mg/kgb

One-way ANOVA

0 2.0 3.5 5.0

163 226 256 257

(31) (46)** (31)** (42)**

100 139 157 158

F (3, 42)=14.24, P< 0.01

0.21 (0.94) – 0.11 (0.57) 0.14 (0.45)

2.08 1.96 1.03 0.49

(0.48) (0.73) (0.56)** (0.76)**

Interaction: F (2, 65)=6.31, P< 0.01

Imipramine

0 3.75 7.5 15 30 60

189 170 166 141 133 139

(53) (52) (86) (40)* (33)* (50)*

100 90 88 75 70 74

F (5, 42)=11.69, P< 0.01

0.10 (0.45) – – – 1.53 (0.78)** 1.80 (0.83)**

1.81 1.45 1.68 1.24 0.14 0.28

(0.85) (0.46) (0.76) (0.49) (1.69)** (1.30)**

Interaction: F (2, 62)=0.17, P=0.84

Venlafaxine

0 4 8 16 24 32

231 220 198 166 155 153

(29) (37) (30)* (31)** (33)** (36)**

100 95 86 72 67 66

F (5, 55 )=11.45, P< 0.01

0.54 0.30 0.49 0.29 – 0.14

(0.65) (1.1)** (0.40)* (1.06 ) (0.52)* (0.83)**

One-way for TST group: F (5, 46)=3.38, P< 0.05

(0.79)

2.05 1.04 1.22 1.53 1.00 0.86

Sertraline

0 8 16 24

207 217 203 241

(43) (41) (47) (44)

100 105 98 116

F (3, 35)=1.53, P=0.22

0.05 (0.59) – 0.27 (0.94) 1.35 (0.44)

2.23 1.50 0.98 1.43

(0.32) (1.07)* (0.86)** (0.77)*

Interaction: F (2, 56)=4.46, P< 0.05

Paroxetine

0 4 8 16

223 215 223 233

(29) (38) (30) (36)

100 96 100 105

F (3, 27)=0.37, P=0.77

– – – –

0.8 (1.1) 1.36 (1.24) 1.63 (0.80) 0.30 (1.53)

One-way: F (3, 27)=1.93, P=0.15

Propranolol

0 5 10 20

182 195 171 175

(28) (40) (45) (41)

100 107 94 96

F (3, 41 )=0.87, P=0.46

0.46 (0.45 ) – 0.98 (1.11) 0.33 (0.70)

2.54 2.00 1.93 0.33

Interaction: F (2, 68)=7.14, P< 0.01

Nadolol

0 10 20

207 (27) 188 (35) 207 (30)

100 91 100

F (2, 28 )=1.28, P=0.30

– – –

1.93 (0.68) 1.95 (0.73) 1.86 (0.78)

One-way: F (2, 27)=0.04, P=0.96

CP-154,526

0 10 20 40

190 198 201 211

(30) (39) (43) (44)

100 104 106 111

F (3, 94)=1.16, P=0.33

– – – –

1.97 2.01 1.77 1.91

One-way: F (3, 94)=1.22, P=0.31

Indomethacin

0 10

165 (36) 172 (33)

100 104

F (1, 31)=0.26, P=0.62

– –

2.20 (0.52) 2.77 (0.47)**

Diazepam

No-TST

(0.74) (0.37) (0.61) (0.78)

TST

ANOVA

(0.76) (0.78) (0.50) (1.49)**

(0.50) (0.60) (0.38) (0.40)

F (1, 31)=8.23, P< 0.01

a Values represent means (S.D.). Each group consisted of 8–12 mice/ dose. All drugs were administered i.p. 30 min prior to TST, except for paroxetine which was administered via gavage. The 0 mg/kg dose corresponds to vehicle injection. b % from 0 mg/kg was calculated over the immobility at 0 mg/kg dose level for each drug. One-way ANOVA tested the variations of immobility across dose levels. Asterisks in the column of Immobility (s) indicate the significance levels when compared to the control group (0 mg/kg), by using the Bonferroni post hoc tests (*P<0.05, **P<0.01). c
T was calculated by subtracting the mean of pre-TST or pre-No-TST temperature from post-TST or post-No-TST temperature. Two-way ANOVA tested interaction effects on
T between the factor of TST/No-TST groups and the factor of dose levels where a No-TST group was included. One-way ANOVA tested variations of DT for TST group where only TST group was included. Asterisks in the last column indicate the significance levels when compared to the control group (0 mg/kg) by applying simple contrasts (*P <0.05, ** P <0.01).


T in the No-TST mice, shown by a significant twoway interaction effects [F (2, 68)=7.14, P < 0.01]. Simple contrasts indicated
T at 20 mg/kg of propranolol in the TST mice was significantly lower than that at 0 mg/ kg (p < 0.01). Nadolol, the peripherally active, hydro-

philic non-specific b-blocker, showed no significant effect on either immobility or
T. CP-154,526, a non-peptide corticotropin-releasing factor-1 (CRF1) receptor antagonist, affected neither immobility nor
T. Indomethacin, a cyclo-oxygenase

X. Liu et al. / Journal of Psychiatric Research 37 (2003) 249–259

inhibitor, at a dosage (10 mg/kg) known to block fever mediated by prostaglandins, did not affect immobility, while showing an increased effect on
T [F (1, 31) =8.23, P < 0.01].

4. Discussion The extent of hyperthermia (TSIH) following the TST paradigm provides a simple and robust measure of a mouse’s response to the inescapable, uncontrollable stress in the TST. This TST-induced hyperthermia correlated poorly with the traditional measure of TST (viz., the duration of immobility). The distinct anti-immobility vs. anti-TSIH profiles of diazepam, propranolol and some antidepressants in the TST demonstrate the independence of these two measures. The blockade of TSIH by propranolol, yet not nadolol, suggests a central mechanism for this TSIH. TSIH offers a supplementary, unexplored dimension of the TST reflecting the multifaceted nature of stress responses. The rise in temperature to the stress of TST was rapid and transient. The degree of hyperthermia varied significantly across strains and by sex, validating sensitivity to detect known genetic and hormonal influences. This TST-induced hyperthermia may model some aspects of human emotional hyperthermia and may be interpreted as a ‘‘stress reactivity’’. These findings confirm prior assays of acute stressors inducing hyperthermia (Borsini et al., 1989; Briese and D Quijada, 1970; Briese et al., 1991b; Hajos et al., 1986; Keeney et al., 2001; Lecci et al., 1990b; Renburn, 1960; Rodgers et al., 1994; Zethof et al., 1994, 1995) and extend the finding of strain differences in temperature responses to stress (Brown and Julian, 1968; Hajos et al., 1986; Muraki and Kato, 1986; Slade et al., 1997). This paradigm facilitates phenotyping individual animals with a high throughput and exploits an existing model to obtain additional information. The physiologic pathways and mechanisms of this hyperthermia remains an unresolved issue (Briese and Cabanac, 1991a; Oka et al., 2001). The degree of resting (i.e. immobility vs. struggling) and the degree of TSIH were weakly correlated. In fact, the small contribution of struggling to the hyperthermia suggests a more CNS component to the behavior. Similarly, other paradigms of stress induced hyperthermia, including restraint stress, acoustic stress, repeated disturbance, novelty, handling, and serial removal of animals, involve little role for motor activity (Borsini et al., 1989; Briese et al., 1991b; Van der Heyden et al., 1997; Zethof et al., 1994, 1995). In rats, the novelty of an open field induced hyperthermia. This hyperthermia was blocked by propranolol via centrally acting non-selective beta adrenergic receptors and evidence suggested open field exposure induced a rise in plasma interleukin-6 (Mayfield et al.,

255

1999; Soszynski et al., 1996). Prior treatment with low dose X-irradiation (15 cGy) confined to the head region abolished SIH, favoring a CNS explanation (Miyachi, 2000). Finally, the emotional hyperthermia among boys (12–14 years) measured before engaging in amateur boxing was significant, yet 30 min of training exercises that appreciably raised their pulse, negligibly raised oral temperatures (Renburn, 1960). The inhibition of TSIH by propranolol, but not nadolol, also suggests that TSIH may largely be centrally mediated via the sympathetic autonomic system. These experiments harmonize with prior work where intra-cerebral stimulation of the rat paraventricular nucleus of the hypothalamus or the posterior hypothalamus can raise core body temperature, which can be inhibited by intra-cerebral propranolol (Amir, 1990a–c). Finally, neural activity (c-fos mapping) during stress demonstrated increased activity in the hypothalamic paraventricular nucleus and the locus ceruleus (Dayas et al., 2001; Lopez et al., 1999). This TSIH stress paradigm models and detects ‘‘individual differences’’ in stress reactivity as demonstrated by the distinctive stress responses between genders and among strains. The
T in response to TST varied by strain with the ICR, NMRI, C57BL/6J, 129S6 and A/J strains increasing by 2.3, 2.2, 1.6, 1.4, and 0.6  C respectively. Likewise, sex differences were found in some strains, but not all strains. While the 129S6, A/J, and ICR strains showed no sex differences, the C57BL/ 6J and NMRI strains exhibited significant differences (
Tmale >
Tfemale). This sex difference does not necessarily mean male mice were more responsive to the stress of TST. The male mice had lower basal temperature than the female mice and following TST both genders reached a comparable level of hyperthermia, which may represent a ceiling effect for temperature. The time course of TST-induced hyperthermia is consistent with a brief, discrete stressor resolving in greater than 120 min. The time course of the ‘‘stress-induced hyperthermia’’ (SIH) paradigm has a similar but distinct pattern (Zethof et al., 1994). In Zethof et al.’s study (1994), group-housed mice (n=10 per cage) were measured sequentially at 1-min intervals by rectal temperature. The SIH was defined as the difference between the average temperature of the first 3 mice (baseline) and the average temperature of the last three mice (stressinduced). The authors noted a peak temperature rise of 1.5  C at 10 min and recovery to the basal temperature 60 min after taking temperature. Also, singly housed mice recovered from ‘‘stress-induced hyperthermia’’ (1.2  C) 60 min following temperature measurement (Van der Heyden et al., 1997). ‘‘Stress induced hyperthermia’’ has been interpreted as being an anticipatory stress response (Borsini et al., 1989; Lecci et al., 1990b). In the TST paradigm, 6-min tail suspension induced a rise of temperature by about 2  C in the ICR strain.

256

X. Liu et al. / Journal of Psychiatric Research 37 (2003) 249–259

Two hours after TST, both ELAMSTM and rectal temperatures remained significantly elevated compared to the pre-TST level. The stimuli evoking TST induced hyperthermia are complex and poorly characterized. Perhaps, tail suspension was perceived as a more potent stressor than sequential removal from home cages, encompassing stronger somatosensory, orthostatic, novelty, and acrophobic components than mere handling stress. The time course of TSIH measured via rectal probe confirms the data obtained by microchip. The high correlation between ELAMSTM and rectal
T from isogenic F1 animals validates the rectal temperature method. The mediators of psychogenic hyperthermia remain unknown (Oka et al., 2001), but probably involves a complex integration of multiple afferent and efferent signals in the hypothalamus and pre-optic areas. In the TST and the TSIH paradigm, no data exists on the relative contributions of ACTH or glucocorticoids to hyperthermia. However, in two studies with other stress-induced hyperthermia paradigms the levels of ACTH and glucocorticoid were increased (Groenink et al., 1996; Moe and Bakken, 1997). Preliminary mechanistic studies have also suggested roles for 5-HT and GABA neurotransmitters (Borsini et al., 1993; Bouwknecht et al., 2000; Gouret et al., 1990; Groenink et al., 1995, 1996; Lecci et al., 1990a, 1991; LopezMendoza et al., 1998; Mendoza et al., 1999; Olivier et al., 1998; Vidal et al., 1983; Zethof et al., 1995), the sympathoadrenal system, and the hypothalamic-pituitary-adrenocortical (HPA) axis (Groenink et al., 1994; Stratakis and Chrousos, 1995; Yamano et al., 2000), and inflammatory associated cytokines (IL-1 and IL6)(Leon et al., 1997; Soszynski et al., 1996). The blockade of TSIH by drugs has a unique pattern, distinguishable from the SIH and TST paradigms. The TST has been demonstrated to have high predictive validity for antidepressant activity using the antiimmobility effect (Porsolt et al., 1987). Likewise, the SIH paradigm has been shown to be blocked by benzodiazepines and to a lesser extent by buspirone, but not by antidepressants or neuroleptics (Lecci et al., 1990a,b; Zethof et al., 1995). In contrast, the TSIH measured following 6-min of tail suspension was blocked by diazepam, propranolol, and some doses of antidepressants. The effects of acute antidepressants on TSIH demonstrated a more complex pattern, suggestive of neurotransmitter interactions and indirect actions, favoring a serotonergic component. While the SSRI antidepressant sertraline blocked the hyperthermia in ICR mice, paroxetine blocked TSIH only in the NMRI strain and not in the ICR strain. Venlafaxine, a SNRI, inhibits serotonin (5-HT) reuptake at low doses and at higher doses inhibits both serotonin and norepinephrine (NE) reuptake (Sanchez and Hyttel, 1999). The pattern of venlafaxine selectively antagonizing the hyperthermia at the

lowest dose (4 mg/kg) with minimal effect on immobility emphasizes that (1) hyperthermic responses can be distinguished from the immobility response and (2) 5-HT reuptake blockade alone can block hyperthermia. Also, prior work with low dose venlafaxine (1, 2, and 4 mg/kg) showed insignificant effects on locomotor activity and in the forced swim test. However, at a higher dose (16 mg/ kg), venlafaxine had no effect on TSIH, while at still higher doses (24 and 32 mg/kg) the blockade of hyperthermia returned along with the anti-immobility effect. Our TST immobility results extend and confirm prior results with venlafaxine in the TST and forced swim test, where efficacy was detected at 10 mg/kg s.c. and maximal at 40 mg/kg s.c. in the TST and at 8 mg/kg and maximal at 32 mg/kg in the FST (Millan et al., 2001; Redrobe et al., 1998). This dose response pattern might be interpreted as favoring a 5-HT reuptake blockade inhibiting TSIH at low doses, while increased efficacy occur only at higher doses when both 5-HT and NE reuptake blockade ensue. This TSIH dose–response pattern resembles the dose–response curve of venlafaxine in the four plates anxiety test (Hascoet et al., 2000). The restoration of TSIH inhibition may emerge from the ratio of 5-HT and NE reuptake blockade along with the distribution of pre- and post-receptors ultimately determining the net balance of effects and the extent of TSIH inhibition. Many stress responses result from the hypothalamic-pituitary-adrenal axis via the secretion of corticotropin-releasing factor (CRF). A CRF1 receptor antagonist affected neither TSIH nor TST immobility suggesting a limited role for this pathway. Likewise, the lack of blockade by indomethacin suggests little role for prostaglandins and distinguishes this hyperthermia from fever. Some limitations in these results should be noted. The rectal method of temperature measurement has high accuracy (  0.1  C.) while incurring some stress during the minimal restraint of measurement compared to the alternative of an implantable subcutaneous microchips with less accuracy ( 0.4  C) but no restraint stress. By utilizing the within individual
T as a measure to characterize TSIH, the accuracy of the microchip system is comparable to the rectal probe method. Second, we did not replicate the SSRI anti-immobility effects in TST, suggesting a diminished assay sensitivity to SSRI’s in ICR mice. Some groups have demonstrated an antiimmobility effect for SSRI’s in the tail suspension test acutely (Mayorga et al., 2001; Perrault et al., 1992; Teste et al., 1993; Ukai et al., 1998). Imipramine and venlafaxine gave robust responses providing a positive control. We used the ICR strain, whereas most European investigators used an outbred NMRI mice (not commercially available in the USA). Although ICR and NMRI are both strains Swiss derived outbred strains, subtle strain differences probably explain the discrepancy in results as others have found differences in

257

X. Liu et al. / Journal of Psychiatric Research 37 (2003) 249–259

antidepressant response by strain (Bai et al., 2001; Liu et al., 2001; Lucki et al., 2001; Porsolt et al., 1978; van der Heyden et al., 1987). Outbred NMRI mice seem to be the most sensitive to SSRI’s and this may explain our detection of TSIH blockade with inbred NMRI, but not ICR mice. Third, drugs with hyperthermic activity or devoid of activity in TSIH blockade were uninformative in dissecting the mediators of TSIH, obviating the need of the NO-TST/control. Clinically, the pharmacological profile of TSIH with blockade by diazepam, propranolol, and some antidepressants resembles the profile of drugs effective among individuals with generalized anxiety, social phobia and performance anxiety, particularly inhibiting the sensations of ‘‘feeling hot’’ and sweating. Interestingly, venlafaxine demonstrated the ability to prevent hot flashes in menopausal women surviving breast cancer (Barton et al., 2001; Loprinzi et al., 2000). While the etiology of hot flashes and TSIH remains unknown, parsimony might suggest a similar mechanism of action. Future areas for study include the mechanisms of hyperthermia, neural circuits, further validation with novel compounds and the role of chronic treatments. The most important finding from these experiments is the characterization and validation of the TSIH as a simple, additional, discrete stress paradigm for detecting stress reactivity. Anxiety-like behaviors encompass diverse endpoints (e.g. cognitive, psychological, somatic and behavioral components). We believe this TSIH taps a unique aspect of ‘‘anxiety-like’’ behavior by an innate, physiological response of unknown mechanism. This paradigm offers the advantages of being an easy, robust phenomenon without expensive equipment, enhancing an existing well validated TST paradigm. The interstrain variability and sex differences detected by TSIH further validates the paradigm (Porsolt, 2000), homologous to the known individual and gender differences in anxiety disorders. The hyperthermic response is blocked by diazepam, propranolol and some antidepressants, but not by indomethacin, a CRF1 receptor blocker, nor nadolol. Ideally, this additional TSIH measure may facilitate the screening of transgenic mice and of lead compounds possessing both anti-depressive and anti-anxiety efficacy via the TST.

Acknowledgements We appreciated helpful conversations with our UTSW colleagues, I. Lucki, P. Skolnick, V. Perrotta, and R. Mansbach. We are grateful to anonymous reviewers for their critiques. This project has been generously supported by a NARSAD Young Investigators Award (HG), the Southwestern Medical Foundation, the Seay Fellowship (XL), THECB ARP Award #010019-0081-1999, and NIH MH67211 (HG).

Appendix. For Web publication of supplementary data Mean baseline and TST temperatures for both male and female mice in four inbred strains (129S6, A/J, C57BL/6J and NMRI), and one outbred strain (ICR). All values were rectal measurements 129S6

A/J

C57BL/6J

ICR

NMRI

Baseline temp (M)

36.2 (0.50)

35.7 (0.52)

35.7 (0.53)

35.6 (0.40)

36.0 (0.72)

TST temp (M)

37.5 (0.43)

36.4 (0.45)

37.5 (0.41)

38.1 (0.32)

38.5 (0.49)

Baseline temp (F)

36.6 (0.56)

36.0 (0.40)

36.5 (0.60)

36.3 (0.38)

36.5 (0.77)

TST temp (F)

37.9 (0.22)

36.5 (0.41)

37.7 (0.32)

38.5 (0.35)

38.5 (0.35)

Standard deviations are included in parenthesis. N=8–14/per group.

References Amir S. Activation of brown adipose tissue thermogenesis by chemical stimulation of the posterior hypothalamus. Brain Res 1990a;534: 303–8. Amir S. Intra-ventromedial hypothalamic injection of glutamate stimulates brown adipose tissue thermogenesis in the rat. Brain Res 1990b;511:341–4. Amir S. Stimulation of the paraventricular nucleus with glutamate activates interscapular brown adipose tissue thermogenesis in rats. Brain Res 1990c;508:152–5. Bai F, Li X, Clay M, Lindstrom T, Skolnick P. Intra- and interstrain differences in models of ‘‘behavioral despair’’. Pharmacol Biochem Behav 2001;70:187–92. Barton D, Loprinzi C, Wahner-Roedler D. Hot flashes: aetiology and management. Drugs Aging 2001;18:597–606. Borsini F, Brambilla A, Cesana R, Donetti A. The effect of DAU 6215, a novel 5HT-3 antagonist, in animal models of anxiety. Pharmacol Res 1993;27:151–64. Borsini F, Lecci A, Volterra G, Meli A. A model to measure anticipatory anxiety in mice? Psychopharmacology 1989;98:207–11. Boulant JA. Thermoregulation. In: Mackowiak PA, editor. Fever: basic mechanisms and management. Philadelphia: LippincottRaven; 1997. p. 35–58. Bouwknecht JA, Hijzen TH, van der Gugten J, Maes RA, Olivier B. Stress-induced hyperthermia in mice: effects of flesinoxan on heart rate and body temperature. Eur J Pharmacol 2000;400:59–66. Briese E. Emotional hyperthermia and performance in humans. Physiol Behav 1995;58:615–8. Briese E, Cabanac M. Stress hyperthermia: physiological arguments that it is a fever. Physiol Behav 1991a;49:1153–7. Briese E, De Quijada MG. Colonic temperature of rats during handling. Acta Physiol Lat Am 1970;20:97–102. Briese E, Hui-Wan H, Parada MA. Stress hyperthermia in mice. J Therm Biol 1991b;16:333–6. Brown AM, Julian T. The body temperature response of two inbred strains of mice to handling, saline and amphetamine. Int J Neuropharmacol 1968;7:531–41. Dayas CV, Buller KM, Crane JW, Xu Y, Day TA. Stressor categorization: acute physical and psychological stressors elicit distinctive recruitment patterns in the amygdala and in medullary noradrenergic cell groups. Eur J Neurosci 2001;14:1143–52.

258

X. Liu et al. / Journal of Psychiatric Research 37 (2003) 249–259

Gouret CJ, Porsolt R, Wettstein JG, Puech A, Soulard C, Pascaud X, Junien JL. Biochemical and pharmacological evaluation of the novel antidepressant and serotonin uptake inhibitor 2-(3,4-dichlorobenzyl)-2-dimethylamino-1-propanol hydrochloride. Arzneimittelforschung 1990;40:633–40. Groenink L, Compaan J, van der Gugten J, Zethof T, van der Heyden J, Olivier B. Stress-induced hyperthermia in mice. Pharmacological and endocrinological aspects. Ann N Y Acad Sci 1995;771:252–6. Groenink L, van der Gugten J, Zethof T, van der Heyden J, Olivier B. Stress-induced hyperthermia in mice: hormonal correlates. Physiol Behav 1994;56:747–9. Groenink L, van der Gugten J, Zethof TJ, van der Heyden JA, Olivier B. Neuroendocrine effects of diazepam and flesinoxan in the stressinduced hyperthermia test in mice. Pharmacol Biochem Behav 1996; 54:249–54. Hajos M, Engberg G. Emotional hyperthermia in spontaneously hypertensive rats. Psychopharmacology 1986;90:170–2. Hasan MK, White AC. Psychogenic fever: entity or nonentity? Postgrad Med 1979;66:152–4. Hascoet M, Bourin M, Colombel MC, Fiocco AJ, Baker GB. Anxiolytic-like effects of antidepressants after acute administration in a four-plate test in mice. Pharmacol Biochem Behav 2000;65:339–44. Keeney AJ, Hogg S, Marsden CA. Alterations in core body temperature, locomotor activity, and corticosterone following acute and repeated social defeat of male NMRI mice. Physiol Behav 2001;74: 177–84. Kleitman N. The effect of motion pictures on body temperature. Science 1945;101:507–8. Kort WJ, Hekking-Weijma JM, TenKate MT, Sorm V, VanStrik R. A microchip implant system as a method to determine body temperature of terminally ill rats and mice. Lab Anim 1998;32:260–9. Lecci A, Borsini F, Gragnani L, Volterra G, Meli A. Effect of psychotomimetics and some putative anxiolytics on stress-induced hyperthermia. J Neural Transm Gen Sect 1991;83:67–76. Lecci A, Borsini F, Mancinelli A, D’Aranno V, Stasi MA, Volterra G, Meli A. Effect of serotoninergic drugs on stress-induced hyperthermia (SIH) in mice. J Neural Transm Gen Sect 1990a;82:219–30. Lecci A, Borsini F, Volterra G, Meli A. Pharmacological validation of a novel animal model of anticipatory anxiety in mice. Psychopharmacology 1990b;101:255–61. Leon LR, Kozak W, Peschon J, Glaccum M, Kluger MJ. Altered acute phase responses to inflammation in IL-1 and TNF receptor knockout mice. Ann N Y Acad Sci 1997;813:244–54. Liu X, Gershenfeld HK. Genetic differences in the tail-suspension test and its relationship to imipramine response among 11 inbred strains of mice. Biol Psychiatry 2001;49:575–81. Lopez JF, Akil H, Watson SJ. Neural circuits mediating stress. Biol Psychiatry 1999;46:1461–71. Lopez-Mendoza D, Aguilar-Bravo H, Swanson HH. Combined effects of Gepirone and (+)WAY 100135 on territorial aggression in mice. Pharmacol Biochem Behav 1998;61:1–8. Loprinzi CL, Kugler JW, Sloan JA, Mailliard JA, LaVasseur BI, Barton DL, et al. Venlafaxine in management of hot flashes in survivors of breast cancer: a randomised controlled trial. Lancet 2000;356:2059–63. Lucki I, Dalvi A, Mayorga AJ. Sensitivity to the effects of pharmacologically selective antidepressants in different strains of mice. Psychopharmacology (Berl) 2001;155:315–22. Mackowiak PA. Fever: Basic Mechanisms and Management. (2nd ed). Philadelphia: Lippincott-Raven; 1997. Mayfield KP, Soszynski D, Kozak W, Kozak A, Rudolph K, Kluger MJ. Beta-adrenergic receptor subtype effects on stress fever and thermoregulation. Neuroimmunomodulation 1999;6:305–17. Mayorga AJ, Dalvi A, Page ME, Zimov-Levinson S, Hen R, Lucki I. Antidepressant-like behavioral effects in 5-hydroxytryptamine(1A) and 5-hydroxytryptamine(1B) receptor mutant mice. J Pharmacol Exp Ther 2001;298:1101–7. Mendoza DL, Bravo HA, Swanson HH. Antiaggresive and anxiolytic

effects of gepirone in mice, and their attenuation by WAY 100635. Pharmacol Biochem Behav 1999;62:499–509. Millan MJ, Dekeyne A, Papp M, La Rochelle CD, MacSweeny C, Peglion JL, et al. S33005, a novel ligand at both serotonin and norepinephrine transporters: II. Behavioral profile in comparison with venlafaxine, reboxetine, citalopram, and clomipramine. J Pharmacol Exp Ther 2001;298:581–91. Miyachi Y. Disappearance of stress-induced hyperthermia following a low dose of X-irradiation: involvement of the vomeronasal system in the modulation of the radiation-induced effects. Br J Radiol 2000;73:51–7. Moe RO, Bakken M. Effects of handling and physical restraint on rectal temperature, cortisol, glucose and leucocyte counts in the silver fox (Vulpes vulpes). Acta Vet Scand 1997;38:29–39. Muraki T, Kato R. Strain difference in the effects of morphine on the rectal temperature and respiratory rate in male mice. Psychopharmacology 1986;89:60–4. Oka T, Oka K, Hori T. Mechanisms and mediators of psychological stress-induced rise in core temperature. Psychosom Med 2001;63: 476–86. Olivier B, Zethof TJ, Ronken E, van der Heyden JA. Anxiolytic effects of flesinoxan in the stress-induced hyperthermia paradigm in singlyhoused mice are 5-HT1A receptor mediated. Eur J Pharmacol 1998; 342:177–82. Perrault G, Morel E, Zivkovic B, Sanger DJ. Activity of litoxetine and other serotonin uptake inhibitors in the tail suspension test in mice. Pharmacol Biochem Behav 1992;42:45–7. Porsolt RD. Animal models of depression: utility for transgenic research. Rev Neurosci 2000;11:53–8. Porsolt RD, Bertin A, Jalfre M. ‘‘Behavioural despair’’ in rats and mice: strain differences and the effects of imipramine. Eur J Pharmacol 1978;51:291–4. Porsolt RD, Chermat R, Lenegre A, Avril I, Janvier S, Steru L. Use of the automated tail suspension test for the primary screening of psychotropic agents. Arch Int Pharmacodyn Ther 1987;288:11–30. Redrobe JP, Bourin M, Colombel MC, Baker GB. Dose-dependent noradrenergic and serotonergic properties of venlafaxine in animal models indicative of antidepressant activity. Psychopharmacology (Berl) 1998;138:1–8. Renburn ET. Body temperature and pulse in boys and young men prior to sporting contests. A study in emotional hyperthermia with a review of the literature. J Psychosomatic Res 1960;4:149–75. Rodgers RJ, Cole JC, Harrison-Phillips DJ. ‘‘Cohort removal’’ induces hyperthermia but fails to influence plus-maze behaviour in male mice. Physiol Behav 1994;55:189–92. Sanchez C, Hyttel J. Comparison of the effects of antidepressants and their metabolites on reuptake of biogenic amines and on receptor binding. Cell Mol Neurobiol 1999;19:467–89. Slade R, Watkinson WP, Hatch GE. Mouse strain differences in ozone dosimetry and body temperature changes. Am J Physiol 1997;272: L73–0L77. Soszynski D, Kozak W, Conn CA, Rudolph K, Kluger MJ. Betaadrenoceptor antagonists suppress elevation in body temperature and increase in plasma IL-6 in rats exposed to open field. Neuroendocrinology 1996;63:459–67. Steru L, Chermat R, Thierry B, Mico JA, Lenegre A, Steru M, Simon P, Porsolt RD. The automated Tail Suspension Test: a computerized device which differentiates psychotropic drugs. Prog Neuropsychopharmacol Biol Psychiatry 1987;11:659–71. Stratakis CA, Chrousos GP. Neuroendocrinology and pathophysiology of the stress system. Ann N Y Acad Sci 1995;771:1–18. Teste JF, Pelsy-Johann I, Decelle T, Boulu RG. Anti-immobility activity of different antidepressant drugs using the tail suspension test in normal or reserpinized mice. Fundam Clin Pharmacol 1993;7:219–26. Ukai M, Maeda H, Nanya Y, Kameyama T, Matsuno K. Beneficial effects of acute and repeated administrations of sigma receptor

X. Liu et al. / Journal of Psychiatric Research 37 (2003) 249–259 agonists on behavioral despair in mice exposed to tail suspension. Pharmacol Biochem Behav 1998;61:247–52. van der Heyden JA, Molewijk E, Olivier B. Strain differences in response to drugs in the tail suspension test for antidepressant activity. Psychopharmacology 1987;92:127–30. Van der Heyden JA, Zethof TJ, Olivier B. Stress-induced hyperthermia in singly housed mice. Physiol Behav 1997;62:463–70. Vidal C, Suaudeau C, Jacob J. Hyper- and hypothermia induced by nonnoxious stress: effects of naloxone, diazepam and gamma-acetylenic GABA. Life Sci 1983;33:587–90.

259

Yamano M, Yuki H, Yasuda S, Miyata K. Corticotropin-releasing hormone receptors mediate consensus interferon-alpha YM643induced depression-like behavior in mice. J Pharmacol Exp Ther 2000;292:181–7. Zethof TJ, Van der Heyden JA, Tolboom JT, Olivier B. Stress-induced hyperthermia in mice: a methodological study. Physiol Behav 1994; 55:109–15. Zethof TJ, Van der Heyden JA, Tolboom JT, Olivier B. Stress-induced hyperthermia as a putative anxiety model. Eur J Pharmacol 1995; 294:125–35.