Effect of intracerebroventricular administration of vasopressin on stress-induced hyperthermia in rats

Physiology & Behavior, Vol. 60, No. 2, pp. 417-424, 1996 Copyright © 1996 Elsevier Science Inc. Printed in the USA. All rights reserved 0031-9384/96 $15.00 + .00

PII S0031-9384(96) 00054-6

ELSEVIER

Effect of Intracerebroventricular Administration of Vasopressin on Stress-Induced Hyperthermia in Rats E. M . C L A U D I A

TERLOUW,

.1 S T E P H E N

KENT,t

SANDRINE

CREMONAt

AND

ROBERT

DANTZERt

* INRA, Centre de Clermont-Ferrand-Thebc, Station de Recherches sur la Viande, 63122 St-Genbs-Champanelle, France and "plNSERM U-394, Domaine de Carreire, Rue Camille St-Sa~ns, 33077 Bordeaux Cddex, France R e c e i v e d 9 A u g u s t 1995 TERLOUW, E. M. C., S. KENT, S. CREMONA AND R. DANTZER. Effect of intracerebroventricularadministration of vasopressin on stress-induced hyperthermia in rats. PHYSIOL BEHAV 60(2) 417-424, 1996.--Vasopressin has been reported to be an endogenous antipyretic peptide. The present study assessed whether this peptide has similar effects on stress-induced hyperthermia. Infusion of 3 ng of vasopressin into the lateral ventricle prior to a 40-min restraint stress reduced significantly the hyperthermic response of rats to this stress, compared to saline-injected controls. Half of the vasopressin-injected animals showed an immediate hypotherrnic response, with a significant reduction in body temperature of 0.34°C or more within 10 min; however, the effect of vasopressin on stress-induced hyperthermia remained significant after exclusion of these animals from the analysis. Administration of a V j receptor antagonist prior to the stress did not affect the hyperthermic response, which may suggest that the hyperthermic response had reached maximal (ceiling) levels. Administration of vasopressin, or of the V, receptor antagonist immediately after the stress, did not affect defervescence, suggesting that vasopressinergic systems are not implicated in the defervescence process. Thus, the results show that ICV admininstration of vasopressin reduces stress-induced hyperthermia. The mechanisms underlying the effects remain to be elucidated. Fever

Body temperature

Restraint stress

Stress-induced hyperthermia

T H E term stress-induced hyperthermia has been coined to describe the rise in body temperature following a psychological stress. Stress-induced hyperthermia may be induced by various stressors, such as exposure to an open field ( 2 0 ) , cage switching (24,30), or mild restraint, ( 4 1 ) in rats, and the anticipation of a school exam ( 2 6 ) or of a boxing match ( 3 7 ) in humans. Previous work has found that stress-induced hyperthermia is not merely a secondary effect of stress-induced physiological changes. For example, in rats, it has been found that during the rise in body temperature, heat retention occurs through vasoconstriction in the tail, suggesting a regulated change in body temperature set point ( 3 ) . It has further been shown that stress-induced hyperthermia may, similar to the febrile response, involve prostaglandin- and interleukin-based mechanisms. Thus, stressinduced hyperthermia was attenuated after IP or ICV administration of sodium salicylate or IP administration of indomethacin, both of which interfere with prostaglandin synthesis (20,40). Similarly, immobilization stress induced interleukin-1/3 m R N A in the rat hypothalamus ( 2 9 ) , and was accompanied by increased plasma levels of interleukin-6 ( 2 3 ) , whereas IP administration of the IL-1 receptor antagonist attenuated stress-induced hyperthermia ( 1 9 ) .

Vasopressin

The ventral septal area ( V S A ) is one of the thermoresponsive sites of the brain ( 1 2 ) . Arginine vasopressin ( A V P ) - c o n t a i n i n g neural pathways in the VSA, having their major source in the bed nucleus of the stria terminalis ( B S T ) ( 8 ) , appear to be involved in the downregulation of the febrile response. Initial p u s h - p u l l studies in febrile sheep found that in the VSA, less A V P was released when body temperature was rising, and conversely, more A V P was released when body temperature was falling ( 5 ) . Similar results have been found in urethane-anesthesized rats ( 2 1 ) . Depletion of A V P in the V S A by chemical lesions or following castration (9,10) resulted in augmented and/ or prolonged prostaglandin- and interleukin-induced fever responses (27,34), whereas ICV administration of A V P and electrical stimulation of the BST attenuated prostaglandin-induced fever (5,33). The action of A V P appears to be mediated by V~ type receptors, as the antipyretic effect of A V P administration to the V S A or of electrical stimulation of the B S T could be reversed by ICV administration of the V~ antagonist, whereas administration of a V2 antagonist was without effect (6,17,33). The aim of the present experiment was to investigate whether the vasopressinergic neurons in the ventral septal area may be involved in the regulation of stress-induced hyperthermia. A V P

' To whom requests for reprints should be addressed. 417

418

TERLOUW ET AL.

appears to have a role in thermoregulation during fever (5), as well as during defervescence (6,34). Therefore, the role of AVP in the rise of body temperature during exposure to a stressor, and in defervescence after the stress, was studied. To this aim, AVP or its VI type receptor antagonist were injected into a lateral ventricle of the brain, either prior to or after exposure to restraint stress. Work by Kasting and Wilkinson (18) has confirmed that AVP injected into the lateral cerebral ventricles of the rat can be detected in dialysates from the VSA. METHOD

Animals and Surgery Subjects were 42 laboratory-bred healthy, male Wistar rats, weighing 300-360 g at the beginning of the experiment. The animals were anaesthesized by IP injection of a mixture of ketamine hydrochloride (61 mg/kg) and xylazine (9 mg/kg), and placed in a stereotaxic apparatus, with the incisor bar at +5.0 ram. Three stainless steel jeweler's screws were threaded through holes drilled in the skull. A 23-gauge stainless steel guide cannula was implanted 1 mm above the right or left lateral ventricle of the brain (0.6 mm posterior, 2.0 mm lateral to the bregma, and 3.2 mm ventral to the skull surface) and secured with dental acrylic cement to the screws. A stylet from solid 30-gauge wire remained in the guide cannula between injections. A silastic covered temperature-sensitive radiotransmitter (Mini Mitter, Sunriver, OR) was implanted in the intraperitoneal cavity. After surgery, the animals were housed in 42 × 22 × 17 cm cages with ad lib access to water and standard rat chow. They were kept on a 12:12 light:dark cycle (lights on from 0800 to 2000 h) and weighed each morning at 0900 h. Ambient temperature was maintained at 22 ___2°C.

Experimentation Following surgery, a receiver was placed under each animal's cage and connected to a peripheral processor attached to an IBM computer, which was loaded with the Dataquest IV system (Data Sciences, Inc., Frankfurt, Germany). Each animal's body temperature was recorded every 10 min. Testing took place once the animal had achieved its presurgery body weight, and its body temperature showed a clear circadian rhythm ( 1 - 2 weeks following surgery). The animals were stressed by submitting them to 40-min restraint in a transparent Plexiglas tube ( 19 cm in length and 7 cm in diameter) with perforated sides to allow dissipation of body heat. Each animal received a 4-#1 ICV injection (3 #l/min), immediately before and immediately after the restraint stress. The injection procedure, consisting of removing the stylet, inserting the injection cannula, and injecting the solution, lasted 3 min per animal. The injection cannula was kept in place for 30 s before and after the infusion period. From 10 rain before the first injection until 40 min after removal from the tube, body temperature was recorded every minute. Stylets and injection cannulae were kept clean with alcohol. The injection cannula was 1 mm longer than the guide cannula, ensuring that the solution was injected into the lateral ventricle. Injections contained either saline alone, 3 ng AVP (Sigma, St. Quentin Fallavier, France), or 12 ng of the V1 type receptor AVP antagonist dPTyr [Me ] AVP (Sigma), both dissolved in saline. There were five treatments in total: 1 ) saline + saline (n = 10), 2) AVP antagonist + saline (n = 9), 3) AVP + saline (n = 9), 4) saline + AVP antagonist (n = 7), and 5) saline + AVP (n = 7), for the first and second injection, respectively.

Each animal was handled on each of the 3 days prior to testing: on the first 2 days this consisted of handling the animal and removing the stylet, and on the third day the animal was subjected to the entire injection procedure (insertion of the injection cannula, connection to the injection tube, and syringe during 120 s, but no injection, before being returned to their home cage). It was decided not to inject the animals with saline on the day before testing to avoid infections. Poulin and Pittman (36) have shown that infections leading to fevers may sensitize rats to the (behavioural) effects of vasopressin 24 h later. Body temperature was recorded every minute from 10 min before until 40 min after the simulated injection procedure. Handling and experimental testing took place between 1100 and 1300 h, when body temperature was near the nadir of the circadian rhythm. To test the effect of AVP on body temperature under conditions of minimal stress, 2 weeks after the restraint stress, eight rats, arbitrarily selected from the experimental group, received an injection of 3 ng of AVP and were replaced in their home cage. Body temperature was recorded as described above. Correct placement of the guide cannula was checked after experimentation, by verifying that there was easy flow of saline into the ventricle by gravity. If any doubt, the placement was histologically checked.

Statistical Analysis The average body temperatures were calculated over 5-rain periods. As animals differ in body temperature in absolute terms, the change in body temperature was calculated by substracting the preinjection body temperature from the hyperthermic response (based on the 5-min average preceding the injection). Analyses of variance (ANOVA) for repeated measures were carried out on 5-min averages of the change in body temperature (10-min averages for the period between 40 and 80 rain after application of the stress). Effects of prerestraint injections and postrestraint injections were compared separately (i.e., comparisons between treatments 1, 2, and 3; and 1, 4, and 5). ANOVA was also carried out using each individual's maximal increase in body temperature, and the time until this maximal increase was reached. Individual differences in the thermic response were controlled for by fitting temperature responses recorded following the simulated injection procedure as a covariate. Where interaction were found in the ANOVA, the least significant difference test was used to locate where differences were largest. It was found that six rats showed a sharp decline of 0.34°C or more in body temperature. The ANOVA was therefore repeated, excluding rats showing such a decline. The WilcoxonMann-Whitney (39) test was used to test differences in occurrence of a reduction in body temperature between groups. RESULTS Following the prestress AVP injection, one rat showed shortlasting ( < 2 min) behavioural inhibition and one showed barrel rotation. The latter rat was replaced. Six animals did not show a rise in body temperature of 0.15°C or more 20 min after handling on the day prior to testing, and one cannula was misplaced. These animals were excluded from the analysis (n = 1, 1, 2, 1, and 2 animals for treatments 1 to 5, respectively). Of the remaining animals, all cannulae were correctly placed. Fitting each individual's thermic response to the simulated injection as a covariate in the ANOVA showed a significant covariate effect on the within-subjects level (e.g., controls, during restraint), F( 1, 55) = 4.71, p < 0.05. Compared to the simulated injection, the injection followed by restraint stress was accompanied by a slightly stronger increase in body temperature, but

AVP A N D S T R E S S - I N D U C E D H Y P E R T H E R M I A

419

after removal from the restraint tube [average changes in body temperature: 0.46 ± 0.18, 0.53 ___0.15, 0.12 _ 0.08°C, for saline-, AVP antagonist-, and AVP-injected animals, respectively, F ( 1 , 21) = 2.28, p = 0.13]. There were no differences between controls and animals that had received the AVP antagonist either during or after the stress (least significant difference test, t = 1.06, NS). More detailed inspection of the results revealed that a number of rats showed a rapid decline in body temperature after the prerestraint injection. The average body temperature reached its lowest levels 10 min after introduction into the tube (see Fig. 1 ). The W i l c o x o n - M a n n - W h i t n e y test revealed that at 10 min, the ranked body temperature of AVP-injected rats (ranking starting with the highest values) was significantly lower than controls (p < 0.01 ), whereas AVP antagonist-injected rats did not differ from controls (see Fig. 2). Based on the data distribution, a hypothermic response was arbitrarily defined as a reduction of 0.34°C or more in body temperature 10 rain after the injection. One rat from the control group, one rat from the AVP antagonist group, and four rats from the AVP group showed a thus defined hypothermic response. Additionally, none of the rats from treatments 4 and 5 showed a hypothermic response after the prestress saline injection. It was found that the body temperature of rats, showing a hypothermic response post-AVP, was already significantly reduced 5 min after introduction into the tube [ time effect: F ( 7 , 21 ) = 33.62, p < 0.0001] and remained significantly lower than the preinjection body temperature until 20 min after the injection (see Fig. 3). However, from 10 min after introduction into the tube, the average body temperature of hypothermic rats showed a rise of 1.46°C over 30 min, that is, 0.05 +_ 0.03°C/min, which was a significantly faster increase than the one shown by the other two groups [e.g., controls: 0.03 + 0.01°C/min, F ( 2 , 18) = 7.40, p = 0.005] (see Fig. 3). To control for the initial hypothermic effects apparently induced by the AVP injection, the A N O V A was repeated, but excluding all rats that showed a hypothermic response. The analysis, which was therefore conducted on eight controls, seven AVP antagonist-injected rats, and four AVP-injected rats, still found a significant effect of AVP on the thermic response to restraint stress (Fig. 4), F ( 2 , 15) = 4.48,

this difference was not significant (e.g., controls: average increase in body temperature at 3 0 - 4 0 min: 0.57 ± 0.16 vs. 0.94 ± 0.15°C), F ( 1 , 7) = 2.91, p = 0.13.

Effect of Prerestraint Injection of A VP or A VP Antagonist on Body Temperature During and After Restraint Overall, rats showed a significant increase in body temperature after introduction into the restraint tube (see Fig. 1) [time effect: F ( 7 , 153) = 12.14, p < 0.001]. However, animals that had received an ICV injection with AVP prior to the restraint stress showed a reduced response [ treatment effect: F ( 2 , 21 ) = 5.94, p < 0.01]. There was a treatment × time interaction, F ( 14, 153) = 2.91,p < 0.001; animals having received AVP had significantly or nearly significantly lower body temperature between 10 and 40 min than the other two treatment groups (p < 0.001 at 10 and 15 min; p < 0.05 between 20 and 30 min; p < 0.08 at 35 and 40 min). AVP-injected animals had also a lower maximal body temperature than the other two treatment groups [1.08 ± 0.14, 1.11 ± 0.18, 0.62 ± 0.10, for saline, AVP antagonist- and AVP-injected animals, respectively, F ( 2 , 22) = 3.33, p = 0.05 ], but there were no differences in the time to reach the maximal body temperature obtained during the stress, F ( 2 , 22) = 0.56, NS (Fig. 1). The maximal body temperature and the time to reach the maximal body temperature were also not correlated ( r = 0.28, N S ) . The maximal body temperature occurring during the restraint was positively correlated with all body temperature values obtained during the stress (e.g., for control animals: p < 0.05 at 5 and 10 re_in of restraint stress;p < 0.005 from 15 to 40 rain during the stress and at 5 rain after removal from the tube; p nonsignificant thereafter). Rats that received AVP prior to the restraint stress also had lower body temperatures during the 40 rain following removal from the tube, compared to the other two treatment groups (Fig. 1 ), F ( 2 , 21 ) = 4.37, p < 0.05. This effect was removed by fitting each individual's average change in body temperature during the stress in the analysis, F ( 2 , 21 ) = 1.06, NS, but not by fitting the maximal change in body temperature, F ( 2 , 19) = 4.79,p < 0.05. No significant differences were found between 40 and 80 min

1.50

Oo

T.T T'TT T.T

1.00

m

T

.9. = •

o..

0.50

T,,'/"t ~- •T']''''"

E o

..o

0.00

,,,! Y

.___ OJ,

-~

"'¢"~.X.-;- ~-_~

-0.50

-1.00

= -20

-10

0

10

20

30

40

I

I

J

=

50

60

70

80

time (rain)

FIG. l. Change in body temperature during restraint stress (0-40 min) in rats after receiving ICV saline (I, n = 9), 12 ng of AVP antagonist (A, n = 8), or 2.5 ng of AVP ( 0 , n = 8) at time= -2, and saline at time = 40. Vertical bars indicate SEM.

420

T E R L O U W ET AL.

1.00

o~"

o.

0.50

0.00

E

o

J= ._c @

-0.50

5

-1.oo

lillllli i,,,, "°'°'lal Z

-1.50 59 63 39 41

61

42

33 28 50 49 38 64 35 30 71 31 74 25

37 26 23 72 43 55

rat number

FIG. 2. Change in body temperature of individual rats at 10 min after introduction into the tube, having received ICV saline (filled bars), 12 ng of AVP antagonist (hatched bars), or 2.5 ng of AVP (striped bars) at time= -2. p < 0.05, as well as thereafter [e.g., during the 40 min following removal of the tube; treatment effect: F ( 2 , 15) = 3.76, p < 0.05 ]. The rise in body temperature was delayed, and not steeper (0.02 _+ 0.01°C/min) than the one shown by the other two groups [e.g., controls: 0.02 ___ 0.01°C/rain, F ( 2 , 15) = 0.51, NS]. The reduced increased in body temperature was not removed by fitting average body temperature values during stress as a covariate in the analysis, F ( 2 , 14) = 4.34, p < 0.05, but it was by fitting the maximal body temperature values [e.g., during the 40 rain following removal of the tube; treatment effect: F ( 2 , 14) = 1.78, NS]. However, maximal body temperature no longer differed significantly among the treatments (1.18 _+ 0,12, 1.16 + 0.21, 0.67 + 0.06), F ( 2 , 16) = 2.40, NS.

Effects of Postrestraint Injection o f A VP or A VP Antagonist on Defervescence After Restraint Rats receiving saline before and one of the tested compounds after the restraint did not differ from controls in their hyperthermic response during the restraint stress, F ( 1, 15) = 0.16, NS. Removal from the tube was followed by a decline in body temperature; rats having received either of the compounds immediately after removal from the tube did not differ from saline-injected controls (treatment 1), F ( 1 , 14) = 2.27, NS. No hypothermic reactions similar to those described above (i.e., during restraint stress) were found after injections following removal from the tube.

Effect o f A VP on Body Temperature Under Conditions of Minimal Stress After the injection and replacement into the home cage, body temperature rose ]time effect: F(15, 105) = 11.89, p < 0.0001] and remained significantly higher than preinjection levels for about 80 rnin ]at 80 min: F ( 1 , 7) = 4.51, p = 0.07]. None of the rats showed a hypothermic reponse. Comparisons were subsequently made between simulated (saline) or AVP injection only (involving handling) and saline or AVP injection followed by 40-min restraint (involving handling and restraint). Rats that have been excluded in other analyses (hypothermic effect) were again excluded, thus reducing the chance to find significant dif-

ferences. All treatment groups were subjected to simulated injection and several treatment groups were subjected to saline injection followed by restraint, but the different possible analyses (depending on which treatment group was included) gave similar results. There was an overal effect (Fig. 5) [e.g., 0 - 4 0 rain: F ( 3 , 18) = 3.48, p < 0.05]; AVP-injected rats, whether subjected to restraint or not, showed similar average body temperatures (0.31 __+0.14°C and 0.36 _ 0.16, respectively; t = 0.44, N S ) , and these were lower than the average body temperature of saline-injected rats subjected to restraint (0.73 _+ 0.14°C; t = 3.00, p < 0.01 and t = 2.57, p < 0.02, respectively). Rats subjected to a simulated injection, without restraint, were intermediate, showing no significant differences with any of the other groups (e.g., compared with AVP, no restraint: t = 1.27, NS). DISCUSSION

The results show that restraint stress induced a significant rise in body temperature and that ICV AVP administration prior to the stress reduced this response. Analysis of the control group showed a positive correlation between the thennic response to the simulated injection on the day before the restraint test and to the restraint stress itself. This indicates that the animals were consistent in their thermic response to stress and that this response was not simply due to a reduced heat dissipation in the tubes. These observations suggest that the rats of the present experiment showed a stress-induced hyperthermia as described in previous reports (3,19,20,40). It was further found that the individual's initial thermic response to the stress had a predictive value for the maximal temperature increase, as illustrated by the positive correlation between these two indices. Thus, in accordance with its antipyretic effects, ICV administration of AVP reduces stress-induced hyperthermia. The AVPinduced reduction in the hyperthermic response to stress appeared to be at least partially related to a reduction in body temperature, apparent within 5 min after the administration of AVP, in a number of animals. However, the effect was significant, even after exclusion of these animals. The mechanisms through which AVP affects body temperature remain largely unclear, and contrasting results have been reported. The ICV administration of AVP increases sympathetic nerve activity, lead-

AVP AND S T R E S S - I N D U C E D H Y P E R T H E R M I A

421

1.00

O"

0.50

¢..,

0.00

o v

0.

,,~"

,,T..T"I "

t

E

'I o

..,o .E

4

-0.50

a°

g E m

"3

-1.00 -1.50

1 i

-20

-10

0

10

20

30

40

I

t

i

p

50

60

70

80

time (rnin)

FIG. 3. Change in body temperature of rats showing a hypothermic response after receiving ICV 2.5 ng of AVP (n = 4) at time= -2. See text for details. Vertical bars indicate SEM.

ing to increases in heart rate and mean arterial blood pressure (15,22). It also increases sympathetic nerve activity in brown adipose tissue in Zucker rats, which is accompanied by an increase in body temperature (14). It is therefore unlikely that the effects found in the present study are due to indirect peripheral effects of AVP. Much of the research has been carried out on the antipyretic effects of AVP on febrile body temperature (see Introduction). The mechanism responsible for the AVP-induced reduction in the febrile response was found to depend on the ambient temperature ( 4 - 3 2 ° C ) (42,43). The authors suggested, therefore, that the effects of AVP are not due to changes in a single effector mechanism, but probably to its effects on the febrile body temperature set point. AVP has further been shown to antagonize prostaglandin-induced fevers, suggesting that it is not acting to inhibit prostaglandin synthesis (38). Central microin-

jections of a V~ receptor antagonist inhibit the antipyretic effects of peripherally administered indomethacin (44). Based on these results, it has been suggested that prostaglandins and AVP may be acting on the same postsynaptic neurons, or even on the same receptor (17). However, a lack of interaction between V~ type receptors and indomethacin has also been reported (13). An interesting observation in the present study is that the average body temperature rose about 0.50°C above the basal levels in both the hypothermic and nonhypothermic rats receiving AVP, despite the initial drop in the hypothermic rats. This might suggest that, similar to Wilkinson and Kasting's study (42,43), AVP acted on the body temperature set point. Further investigation is needed to study the mode of action of AVP on stress-induced hyperthermia. The present study found a rapid decline in body temperature after AVP administration in half of the animals. A similar action

2.00

O"

1.50

m

1.00

"L2.!.T T !

E

o

0.50 *" .zP T.."'"

E

T 2. T ! T_-!-

"*T

o.oo

°0.50

I -20

-10

0

10

20

30

40

i

I

i

i

50

60

70

80

time (min)

FIG. 4. Same as in Fig. l, but after exclusion of rats showing a hypothermic response. See text for details.

422

TERLOUW ET AL.

1.50

TTTT T ~"

1.00

Q.

0.50

O v

E Q~ "O O

J=

0.00

.__.

-0.50

-1.00

i

-20

-10

0

10

20

30

40

i

i

i

i

50

60

70

80

time (rain)

FIG. 5. Change in body temperature after (time = -2) simulated ICV administration of saline (©), after ICV administration of 2.5 ng of AVP (O), after ICV administration of saline followed by restraint (O) or of 2.5 ng of AVP followed by restraint ( • ). Simulated ICV administration: data of group of treatment 1; ICV administration of saline followed by restraint: data of group of treatment 4; other groups have not been included for sake of clarity. See Fig. 1 for additional details.

of AVP may have taken place in all rats: an overt hypothermic response may have been overridden in some of the rats by the restraint stress. In addition, it may be that AVP-induced hypothermia is specific to more serious stress conditions, as no hypothermic responses were found after admininstration of AVP under conditions of minimal stress. It is often reported that AVP has mostly antipyretic effects, that is, at lower doses it only affects the body temperature of the febrile animal, and has no influence on the body temperature of the nonfebrile animal [e.g., (6,33,35,42,43)]. Other authors, however, do report AVP-induced hypothermic responses in nonfebrile rats. Diamant and De Wied ( 11 ) found a reduction of 0.50°C after administering 30 ng of AVP into the lateral ventricle. Wilkinson and Kasting (43) found a reduction in brain temperature in nonfebrile animals kept at an ambient temperature of 4°C, after ICV administration of 2.5 ng of AVP. Although they report the overall effect as nonsignificant, there was a reduction in body temperature of about 0.6 +_ 0.3°C at 10 and 20 min after the injection. Interestingly, the effect was not found in animals kept at an ambient temperature of 25 or 32°C. They suggested that at an ambient temperature of 4°C rats may become more sensitive to the effects of AVP, although the mechanisms remain unclear. Banet and Wieland (1) found a similar effect of ambient temperature on the occurrence of a AVP-induced hypothermic response, but their results are difficult to interpret as they injected AVP into the lateral septum, a site not known to be involved in vasopressinergic body temperature regulation [e.g. (32)]. In other studies it was found that ICV injection of a large dose (t #g) of AVP caused an immediate sharp decrease in body temperature (16,28). Antipyretic compounds may affect nonfebrile body temperature at high doses (2,43), and one could argue that some of the above studies found hypothermic effects because high doses were used. Comparison of doses between studies remains difficult, given the large variation in sensitivity between rats from different colonies. Some authors reported behavioural changes, such as

convulsions, accompanying the changes in body temperature (16,28), which might indicate that a relatively high dose was used. However, behavioural effects of a given dose may not necessarily give a good appreciation of the level of the dose as far as body temperature regulation is concerned, as there is no evidence that these two indices are correlated. On the contrary, Kasting et al. (16) found that repeated ICV AVP administration, at doses that induced changes in both behaviour and body temperature, induced sensitization to the drug on the behavioural level, but not on the level of body temperature. The hypothermic response of the rats in the present study may simply indicate a relatively high sensitivity to the effects of AVP on body temperature regulation of these rats (at least under the conditions of restraint stress), without necessarily being accompanied by behavioural changes. Such high sensitivity may be inherent to these rats, or may have been inadvertedly induced by prior infections. Repeated injections of endotoxin result in the the gradual reduction of the febrile response to endotoxin. AVP appears to be responsible for this tolerance, as administration of the AVP V~ receptor antagonist restored the suppressed fevers to levels seen prior to the induction of tolerance (45). The lack of enhanced hypothermia in response to repeated AVP administration found by Kasting et al. (16) may suggest that the sensitization process itself is not based on vasopressinergic mechanisms. Concerning the present study, however, none of the rats showing a hypothermia had given signs of a fever due to the simulated injection procedure. Close inspection of the circadian rhythm of body temperature of the 7 days preceding the test also revealed no relationship between occasional irregularities in body temperature rhythm (i.e., potential fevers) and the hypothermic response. In contrast to AVP, the V1 type receptor antagonist did not affect the hyperthermic response to stress. It is possible that this is due to an insufficient dose. However, a fivefold dose of AVP antagonist is usually sufficient to reverse the effects of AVP, and similar doses of AVP and AVP antagonist as used in the present experiment were previously found to reverse and potentiate, re-

AVP AND STRESS-INDUCED HYPERTHERMIA

423

spectively, the behavioural effects of ICV interleukin-1 injections (5 n g ) in male rats in this laboratory ( 7 ) . Furthermore, in a separate test, we determined that prior administration of 18 ng of the A V P antagonist prevented the behavioural alterations induced by 6 ng of A V P (data not s h o w n ) . It is also unlikely that the lack of effect of the A V P antagonist is related to either a short-lasting or delayed action, as previous work found that administration of the compound into the V S A 5 - 1 5 min prior to the ICV administration of IL-1 prolonged and augmented the fever and behavioural responses to the latter (6,7). One explanation may be that the stress-induced increase in body temperature was maximal, and therefore difficult to increase further. In support of this, body temperature often reaches plateau values after 30 min, and does not further increase by a longer duration of stress exposure (3,31). In addition, peak values reported are usually in the range of 0.75 to 1.25°C (i.e., similar found to those in the present study). This would also explain the relatively small differences in thermic response between procedures involving handling and injection, and handling and injection followed by restraint. Furthermore, it has been found that in stressed animals, an injection with IL-1 induces only a small (and apparently nonsignificant) further rise in body temperature (25,30). Another possibility is that whereas exogenous A V P can affect the hyperthermia, endogenous A V P plays no role. In support of this, Landgraf et al. ( 2 1 ) showed that passive (environmentally induced hyperthermia) did not cause A V P to be released into p u s h - p u l l perfusates. Rats that received A V P prior to the stress tended to have a lower body temperature during the 40 min following the stress compared to controls. However, this difference appeared to be related to the reduced hyperthermic reponse during the stress, as fitting the average body temperature or the maximal body temperature value as a covariate in the analysis removed the effect. Previous work found that bolus injections of A V P are metabolised within minutes ( 4 ) , which is in accordance with the lack of

effect. Postrestraint ICV administration of A V P and pre- or postrestraint ICV administration of the A V P Vl receptor antagonist did not affect defervescence after stress-induced hyperthermia. In contrast, previous work found that degeneration of the steroid-dependent vasopressinergic innervation of the brain by castration (9,10) resulted in reduced defervescence after a PGEinduced fever ( 3 4 ) . These authors did not, however, report to what extent this effect could have been due to an increased fever response in the castrated animals. To our knowledge, this is the first study to assess the interaction between central A V P and stress-induced hyperthermia. The results show that the adminstration of A V P into one of the lateral ventricles significantly reduces the hyperthermic response to restraint stress and handling stress. The reduced increase in body temperature during restraint stress can be explained, at least partially, by a hypothermic effect on the nonfebrile body temperature in half of the animals. The mechanisms underlying the hypothermic response are presently unknown. Administration of the V~ type A V P antagonist did not affect the thermic response to stress, which may suggest that the hyperthermic response had reached maximal (ceiling) levels. Defervescence was not affected by AVP, nor by a specific antagonist V~ type receptors administered after removal from the tube. An apparent increase in defervescence after the stress in rats having received A V P prior to the stress was due to the lower body temperatures of these rats during the stress. Further investigations using more direct techniques should be used to study the interaction between A V P and the thermic response to stress. ACKNOWLEDGEMENTS The authors wish to thank Dr. Q. J. Pittman, for help in the interpretation of the data and V. Tridon for technical assistance. They also thank an anonymous referee for helpful suggestions. S. K. was supported by an NSF-NATO Postdoctoral fellowship.

REFERENCES 1. Banet, M.; Wieland, U. E. The effect of intraseptally applied vasopressin on thermoregulation in the rat. Brain Res. Bull. 14:113-116; 1985. 2. Barbour, H. G. The heat regulating mechanism of the body. Physiol. Rev. 1:295-326; 1921. 3. Briese, E.; Cabanac, M. Stress hyperthermia: Physiological arguments that it is a fever. Physiol. Behav. 49:1153-1157; 1991. 4. Burbach, J. P. H.; LeBouille, J. L. M. Proteolytic conversion of arginin-vasopressin and oxytocin by brain synaptic membranes. J. Biol. Chem. 258:1487-1494; 1983. 5. Cooper, K. E.; Kasting, N. W.; Lederis, K.; Veale, W. L. Evidence supporting a role for endogenous vasopressin in natural suppression in the sheep. J. Physiol. 295:33-45; 1979. 6. Cooper, K. E.; Naylor, A. M.; Veale, W. L. Evidence supporting a role for endogenous vasopressin in fever suppression in the rat. J. Physiol. 387:163-172; 1987. 7. Dantzer, R.; Bluthr, R. M.; Kelley, K. W. Androgen-dependent vasopressinergic neurotransmission attenuates interleukin-l-induced sickness behavior. Brain Res. 557:115-120; 1991. 8. De Vries, G. J.; Buijs, R. M. The origin of the vasopressinergic and oxytocinergic innervation of the rat brain with special reference to the lateral septum. Brain Res. 273:307-317; 1983. 9. De Vries, G. J.; Buijs, R. M.; Sluijter, A. A. Gonadal hormone actions on the morphology of the vasopressinergic innervation of the adult brain. Brain Res. 298:141 - 145; 1984. 10. De Vries, G. J.; Buijs, R. M.; Van Leeuwen, F. W.; Caffe, W. R.; Swaab, D. F. The vasopressinergic innervation of the brain in normal and castrated rats. J. Comp. Neurol. 233:236-254; 1985.

11. Diamant, M.; De Wied, D. Differential effects of centrally injected AVP on heart rate, core temperature, and behavior in rats. Am. J. Physiol. Regul. Integr. Comp. Physiol. 264:R51-R61; 1993. 12. Disturnal, J. E.; Veale, W. L.; Pittman, Q. J. Modulation by arginine vasopressin of glutamate excitation in the ventral septal area of the rat brain. Can. J. Physiol. Pharmacol. 65:30-35; 1987. 13. Fyda, D. M.; Cooper, K. W.; Veale, W. L. Indomethacin-induced antipyresis in the rat: Role of vasopressin receptors. Brain Res. 494:307-314; 1989. 14. Holt, S. J.; York, D. A. The effects of adrenalectomy, corticotropin releasing factor and vasopressin on the sympathetic firing rate of nerves to intercapsular brown adipose tissue in the Zucker rat. Physiol. Behav. 45:1123-1129; 1989. 15. Janiak, P. B.; Kasson, G.; Brody, M. J. Central vasopressin raises arterial pressure by sympathetic activation and vasopressin release. Hypertension 13(6, Pt. 2):935-940; 1989. 16. Kasting, N. W.; Veale, W. L.; Cooper, K. E. Convulsive and hypothermic effects of vasopressin the the brain of the rat. Can. J. Physiol. Pharmacol. 58:316-319; 1980. 17. Kasting, N. W. Criteria for establishing a physiological role for brain peptides. A case in point: The role of vasopressin in thermoregulation during fever and antipyresis. Brain Res. Rev. 14:143-153; 1989. 18. Kasting, N. W.; Wilkinson, M. F. Microdialysis evaluation of brain antipyretic concentrations of AVP following ICV injection. Soc. Neurosci. Abstr. 14:149; 1988. 19. Kent, S.; Kelley, K. W.; Dantzer, R. Stress-induced hyperthermia is partially mediated by interleukin-I (IL-1). Soc. Neurosci. Abstr. 19:226; 1993.

424

20. Kluger, M. J.; O'Reilly, B.; Shope, T. R.; Vander, A. J. Further evidence that stress hyperthermia is a fever. Physiol. Behav. 39:763766; 1987. 21. Landgraf, R.; Malkinson, T. J.; Veale, W. L.; Lederis, K.; Pittman, Q. J. Vasopressin and oxytocin in rat brain in response to prostaglandin fever. Am. J. Physiol. Regul. Integr. Comp. Physiol. 259)28):R1056-R1062; 1990. 22. Lebrun, C. J.; Rohmeiss, P.; Demmert, G.; Unger, T. Sensitization to pressor effects by repeated central injections of vasopressin in conscious rats. Brain Res. 505:39-43; 1989. 23. LeMay, L. G.; Vander, A. J.; Kluger, M. J. The effects of psychological stress on plasma interleukin-6 activity in rats. Physiol. Behav. 47:957-961; 1990. 24. Long, N. C.; Vander, A. J.; Kluger, M. J. Stress-induced rise of body temperature in rats is the same in warm and cool environments. Physiol. Behav. 47:773-775; 1990. 25. Long, N. C.; Morimoto, A.; Nakamori, T.; Murakami, N. The effect of physical restraint on IL-lb- and LPS-induced fever. Physiol. Behav. 50:625-628; 1991. 26. Marazziti, D.; Di Muro, A.; Castrogiovanni, P. Psychological stress and body temperature changes in humans. Physiol. Behav. 52:393395; 1992. 27. Martin, S. M.; Malkinson, T. J.; Veale, W. L.; Pittman, Q. J. Prostaglandin fever in rats is altered by kainic acid lesions of the ventral septal area. Brain Res. 455:196-200; 1988. 28. Meisenberg, G.; Simmons, W. H. Hypothermia induced by centrally administered vasopressin in rats. Neuropharmacology 23:11951200; 1984. 29. Minami, M.; Kuraishi, Y.; Yamaguchi, T.; Nakai, S.; Hirai, Y.; Satoh, M. Immobilization stress induces intedeukin-lb mRNA in the rat hypothalamus. Neurosci. Lett. 123:254-256; 1991. 30. Morimoto, A.; Watanabe, T.; Morimoto, K.; Nakamori, T.; Murakami, N. Possible involvement of prostaglandins in psychological stress-induced responses in rats. J. Physiol. 443:421-429; 1991. 31. Nakamori, T.; Morimoto, A.; Morimoto, K.; Tan, N.; Murakami, N. Effects of a- and/3-adrenergic antagonists on rise in body temperature induced by psychological stress in rats. Am. J. Physiol. Regul. Integr. Comp. Physiol. 264(33):R156-R161; 1993. 32. Naylor, A. M.; Ruwe, W. D.; Veale, W. L. Antipyretic action of centrally administered arginine vasopressin but not oxytocin in the cat. Brain Res. 385:156-160; 1986.

T E R L O U W ET AL.

33. Naylor, A. M.; Pittman, Q. J.; Veale, W. L. Stimulation of vasopressin release in the ventral septum of the rat brain suppresses prostaglandin E1 fever. J. Physiol. 399:177-189; 1988. 34. Pittman, Q. J.; Malkinson, T. J.; Kasting, N. W.; Veale, W. L. Enhanced fever following castration: possible involvement of brain arginine vasopressin. Am. J. Physiol. Regul. Integr. Comp. Physiol. 254(23):R513-R517; 1988. 35. Pittman, Q. J.; Landgraf, R. Vasopressin in thermoregulation and blood pressure control. Vasopressin 208:177-184; 1991. 36. Poulin, P.; Pittman, Q. J. Possible involvement of brain oxytocin in modulating vasopressin antipyretic action. Am. J. Physiol. Regul. Integr. Comp. Physiol. 264(34):R151-R156; 1993. 37. Renbourn, E. T. Body temperature and pulse rate in boys and young men prior to sporting contests. A study in emotional hyperthermia, with a review of the literature. J. Psychosom. Res. 4:149-175. 1960. 38. Ruwe, W. D.; Naylor, A. M.; Veale, W. L. Perfusion of vasopressin within the rat brain suppresses prostaglandin E-hyperthermia. Brain Res. 338:219-224; 1985. 39. Siegel, S.; Castellan, N. J. McGraw-Hill international editions-statistics series. New York: McGraw-Hill; 1988. 40. Singer, R.; Harker, C. T.; Vander, A. J.; Kluger, M. J. Hyperthermia induced by open-field stress is blocked by salicylate. Physiol. Behav. 36:1179-1182; 1986. 41. Terlouw, C.; Kent, S.; Dantzer, R.; Monin, G. L'hyperthermie induite par le stress: Donn6es pr61iminaires. In: Pichard, M.; Porter, R. H.; Signoret, J. P., eds. Comportement et bien-~tre animal. Pads: INRA; 1994:191-201. 42. Wilkinson, M. F.; Kasting, N. W. Antipyresis due to centrally administered vasopressin differentially alters thermoregulatory effects depending on the ambient temperature. Regul. Pept. 19:45-54; 1987. 43. Wilkinson, M. F.; Kasting, N. W. The antipyretic effects of centrally administered vasopressin at different ambient temperatures. Brain Res. 415:275-280; 1987. 44. Wilkinson, M. F.; Kasting, N. W. Central vasopressin V~-receptors mediate indomethacin-induced antipyresis in rats. Am. J. Physiol. Regul. Integr. Comp. Physiol. 256(25):R1164-R1168; 1989. 45. Wilkinson, M. F.; Kasting, N. W. Centrally acting vasopressin contributes to endotoxin tolerance. Am. J. Physiol. Regul. Integr. Comp. Physiol. 258(27):R443-R449; 1990.