Vasopressin into the preoptic area increases grooming behavior in mice

Lateral Habenula Neurons Signal Cold Aversion and Participate in Cold Aversion

The aversion to cold is a fundamental motivated behavior that contributes to the body temperature homeostasis. However, the involvement of the lateral habenula (LHb) as a regulatory hub for negative emotions in this physiological process remains uninvestigated. In this study, we demonstrate an elevation in the population activity of LHb neurons following exposure to cold stimuli. Additionally, we establish the necessity of Vglut2-expressing neurons within the LHb for the encoding of cold aversion behaviors. Furthermore, we have elucidated a neural circuit from excitatory neurons of the dorsomedial hypothalamus (DMH) to LHb that plays a crucial role in this progress. Manipulation of the DMH-LHb circuit has a significant impact on cold aversion behavior in mice. It is worth noting that this circuit does not exhibit any noticeable effects on autonomic thermoregulation or depression-like behavior. The identification of these neural mechanisms involved in behavioral thermoregulation provides a promising avenue for future research.

Paraventricular hypothalamic vasopressin neurons induce self-grooming in mice

Self-grooming plays an essential role in hygiene maintenance, thermoregulation, and stress response. However, the neural populations involved in self-grooming remain largely unknown. The paraventricular hypothalamic nucleus (PVH) has been implicated in the regulation of self-grooming. Arginine vasopressin-producing neurons are among the major neuronal populations in the PVH (PVH^AVP), which play important roles in water homeostasis, blood pressure regulation, feeding, and stress response. Here, we report the critical role of PVH^AVP neurons in the induction of self-grooming. Optogenetic activation of PVH^AVP neurons immediately induced self-grooming in freely moving mice. Chemogenetic activation of these neurons also increased time spent self-grooming. In contrast, their chemogenetic inhibition significantly reduced naturally occurring self-grooming, suggesting that PVH^AVP-induced grooming has physiological relevance. Notably, optogenetic activation of PVH^AVP neurons triggered self-grooming over other adaptive behaviors, such as voracious feeding induced by fasting and social interaction with female mice. Thus, our study proposes the novel role of PVH^AVP neurons in regulating self-grooming behavior and, consequently, hygiene maintenance and stress response. Furthermore, uncontrolled activation of these neurons may be potentially relevant to diseases characterized by compulsive behaviors and impaired social interaction, such as autism, obsessive–compulsive disorder, and anorexia nervosa.

Social status in mouse social hierarchies is associated with variation in oxytocin and vasopressin 1a receptor densities

The neuropeptides oxytocin and vasopressin and their receptors have established roles in the regulation of mammalian social behavior including parental care, sex, affiliation and pair-bonding, but less is known regarding their relationship to social dominance and subordination within social hierarchies. We have previously demonstrated that male mice can form stable linear dominance hierarchies with individuals occupying one of three classes of social status: alpha, subdominant, subordinate. Alpha males exhibit high levels of aggression and rarely receive aggression. Subdominant males exhibit aggression towards subordinate males but also receive aggression from more dominant individuals. Subordinate males rarely exhibit aggression and receive aggression from more dominant males. Here, we examined whether variation in social status was associated with levels of oxytocin (OTR) and vasopressin 1a (V1aR) receptor binding in socially relevant brain regions. We found that socially dominant males had significantly higher OTR binding in the nucleus accumbens core than subordinate animals. Alpha males also had higher OTR binding in the anterior olfactory nucleus, posterior part of the cortical amygdala and rostral lateral septum compared to more subordinate individuals. Conversely, alpha males had lower V1aR binding in the rostral lateral septum and lateral preoptic area compared to subordinates. These observed relationships have two potential explanations. Preexisting individual differences in the patterns of OTR and V1aR binding may underlie behavioral differences that promote or inhibit the acquisition of social status. More likely, the differential social environments experienced by dominant and subordinate animals may shift receptor expression, potentially facilitating the expression of adaptive social behaviors.

Persistent increase in hypothalamic arginine vasopressin gene expression during protracted withdrawal from chronic escalating-dose cocaine in rodents

Arginine vasopressin (AVP) from the paraventricular nucleus (PVN) of hypothalamus has important roles in regulation of the hypothalamic-pituitary-adrenal (HPA) axis and stress-related behaviors during chronic stress. It is unknown, however, whether AVP in the PVN is involved in the modulation of HPA activity after chronic cocaine exposure. Here, we examined the gene expression alterations of AVP in the hypothalamus, and V1b receptor and pro-opiomelanocortin (POMC) in the anterior pituitary, as well as HPA hormonal changes, in Fischer rats after chronic cocaine and withdrawal, using two different chronic (14-day) 'binge' pattern administration regimens: steady-dose cocaine (SDC, 45 mg/kg/day) and escalating-dose cocaine (EDC, 45 up to 90 mg/kg/day). There was a significant (7-fold) plasma adrenocorticotropic hormone (ACTH) elevation after chronic EDC (but not SDC), coupled with increased V1b and POMC mRNA levels in the anterior pituitary. From acute (1-day) to protracted (14-day) withdrawal from chronic EDC (but not from SDC), we found persistent elevations of both plasma ACTH and corticosterone levels and AVP mRNA levels in the PVN. Selective V1b antagonist SSR149415 (5 mg/kg) attenuated acute withdrawal-induced HPA activation after EDC. To study potential roles of endogenous opioids in modulating the AVP gene, we administered naloxone (1 mg/kg); we found that opioid receptor antagonism increased AVP mRNA levels in cocaine-naive rats, but not in cocaine-withdrawn rats, suggesting less tonic opioid inhibition of PVN AVP neurons after chronic EDC. To assess the effects of cocaine withdrawal on sub-populations of PVN AVP neurons, we utilized AVP-enhanced green fluorescent protein (EGFP) promoter transgenic mice and found that acute withdrawal following chronic EDC increased the number of AVP-EGFP neurons in the parvocellular PVN (pPVN). These results suggest that during protracted withdrawal, enhanced pPVN AVP gene expression is associated with persistent elevations of basal HPA activity; a hyposensitivity of PVN AVP gene expression to naloxone is indicative of reduced opioidergic tone. Our studies indicate that the AVP and its V1b receptor system may be a potential therapeutic target for treating anxiety and depressive symptoms associated with cocaine addiction.

Deficiency of antinociception and excessive grooming induced by acute immobilization stress in Per1 mutant mice

Acute stressors induce changes in numerous behavioral parameters through activation of the hypothalamic-pituitary-adrenal (HPA) axis. Several important hormones in paraventricular nucleus of the hypothalamus (PVN) play the roles in these stress-induced reactions. Corticotropin-releasing hormone (CRH), arginine-vasopressin (AVP) and corticosterone are considered as molecular markers for stress-induced grooming behavior. Oxytocin in PVN is an essential modulator for stress-induced antinociception. The clock gene, Per1, has been identified as an effecter response to the acute stresses, but its function in neuroendocrine stress systems remains unclear. In the present study we observed the alterations in grooming and nociceptive behaviors induced by acute immobilization stress in Per1 mutant mice and other genotypes (wild types and Per2 mutant). The results displayed that stress elicited a more robust effect on grooming behavior in Per1 mutant mice than in other genotypes. Subsequently, the obvious stress-induced antinociception was observed in the wild-type and Per2 mutant mice, however, in Per1 mutant, this antinociceptive effects were partially-reversed (mechanical sensitivity), or over-reversed to hyperalgesia (thermal sensitivity). The real-time qPCR results showed that in PVN, there were stress-induced up-regulations of Crh, Avp and c-fos in all of genotypes; moreover, the expression change of Crh in Per1 mutant mice was much larger than in others. Another hormonal gene, Oxt, was up-regulated induced by stress in wild-type and Per2 mutant but not in Per1 mutant. In addition, the stress significantly elevated the serum corticosterone levels without genotype-dependent differences, and accordingly the glucocorticoid receptor gene, Nr3c1, expressed with a similar pattern in PVN of all strains. Taken together, the present study indicated that in acute stress treated Per1 mutant mice, there are abnormal hormonal responses in PVN, correlating with the aberrant performance of stress-induced behaviors. Therefore, our findings suggest a novel functional role of Per1 in neuroendocrine stress system, which further participates in analgesic regulation.

Increased cerebral activity suppresses baroreflex control of heart rate in freely moving mice

We assessed whether increased cerebral activity suppressed baroreflex control of heart rate (HR) and, if so, whether this occurred prior to the onset of locomotion in daily activity of mice. We measured mean arterial pressure (MAP, arterial catheter), cerebral blood flow in the motor cortex (CBF, laser-Doppler flowmetry), and electroencephalogram in free-moving mice (n = 8) during 12 daytime hours. The contribution of baroreflex control of HR to MAP regulation was determined during a total resting period for approximately 8 h from the cross-correlation function (R(t)) between spontaneous changes in HR (HR) and MAP (MAP) every 4 s and the sensitivity was determined from HR/MAP where R(t) was significant (P < 0.05). The power density ratio of theta to delta wave band in electroencephalogram (theta/delta), determined every 4 s as an index of cerebral activity, was positively correlated with CBF during 73 +/- 3% of the total resting period (P < 0.05) and with R(t) during 59 +/- 2% (P < 0.05). When each measurement during the resting period was divided into seven bins according to the level of theta/delta, CBF was 91 +/- 2% in the lowest bin and 118 +/- 3% in the highest bin (P < 0.001), R(t) was 0.69 +/- 0.06 and 0.27 +/- 0.04 (P < 0.001) and HR/MAP (beats min(1) mmHg(1)) was 12.4 +/- 0.9 and 7.5 +/- 0.9 (P < 0.001), respectively, with significant correlations with theta/delta (all P < 0.002). Moreover, mice started to move in approximately 30 sec after the sequential increases of theta/delta and R(t), mice started to move at 5 times higher probability than after a given time, followed by a rapid increase in MAP by approximately 10 mmHg. These results suggest that increased cerebral activity suppresses baroreflex control of HR and this might be related to the start of voluntary locomotion with a rapid increase in MAP.

Oxytocin, vasopressin, and human social behavior

There is substantial evidence from animal research indicating a key role of the neuropeptides oxytocin (OT) and arginine vasopressin (AVP) in the regulation of complex social cognition and behavior. As social interaction permeates the whole of human society, and the fundamental ability to form attachment is indispensable for social relationships, studies are beginning to dissect the roles of OT and AVP in human social behavior. New experimental paradigms and technologies in human research allow a more nuanced investigation of the molecular basis of social behavior. In addition, a better understanding of the neurobiology and neurogenetics of human social cognition and behavior has important implications for the current development of novel clinical approaches for mental disorders that are associated with social deficits (e.g., autism spectrum disorder, social anxiety disorder, and borderline personality disorder). This review focuses on our recent knowledge of the behavioral, endocrine, genetic, and neural effects of OT and AVP in humans and provides a synthesis of recent advances made in the effort to implicate the oxytocinergic system in the treatment of psychopathological states.

Characterization of the V1a antagonist, JNJ-17308616, in rodent models of anxiety-like behavior

RATIONALE

Vasopressin (AVP) plays a role in regulating anxiety, which is thought to be partially mediated through the V1a receptor. Recently, JNJ-17308616 was identified as a V1a antagonist.

OBJECTIVES

The purpose of this work was to assess V1a receptor affinity and selectivity of JNJ-17308616 and in vivo efficacy in animal models of anxiety-like behavior.

MATERIALS AND METHODS

The affinity of JNJ-17308616 for the human and rat V1a, V1b, V2, and oxytocin receptors was determined. Central administration of AVP induces a scratching response mediated through the V1a receptor. Inhibition of scratching was used as a behavioral measure of in vivo potency. JNJ-17308616 was tested in five models of anxiety: rat elevated plus-maze (EPM), rat-elevated zero-maze (EZM), rat-conditioned lick suppression (CLS), rat pup separation-induced ultrasonic vocalizations (USV), and mouse marble burying (MMB).

RESULTS

High affinity for the human V1a receptor (K (i) 5.0 nM) was confirmed. However, the rat V1a receptor affinity was more modest (K (i) 216 nM), and the compound was not selective over the rat V2 receptor (K (i) 276 nM). At 100 mg/kg, JNJ-17308616 significantly reduced anxiety-like behavior in EPM, USV, and MMB; at 30 mg/kg, it was effective in EZM and CLS. JNJ-17308616 neither impaired social recognition nor reduced locomotor activity.

CONCLUSIONS

These results demonstrate the potential for V1a receptor antagonists as novel anxiolytics. Tool compounds that have greater V1a receptor selectivity than JNJ-17308616 are necessary to make precise conclusions about the role of the V1a receptor in affective disorders.

Differential blockade of CRF-evoked behaviors by depletion of norepinephrine and serotonin in rats

RATIONALE

Central administration of corticotropin-releasing factor (CRF) elicits a specific pattern of behavioral responses resembling a stress-like state and is anxiogenic in rodent models of anxiety.

OBJECTIVES

Specific behaviors evoked by the administration of CRF were measured. The roles of CRF receptor subtypes and that of serotonergic and noradrenergic systems in mediating these responses were studied.

MATERIALS AND METHODS

Burying, grooming, and head shakes were quantified in rats following intracerebroventricular administration of CRF and urocortin II and after pretreatment with antagonists. The role of forebrain norepinephrine in the behavioral responses to CRF (0.3 microg) was examined following pretreatment with the neurotoxin DSP-4 and that of serotonin after depletion using systemic administration of para-chlorophenylalanine (p-CPA).

RESULTS

CRF at 0.3 and 3.0 microg caused robust increases in burying, grooming, and head shakes, but urocortin II was ineffective. Pretreatment with either antalarmin or propranolol significantly attenuated the CRF-evoked behaviors. Destruction of forebrain norepinephrine pathways blocked spontaneous burying behavior elicited by CRF and conditioned burying directed towards an electrified shock probe. In contrast, depletion of 5-HT selectively attenuated CRF-evoked grooming.

CONCLUSIONS

Overt behavioral responses produced by CRF, burying, grooming, and head shakes appeared to be mediated through the CRF(1) receptor. Spontaneous burying behavior evoked by CRF or conditioned burying directed towards a shock probe was disrupted by lesion of the dorsal noradrenergic bundle and may represent anxiety-like behavior caused by CRF activation of the locus ceruleus. In contrast, CRF-evoked increases in grooming were dependent on serotonin.

The hyperactive syndrome: metanalysis of genetic alterations, pharmacological treatments and brain lesions which increase locomotor activity

The large number of transgenic mice realized thus far with different purposes allows addressing new questions, such as which animals, over the entire set of transgenic animals, show a specific behavioural abnormality. In the present study, we have used a metanalytical approach to organize a database of genetic modifications, brain lesions and pharmacological interventions that increase locomotor activity in animal models. To further understand the resulting data set, we have organized a second database of the alterations (genetic, pharmacological or brain lesions) that reduce locomotor activity. Using this approach, we estimated that 1.56% of the genes in the genome yield to hyperactivity and 0.75% of genes produce hypoactivity when altered. These genes have been classified into genes for neurotransmitter systems, hormonal, metabolic systems, ion channels, structural proteins, transcription factors, second messengers and growth factors. Finally, two additional classes included animals with neurodegeneration and inner ear abnormalities. The analysis of the database revealed several unexpected findings. First, the genes that, when mutated, induce hyperactive behaviour do not pertain to a single neurotransmitter system. In fact, alterations in most neurotransmitter systems can give rise to a hyperactive phenotype. In contrast, fewer changes can decrease locomotor activity. Specifically, genetic and pharmacological alterations that enhance the dopamine, orexin, histamine, cannabinoids systems or that antagonize the cholinergic system induce an increase in locomotor activity. Similarly, imbalances in the two main neurotransmitters of the nervous system, GABA and glutamate usually result in hyperactive behaviour. It is remarkable that no genetic alterations pertaining to the GABA system have been reported to reduce locomotor behaviour. Other neurotransmitters, such as norepinephrine and serotonin, have a more complex influence. For instance, a decrease in norepinephrine synthesis usually results in hypoactive behaviour. However, a chronic increase in norepinephrine may result in hypoactivity too. Similarly, changes in both directions of serotonin levels may reduce locomotor activity, whereas alterations in specific serotonin receptors can induce hyperactivity. The lesion of at least 12 different brain regions can increase locomotor activity too. Comparatively, few focal lesions decrease locomotor activity. Finally, a large number of toxic events can increase locomotor activity, particularly if delivered during the prepuberal time window. These data show that there is a net imbalance in the number of altered genes/brain lesions/toxics that induce hyperactivity versus hypoactive behaviour. Although some of these data may be explained in terms of the activating role of subcortical systems (such as catecholamines), the larger number of alterations that induce hyperactivity suggests a different scenario. Specifically, we hypothesize (i) the existence of a control system that continuously inhibit a basally hyperactive locomotor tone and (ii) that this control system is highly vulnerable (intrinsic fragility) to any change in the genetic asset or to any toxic/drug delivered during prepuberal stages. Brain lesion studies suggest that the putative control system is located along an axis that connects the olfactory bulb and the enthorhinal cortex (enthorhinal-hippocampal-septal-prefrontal cortex-olfactory bulb axis). We suggest that the increased locomotor activity in many psychiatric diseases may derive from the interference with the development of this brain axis during a specific postnatal time window.

Central actions of arginine vasopressin and a V1a receptor antagonist on maternal aggression, maternal behavior, and grooming in lactating rats

Maternal aggression is a robust type of aggression displayed by lactating female rats. Although arginine vasopressin (AVP) has been implicated in the control of male aggression, its involvement in maternal aggression has not been thoroughly investigated. Previous neuroanatomical studies suggest that AVP may mediate the display of aggression during lactation. In the current study, AVP and an AVP V1a receptor antagonist were centrally administered to primiparous rats on days 5 and 15 of lactation, and aggression, maternal behavior, and grooming were recorded. Although AVP did not affect the number of attacks or duration of aggression, it increased the latency to initiate aggression on day 5, in addition to decreasing maternal behavior and increasing grooming. Conversely, V1a antagonist treatment increased maternal aggression on both days of lactation, decreased maternal behavior on day 15, and decreased grooming on day 5. Thus, it appears that central AVP activity modulates maternal aggression, as well as maternal behavior and grooming behavior during lactation.

Estrogen deficient male mice develop compulsive behavior

BACKGROUND

Aromatase converts androgen to estrogen. Thus, the aromatase knockout (ArKO) mouse is estrogen deficient. We investigated the compulsive behaviors of these animals and the protein levels of catechol-O-methyltransferase (COMT) in frontal cortex, hypothalamus and liver.

METHODS

Grooming was analyzed during the 20-min period immediately following a water-mist spray. Running wheel activity over two consecutive nights and barbering were analyzed. COMT protein levels were measured by Western analysis.

RESULTS

Six-month old male but not female ArKO mice develop compulsive behaviors such as excessive barbering, grooming and wheel-running. Excessive activities were reversed by 3 weeks of 17beta-estradiol replacement. Interestingly, the presentation of compulsive behaviors is accompanied by concomitant decreases (p < .05) in hypothalamic COMT protein levels in male ArKO mice. These values returned to normal upon 17beta-estradiol treatment. In contrast, hepatic and frontal cortex COMT levels were not affected by the estrogen status, indicating region- and tissue-specific regulation of COMT levels by estrogen. No differences in COMT levels were detectable between female animals of both genotypes.

CONCLUSIONS

This study describes the novel observation of a possible link between estrogen, COMT and development of compulsive behaviors in male animals which may have therapeutic implications in obsessive compulsive disorder (OCD) patients.

Estrogen receptor alpha and vasopressin in the paraventricular nucleus of the hypothalamus in Peromyscus

The purpose of this study was to determine the presence of estrogen receptor alpha (ERalpha) and the relationship between neurons that express ERalpha and produce vasopressin (AVP) in the paraventricular nucleus of the hypothalamus (PVN) in new world mice of the genus Peromyscus. Brains were collected from male and female Peromyscus californicus, Peromyscus leucopus, Peromyscus maniculatus, and Peromyscus polionotus, and double labeled for the expression of ERalpha and AVP immunoreactivity (IR). The number of cells expressing ERalpha-IR and AVP-IR was determined in the medial and posterior region of the PVN. The results indicate that Peromyscus is the first taxonomic group reported to have ERalpha widely distributed in the PVN, occurring in both medial and posterior regions of the PVN. While estrogen can regulate the production of AVP, AVP and ERalpha were rarely colocalized. There was, however, a significant inverse relationship between the number of cells that expressed ERalpha-IR and the number expressing AVP-IR. There were no sex differences in the expression of ERalpha-IR or AVP-IR.

Contrasting grooming phenotypes in C57Bl/6 and 129S1/SvImJ mice

Since C57 and 129 mice are the commonly used background strains, a better knowledge of all their behavioural characteristics is important in neuroscience research. Grooming is a complex and essential ritual in the rodent behavioural repertoire, normally proceeding in a cephalocaudal progression (paws-nose-face-body-legs-tail and genitals). Various stressors as well as genetic manipulations have been reported to alter mouse grooming and its patterning, underlying the importance of analysis of grooming behaviours in detail. Although strain differences between C57BL/6 and 129S1/SvImJ substrains have been assessed in many studies, no ethological analyses of their grooming have been performed. Here we show strain differences between these mice in spontaneous (novelty-induced) and artificial (water-induced) grooming. Overall, 129S1/SvImJ mice demonstrated less grooming activity, more interrupted and incomplete bouts, and more incorrect transitions (contrary to the cephalocaudal rule) between patterns, accompanied by lower vertical activity and higher defecation/urination in both tests. These results are consistent with general hypoactive anxious phenotype in 129S1/SvImJ mice and suggest that ethological analysis of mouse grooming may be used in neurobehavioural stress research, including behavioural phenotyping of both mutant and background mice.

Mouse grooming microstructure is a reliable anxiety marker bidirectionally sensitive to GABAergic drugs

Grooming is an important part of rodent behavioural repertoire, representing a complex hierarchically ordered cephalo-caudal sequence of patterns sensitive to stress and various drugs. Gamma-aminobutyric acid (GABA) is involved in the regulation of both anxiety and grooming behaviours. This study investigated the predictive validity of grooming behavioural microstructure as a marker of anxiety, by examining the effects of two GABAergic reference compounds, anxiolytic diazepam (0.1 and 0.5 mg/kg i.p.) and anxiogenic pentylenetetrazole (5 and 10 mg/kg i.p.) on mouse grooming. Our data suggest that the percentage of pattern transitions not fitting to the cephalo-caudal progression, and the percentage of interrupted grooming bouts are more reliable behavioural markers of stress bidirectionally sensitive to GABAergic anxiogenic and anxiolytic drugs, compared to the frequency and duration scores. Our study also confirms that detailed ethological analyses of grooming microstructure can be a useful tool in behavioural pharmacology of anxiety.

Role of L-glutamate in systemic AVP-induced hypothermia

It has been reported that systemic injection of arginine vasopressin (AVP) induces a drop in body core temperature (T(c)), but little is known about the mechanisms involved. Because glutamate is an important excitatory neurotransmitter involved in a number of thermoregulatory actions, in the present study, we tested the hypothesis that glutamate plays a role in systemic AVP-induced hypothermia. Wistar rats were pretreated intracerebroventricularly (icv) with kynurenic acid, an antagonist of l-glutamate ionotropic receptors, alpha-methyl-(4-carboxyphenyl)glycine (MCPG), an antagonist of l-glutamate metabotropic receptors, or saline 15 min before intravenous injection of AVP (2 microg/kg) or saline. T(c), brown adipose tissue (BAT) temperature, blood pressure, heart rate, and tail skin temperature were measured continuously. Administration of saline icv followed by intravenous AVP caused a significant drop in T(c) brought about by a reduction in BAT thermogenesis and an increase in heat loss through the tail. MCPG treatment (icv) did not affect the fall in T(c) induced by AVP. Treatment with kynurenic acid (icv) abolished AVP-induced hypothermia but did not affect the AVP-evoked rise in blood pressure or drop in heart rate and BAT temperature. Heat loss through the tail was significantly reduced in animals injected with AVP and pretrated with kynurenic acid. These data indicate that ionotropic receptors of l-glutamate in the central nervous system participate in peripheral AVP-induced hypothermia by affecting heat loss through the tail.

Systemic salt loading decreases body temperature and increases heat-escape/cold-seeking behaviour via the central AT1 and V1 receptors in rats

Salt loading decreases body core temperature (T(core)) at neutral ambient temperature (26 degrees C) and increases heat-escape/cold-seeking behaviour in desalivated rats. In this study, we tested the hypothesis that brain angiotensin II (AII) and arginine vasopressin (AVP) are associated with these responses. Surgically desalivated rats (n = 28) were administered an injection (S.C., 10 ml kg(-1)) of either normal saline (154 mM, NS) or hypertonic saline (2500 mM, HS) following an intracerebroventricular injection (10 microl kg(-1)) of an AII AT(1)-receptor antagonist (candesartan, 5 microg microl(-1)), an AVP V(1)-receptor antagonist ((beta-mercapto-beta, beta-cyclopenta-methylene propionyl(1), O-Me-Tyr(2), Arg(8))-vasopressin, 0.5 microg microl(-1)), or normal saline (154 mM). Each rat was placed in a behaviour box, first at 26 degrees C for 1 h to allow the measurement of baseline T(core) and movement. The ambient temperature was then elevated to 40 degrees C for the next 2 h, during which time the rat was able to trigger a 0 degrees C air reward for 30 s by moving into a specific area of the box (operant behaviour). The S.C. HS significantly decreased baseline T(core) at 26 degrees C (36.5 +/- 0.1 degrees C) and increased counts of operant behaviour at 40 degrees C (57 +/- 3) compared with results obtained following S.C. NS injection (37.4 +/- 0.1 degrees C and 42 +/- 1, respectively). These responses to s.c. HS were inhibited by the intracerebroventricular injection of AT(1) (37.3 +/- 0.1 degrees C and 43 +/- 2, respectively; P < 0.05) and V(1) antagonists (37.2 +/- 0.2 degrees C and 42 +/- 2, respectively; P < 0.05), although administration of both antagonists with S.C. NS had no effect. These results suggest that brain AII and AVP are involved in the decrease in T(core) observed at neutral ambient temperature and the increase in heat-escape/cold-seeking behaviour in response to osmotic stimulation, via the central AT(1) and V(1) receptors, respectively