Preoptic mechanism for cold-defensive responses to skin cooling

Abstract

We recently identified a somatosensory pathway that transmits temperature information from the skin to a median subregion of the preoptic area (POA), a thermoregulatory centre. Here, we investigated in vivo the local neuronal circuit in the rat POA that processes the thermosensory information and outputs thermoregulatory effector signals. Skin cooling-evoked increases in sympathetic thermogenesis in brown adipose tissue, in metabolism and in heart rate were reversed by inhibition of neurons in the median preoptic nucleus (MnPO). Glutamatergic stimulation or disinhibition of MnPO neurons evoked thermogenic, metabolic and cardiac responses that mimicked the cold-defensive responses to skin cooling and were reversed by antagonizing GABA_A receptors in the medial preoptic area (MPO), which is thought to contain neurons providing thermoregulatory output to effectors. These results suggest that GABA inhibition of output neurons in the MPO by MnPO neurons that are activated by cool sensory signals from the skin is a core thermoregulatory mechanism within the POA that is essential for the feedforward defence of body temperature against cold challenges in the environment.

Homeothermic animals, including humans, defend their body temperature against changes in environmental temperature through thermoregulatory mechanisms that are governed by the central nervous system. The preoptic area (POA), which is located rostral to the hypothalamus, plays pivotal roles in the defence of body temperature by integrating information on peripheral and central temperatures and by providing appropriate command signals to peripheral thermoregulatory effectors (Hammel, 1968; Boulant & Hardy, 1974; Nagashima et al. 2000; Nakamura & Morrison, 2007; Romanovsky, 2007). To initiate appropriate thermoregulatory responses to changes in environmental temperature before they affect body core temperature, the POA needs to receive feedforward information on environmental temperature (Huckaba et al. 1971; Boulant & Gonzalez, 1977; Savage & Brengelmann, 1996; Nakamura & Morrison, 2008), which is sensed by thermoreceptors in the cutaneous terminals of primary somatosensory neurons (Lumpkin & Caterina, 2007). Recently, we described a neuronal pathway responsible for the transmission of such thermosensory information from the skin to the POA (Nakamura & Morrison, 2008). In this pathway, a direct glutamatergic projection from the lateral parabrachial nucleus to a median subregion of the POA centred in the median preoptic nucleus (MnPO) transmits cool signals coming from the somatosensory relay mechanism in the dorsal horn (Nakamura & Morrison, 2008). The present study examines how the thermosensory signals to MnPO neurons are processed in the local circuitry of the POA to elicit descending command signals.

The output neurons in the POA that control caudal thermogenic brain regions, such as the dorsomedial hypothalamus and rostral medullary raphe, are thought to be inhibitory neurons that are tonically active when thermogenesis is not needed (Nagashima et al. 2000; Nakamura et al. 2002, 2005; Nakamura & Morrison, 2007). For example, transecting the output fibres from the POA activates sympathetic thermogenesis in brown adipose tissue (BAT) (Chen et al. 1998), a thermogenic organ in rodents (Cannon & Nedergaard, 2004) and humans (Nedergaard et al. 2007). Furthermore, GABA inhibition of neurons in the medial POA (MPO), but not in the MnPO or lateral POA (LPO), increases metabolism and body temperature (Osaka, 2004) and antagonizing GABA_A receptors on neurons in the MPO suppresses BAT thermogenesis evoked by skin cooling (Nakamura & Morrison, 2007). Therefore, we hypothesized the existence of GABAergic interneurons in the MnPO that are activated by cool signals from the skin and reduce the tonic activity of inhibitory output neurons distributed within the MPO, thereby leading to facilitation of sympathetic thermogenesis through disinhibition of BAT sympathoexcitatory neurons in caudal brain regions.

In the present study, we tested this hypothesis by using an in vivo physiological approach to determine (1) whether activation of MnPO neurons mediates cold-defensive responses to skin cooling, (2) whether stimulation of MnPO neurons evokes physiological responses mimicking those observed during cold defence, and (3) how GABA_A receptor antagonism in the MPO affects the physiological responses triggered by MnPO activation.

Methods

Animal preparation

Twenty-two male Sprague–Dawley rats (290–510 g) contributed to the present study. The animals were housed with ad libitum access to food and water in a room air-conditioned at 22–23°C with a standard 12 h light–dark cycle. All procedures conform to the regulations detailed in the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Animal Care and Use Committee of the Oregon Health and Science University.

Animal preparation basically followed our previous procedure (Nakamura & Morrison, 2007). Rats were anaesthetized with intravenous urethane (0.8 g kg⁻¹) and α-chloralose (80 mg kg⁻¹). Adequacy of anaesthesia was verified by the absence of a hindlimb withdrawal to foot pinch and/or by the absence of eye blink response to gentle touch of the cornea. Arterial pressure and heart rate (HR) were recorded from a cannulated femoral artery. The trunk was shaved, a copper-constantan thermocouple was taped to the abdominal skin to monitor skin temperature, and the trunk was wrapped with a plastic water jacket. The animal was positioned in a stereotaxic apparatus, and a needle-type thermocouple was perpendicularly inserted into the basal forebrain to monitor brain temperature. The animal was then artificially ventilated with 100% O₂ through a tracheal cannula and subjected to neuromuscular blockade with d-tubocurarine (0.6 mg i.v. initial dose, supplemented with 0.3 mg h⁻¹) to stabilize BAT nerve recording by preventing respiration-related movements as well as shivering movements that could be caused by skin cooling or drug injections into the brain. Every time the effect of tubocurarine waned, the depth of anaesthesia was re-assessed prior to supplementation of tubocurarine and the anaesthetic was supplemented as necessary. Mixed expired CO₂ was measured to provide an index of changes in whole body metabolism and was maintained between 3.5 and 4.5% under basal conditions. Rectal temperature was monitored with a thermocouple as an indication of core body temperature.

Postganglionic BAT sympathetic nerve activity (SNA) was recorded from the central cut end of a nerve bundle isolated from the right interscapular BAT pad. Nerve activity was filtered (1–300 Hz) and amplified (×2000–50 000) with a CyberAmp 380 (Axon Instruments, Union City, CA, USA). To obtain a continuous measure (4 s bins) of BAT SNA amplitude, the root mean square amplitude of the BAT SNA was calculated (Spike 2, CED, Cambridge, UK) as the square root of the total power in the 0–20 Hz band of the autospectra of sequential 4 s segments of BAT SNA. BAT temperature (T_BAT) was monitored with a thermocouple inserted into the intact left interscapular BAT pad. Physiological variables were digitized and recorded to a computer hard disk using Spike 2.

Experimental procedure and data analysis

Experiment 1

All the animals exhibited consistent cold-defensive physiological responses to repeated cooling of the trunk skin by perfusing the water jacket with cold water (Nakamura & Morrison, 2007). Each cooling episode lasted for 150–200 s until lowering the skin temperature to a level consistent with preceding cooling episodes and then the skin was rewarmed by switching to perfusion with warm water. After a few cooling episodes, the effect of a nanoinjection (100–200 nl) of saline vehicle or glycine (0.5 m) into the MnPO on cold-defensive responses evoked by a subsequent cooling episode was examined. Testing the effect of saline was followed by testing that of glycine in each animal. Baseline values of all physiological variables were the averages during the 1 min period immediately before skin cooling. Skin cooling-evoked response values for T_BAT, expired CO₂ and HR were obtained at the end of the skin cooling episode and those for BAT SNA and mean arterial pressure (MAP) were the averages during the 1 min period immediately before the end of skin cooling. Skin cooling-evoked changes in these variables from their baseline values were compared between skin cooling episodes after saline and after glycine injections into the MnPO (Fig. 1D). Statistical significance was evaluated with Student's two-tailed paired t test.

Figure 1. Glycine injection into the MnPO blocks thermogenic, metabolic and cardiac responses to skin cooling.

A, changes in skin temperature (T_skin), BAT SNA, T_BAT, expired (Exp.) CO₂, HR, arterial pressure (AP), rectal temperature (T_rec) and brain temperature (T_brain) that are evoked by repeated skin cooling. Saline or glycine was nanoinjected into the MnPO at the broken lines. The vertical scale bar for the BAT SNA trace represents 400 μV. B, location of the sites of saline and glycine injections. 3V, third ventricle; ac, anterior commissure; ox, optic chiasm. C, representative view of a site of an injection into the MnPO. The injection site is clearly identified as a cluster of fluorescent beads (arrow). D, group data showing the effect of saline or glycine injection into the MnPO on skin cooling-evoked changes in physiological variables (n = 4). *P < 0.05; **P < 0.01; ***P < 0.001, compared with the saline-injected group.

Experiment 2

Animals received a nanoinjection (60 nl) of saline, NMDA (0.2 mm) or the GABA_A receptor antagonist, (–)-bicuculline methiodide (0.2 mm), unilaterally into the MnPO, MPO or LPO. To prevent thermogenesis dependent on ambient temperature, skin temperature was maintained between 36 and 38°C by perfusing the water jacket with warm water during the experiment. Since skin cooling-evoked responses elicited 30 min after intra-POA drug injections were not different from those during the original skin cooling challenge, we were able to perform injections into multiple POA subregions in most animals, allowing at least a 30 min recovery period between injections.

In experiment 2, baseline values of all physiological variables were the averages during the 1 min period immediately before injection into a POA subregion. Due to a difference in the durations of the drug-evoked responses, we used a response analysis time of 3 min postinjection for NMDA and its accompanying saline injections and 10 min postinjection for bicuculline and its accompanying saline injections. The drug-evoked changes in T_BAT, expired CO₂, HR and MAP were the differences between their peak values within the response analysis time and their baseline values. The drug-evoked changes in BAT SNA were calculated by dividing the area under the curve (AUC) of the ‘power per 4 s’ trace above the baseline level for the response analysis time by the area under the baseline (basal AUC) during a comparable time period just prior to the injection. Drug-evoked changes in variables were compared to those evoked by saline injections (Fig. 2F). Statistical significance was evaluated with a two-tailed unpaired t test.

Figure 2. NMDA or bicuculline injection into the MnPO, but not that in the MPO or LPO, evokes thermogenic, metabolic and cardiac increases.

A–D, changes in physiological variables that are evoked by a nanoinjection of NMDA into the MnPO (A), NMDA into the MPO (B), bicuculline into the MnPO (C) or saline into the MnPO (D). The vertical scale bars for the BAT SNA trace represent 50 μV (A, C and D) and 200 μV (B). E, composite drawing of sites of saline, NMDA or bicuculline injections into POA subregions with their stimulatory effects on BAT SNA during 3 min (NMDA) or 10 min (bicucuclline) after the injection. No saline injections increased BAT SNA by > 50% (of basal AUC) during 3 min nor > 200% during 10 min after the injection. F, group data showing the effect of injection of saline into the MnPO (n = 7), NMDA into the MnPO (n = 7), NMDA into the MPO and LPO (n = 11), bicuculline into the MnPO (n = 6) or bicuculline into the MPO and LPO (n = 13) on the physiological variables. *P < 0.05, **P < 0.01, ***P < 0.001, compared with the saline-injected group. †††P < 0.001, compared with the group of NMDA or bicuculline injection into the MnPO.

Experiment 3

Bicuculline (2 mm) was nanoinjected (60 nl) into the MnPO in animals whose skin temperature was maintained as in experiment 2. During the responses evoked by bicuculline into the MnPO, bilateral nanoinjections (60 nl each) of saline or bicuculline (2 mm) were made into the MPO. Baseline values of all physiological variables were the averages during the 1 min period immediately before bicuculline injection into the MnPO. The values of the increases in T_BAT, expired CO₂ and HR evoked by bicuculline into the MnPO were taken immediately before the first of the subsequent bilateral bicuculline injections into the MPO and those for BAT SNA and MAP were the averages during the 1 min period immediately before the first injection into the MPO. To determine the effect of the bilateral bicuculline injections into the MPO, values for T_BAT, expired CO₂ and HR were taken at 4 min after the completion of the bilateral injections and those for BAT SNA and MAP were the averages during the 1 min period beginning at 3 min after the completion of the bilateral injections. For each variable, the statistical significance of the change due to the bicuculline injections into the MPO was detected with a two-tailed paired t test comparing the peak response value following bicuculline into the MnPO and the value 4 min after bilateral bicuculline into the MPO (Fig. 3C).

Figure 3. Bicuculline injections into the MPO reverse thermogenic, metabolic and cardiac increases evoked by MnPO stimulation.

A and B, changes in physiological variables after a bicuculline nanoinjection into the MnPO followed by bilateral nanoinjections of bicuculline (A) or saline (B) into the MPO. The vertical scale bars for the BAT SNA trace represent 200 μV (A) and 250 μV (B). C, group data showing the effect of saline (n = 4) or bicuculline (n = 5) injections into the MPO on increases in the physiological variables that were evoked by bicuculline injection into the MnPO. *P < 0.05, **P < 0.01, compared with the increase value. D, location of the sites of saline and bicuculline injections. E, representative view of sites of bilateral injections into the MPO (arrows).

All data are presented as the means ± s.e.m. and statistical results with a P-value of < 0.05 were considered significant.

All drugs were dissolved in 0.9% saline and pressure-ejected into the brain at appropriate stereotaxic coordinates through glass micropipettes (tip inner diameter, 20–30 μm). To mark the injection sites, 10–20 nl injections of 0.2% fluorescent microspheres (diameter, 0.1 μm; Invitrogen) in saline were made at the same coordinates through the same micropipette. After the physiological recordings, the animals were transcardially perfused with a 10% formaldehyde solution and the brain tissue was sectioned. The locations of the injections were identified by detecting the fluorescent microspheres under an epifluorescence microscope.

Results

Cold-defensive responses to skin cooling requires activation of MnPO neurons

We tested whether induction of cold-defensive physiological responses to reduced environmental temperature involves activation of MnPO neurons. Repeated cooling of the trunk skin with a water jacket evoked consistent increases in BAT SNA and T_BAT, indicators of sympathetic thermogenic activity; in expired CO₂, an indicator of metabolic rate; and in HR, an indicator of cardiac output (Fig. 1A; Nakamura & Morrison, 2007). We examined the effect of inhibition of MnPO neurons with nanoinjection of glycine (Fig. 1B and C), a widely used inhibitor of neurons, on these skin cooling-evoked responses. Skin cooling following a saline injection into the MnPO increased these physiological variables to the levels comparable to those evoked by preinjection skin cooling challenges (n = 4; Fig. 1A). In contrast, skin cooling following a glycine injection into the MnPO failed to evoke any of the thermogenic, metabolic or cardiac responses (P < 0.05 versus saline injection, n = 4, paired t test; Fig. 1A and D). Following the decay of this inhibitory effect of glycine in 5–10 min, the skin cooling-evoked physiological responses recovered and the second injection of glycine into the MnPO again completely suppressed the skin cooling-evoked responses (Fig. 1A). Recovery of the skin cooling-evoked physiological responses after the second glycine injection was slower than that after the first one (Fig. 1A). These results indicate that activation of MnPO neurons is an essential central process to evoke cold-defensive responses to reduced environmental temperature.

Glutamatergic stimulation or disinhibiton of MnPO neurons, but not that of MPO or LPO neurons, evokes thermogenic responses

To functionally delineate the POA subregions whose activation leads to thermogenesis, we nanoinjected stimulatory reagents into various sites in the POA under thermoregulatory conditions in which BAT SNA was quiescent and examined their effects on thermogenic, metabolic and cardiovascular activities. In contrast to no effect of saline injection into the MnPO (n = 7; Fig. 2D), NMDA injection into the MnPO evoked short, but significant, increases in BAT SNA, T_BAT, expired CO₂ and HR (P < 0.01 versus saline injection, n = 7, unpaired t test; Fig. 2A and F), consistent with glutamatergic transmission of cutaneous cool signals to the MnPO (Nakamura & Morrison, 2008). To determine whether MnPO neurons whose activation leads to thermogenesis receive a tonic GABAergic inhibition, we nanoinjected the GABA_A receptor antagonist, bicuculline, into the MnPO. Bicuculline injection into the MnPO evoked increases in BAT SNA, T_BAT, expired CO₂ and HR (P < 0.01 versus saline injection, n = 6; Fig. 2F) that appeared longer and more intense than those following NMDA injection into the MnPO (Fig. 2A and C). In contrast, neither NMDA (n = 11) nor bicuculline (n = 13) injected into the MPO or the LPO significantly increased any of these physiological variables (P > 0.05 versus saline injection; Fig. 2B and F); rather NMDA injections into these areas slightly, but significantly, reduced expired CO₂ and HR from their basal levels (P < 0.05; Fig. 2B and F). Comparison of the sites of NMDA and bicuculline injections and the resulting stimulations of BAT SNA (Fig. 2E) clearly delineates similar effective regions for the two compounds that were confined to the MnPO. This functional mapping study indicates that glutamatergic activation of neurons in the MnPO is sufficient to trigger cold-defensive responses and that these neurons are tonically inhibited by GABA inputs under conditions of BAT SNA quiescence.

Thermogenic responses evoked from the MnPO require GABA transmission in the MPO

To determine whether activation of MnPO neurons evoked thermogenic responses through a GABA-mediated inhibition of neurons in the MPO, possibly those output neurons that control caudal thermogenic brain regions, we examined the effect of blockade of GABA_A receptors in the MPO with bicuculline on the physiological responses evoked following activation of neurons in the MnPO. The large increases in BAT SNA, T_BAT, expired CO₂ and HR that were evoked by blockade of GABAergic inhibition in the MnPO with bicuculline were rapidly reversed by bilateral injections of bicuculline into the MPO (P < 0.05, peak increase values versus effect values after injections into the MPO, n = 5, paired t test; Fig. 3A and C). In contrast, bilateral saline injections into the MPO did not significantly affect the sustained increases in these variables evoked by bicuculline into the MnPO (P > 0.05, n = 4; Fig. 3B and C). The bicuculline injections into the MPO inhibited BAT SNA, T_BAT, expired CO₂ and HR, with respective changes of −94 ± 4%, −48 ± 14%, −63 ± 9% and −95 ± 5% from their increased levels following bicuculline into the MnPO (Fig. 3C). All of these changes were significantly different (P < 0.01; unpaired t test) from the respective changes after saline injections into the MPO: BAT SNA, −11 ± 15%; T_BAT, +42 ± 18%; expired CO₂, +15 ± 13%; and HR, +24 ± 28% (Fig. 3C). Figure 3D indicates that the initial bicuculline injection sites were restricted to the midline MnPO, dorsal to the third ventricle and that the subsequent bilateral bicuculline injections (Fig. 3E) were all in the MPO, at sites lateral to the MnPO. These results indicate that the thermogenic, metabolic and cardiac responses evoked by activation of MnPO neurons are dependent on a GABAergic inhibition of neurons in the MPO.

Discussion

The present results provide the first indication of the local neuronal mechanism within the POA through which temperature information from the skin can alter the efferent command signals from the POA to evoke appropriate thermoregulatory effector responses. Consistent with our recent demonstration that the POA subregion centred in the midline MnPO receives excitatory cutaneous cool signals (Nakamura & Morrison, 2008), we have shown that BAT thermogenic, metabolic and tachycardic responses to skin cooling are dependent on the activation of neurons in the MnPO and that stimulation of MnPO neurons itself evokes similar responses in a manner requiring induction of a GABAergic inhibition on neurons in the MPO. As shown in the scheme in Fig. 4, we propose that thermal homeostasis is defended from reductions in environmental temperature by a mechanism involving the feedforward activation of a population of GABAergic inhibitory interneurons in the MnPO that reduce the tonic discharge of inhibitory output neurons in the MPO, thereby allowing activation of caudal brain regions, including the dorsomedial hypothalamus and the rostral medullary raphe, that drive thermogenic responses in BAT as well as increases in metabolism and HR.

Figure 4. Schematic model of the mechanism for cold-defensive responses to thermosensory signals from the skin.

In thermoneutral environments, GABAergic projection neurons in the MPO tonically inhibit excitatory pathways that drive cold-defensive responses including BAT thermogenesis. In cold environments, environmental cooling is sensed by thermoreceptors in the cutaneous endings of somatosensory neurons, which transmit the thermal signals from the skin to the dorsal horn (DH). Glutamatergic inputs (Glu) from the DH neurons activate neurons in the lateral parabrachial nucleus (LPB), which then transmit the signals to the MnPO through their glutamatergic projections. This feedforward thermal afferent input activates GABAergic interneurons in the MnPO, which reduce the tonic activity of the inhibitory projection neurons in the MPO. The resulting disinhibition of the descending efferent pathways that excite thermoregulatory effectors leads to induction of cold-defensive responses to counteract the impact of the reduced environmental temperatures. Filled red, filled blue and open black circles denote cell bodies of activated excitatory neurons, activated inhibitory neurons and suppressed neurons, respectively. DRG, dorsal root ganglion.

Inhibition of MnPO neurons with glycine completely blocked cold-defensive responses to skin cooling and glutamatergic activation of MnPO neurons with NMDA evoked thermogenic, metabolic and cardiac responses that mimicked cold-defensive responses to skin cooling. Importantly, sites of NMDA injections that evoked these responses were clearly confined to the MnPO. These results are consistent with our previous findings that the MnPO receives direct projections from lateral parabrachial neurons that are activated by cutaneous cool signals and that antagonizing glutamate receptors in the MnPO blocks thermogenic responses that are evoked by stimulation of the lateral parabrachial nucleus (Nakamura & Morrison, 2008). These data provide strong support for the view that MnPO neurons can be activated by cool thermosensory signals transmitted by a direct glutamatergic input from lateral parabrachial neurons (Fig. 4).

The ability of bicuculline injected into the MPO to eliminate the thermogenic, metabolic and cardiac responses evoked either by activation of MnPO neurons (present study) or by skin cooling (Osaka, 2004; Nakamura & Morrison, 2007) demonstrates the dependence of these responses on induction of a GABA-mediated inhibition on neurons in the MPO. The level of tonic activity of inhibitory projection neurons in the MPO that control thermogenesis-promoting neurons in caudal brain regions, such as the dorsomedial hypothalamus and rostral medullary raphe, is thought to determine the overall level of thermoregulatory thermogenesis, metabolism and cardiac output (Nagashima et al. 2000; Nakamura et al. 2002, 2005; Nakamura & Morrison, 2007). This notion is mostly based on the findings that GABA inhibition of neurons in the MPO increases metabolism, body temperature and HR (Osaka, 2004; Ishiwata et al. 2005) and that transecting POA output pathways increases BAT thermogenesis (Chen et al. 1998). Consistent with this notion, stimulation of neurons in the MPO with NMDA slightly, but significantly, reduced the basal metabolic and cardiac levels in the present study. Whether the inhibitory tonic control by MPO neurons is also involved in the regulation of heat loss through cutaneous vasoconstriction remains unclear. Lesion of a rostral part of the POA, which seems to damage some MPO neurons, evokes hyperthermia without inducing cutaneous vasoconstriction (Romanovsky et al. 2003). In contrast, hyperthermia evoked by a lesion that is limited to the MPO is accompanied by cutaneous vasoconstriction as well as increased metabolism and shivering (Szymusiak & Satinoff, 1982).

The idea that MnPO neurons provide a local GABAergic connection to the MPO (Fig. 4) is also supported by previous anatomical observations that some MnPO neurons innervate the MPO (Uschakov et al. 2007) and that the MnPO contains many GABA neurons (Nakamura et al. 2002; Gong et al. 2004). Furthermore, many neurons in the MnPO, rather than the MPO or LPO, are activated (express Fos protein) in response to reduced environmental temperature (Bratincsák & Palkovits, 2004) and the extracellular level of GABA in the POA is elevated during cold exposure and reduced during heat exposure in free-moving rats (Ishiwata et al. 2005).

In addition to thermosensory information from the skin, local temperature in the POA also affects thermoregulatory functions (Hammel et al. 1960; Baldwin & Ingram, 1967; Imai-Matsumura et al. 1984) and the POA contains abundant thermosensitive neurons (Nakayama et al. 1961), most of which are activated by local warm temperatures (warm-sensitive neurons; Boulant & Dean, 1986). Since brain temperature is far less susceptive to changes in environmental temperature than skin temperature in many homeothermic mammals (Hammel, 1968), it seems likely that the local thermosensation of POA neurons would play a role (a) in setting the basal tone of thermoregulatory effector efferents, (b) in enhancing thermoregulatory responses in situations of extreme thermal environments when the feedforward thermoregulatory responses driven by changes in skin temperature have proven inadequate to prevent changes in brain or body core temperature, and (c) in responding to challenges to thermal homeostasis involving changes in temperature within the body, as might occur during exercise, intake of cold fluids or haemorrhage. Warm-sensitive neurons in the POA are tonically active at neutral local temperatures (Nakayama et al. 1961) and skin cooling reduces the tonic discharge of warm-sensitive POA neurons (Boulant & Hardy, 1974), suggesting that warm-sensitive neurons in the MPO integrate both cutaneous and local thermal information. Furthermore, that the activity of warm-sensitive POA neurons is expected to inhibit BAT thermogenesis and that transecting POA output pathways increases BAT thermogenesis (Chen et al. 1998) are consistent with our model (see above) in which warm-sensitive neurons project from the MPO to inhibit BAT thermogenesis, although whether warm-sensitive neurons send axons outside of the POA remains unknown.

The POA is also known as a site of action of prostaglandin (PG) E₂ for its pyrogenic function. That PGE₂, which is produced in response to inflammatory immune signals, can trigger fever by reducing the activity of inhibitory neurons in the MPO that express the EP3 subtype of PGE receptors and that project to the thermogenesis-promoting caudal brain regions (Nakamura et al. 2002, 2005) raises the possibility that the MPO neurons integrating thermal information also detect pyrogenic PGE₂ signals. A recent finding that deletion of EP3 receptors in the POA eliminates febrile responses to PGE₂ or endotoxin (Lazarus et al. 2007) further enhances the importance of EP3-expressing MPO neurons in fever development. However, rats that have a lesion of the whole POA, which cannot defend their body temperatures even in moderately cold or heated environment, exhibit intact warm-seeking behaviour in response to endotoxin (Almeida et al. 2006), suggesting the existence of a POA-independent mechanism for behavioural thermoregulation during inflammatory fever.

The lamina terminalis, including the MnPO, detects changes in the tonicity of body fluids and provides effector signals regulating fluid homeostasis (McKinley et al. 1999) and the systems for osmoregulation and thermoregulation are closely related. Actually, lesion of the MnPO cancelled enhancement of thermoregulatory behaviours by systemic salt loading (Konishi et al. 2007). Thus, MnPO neurons are likely to integrate thermosensory and osmosensory information and provide the integrated signals to effector pathways leading to either thermoregulation or osmoregulation.

The present study provides evidence for an interaction of two distinct functional groups of thermoregulatory neurons within the POA: one that receives thermosensory information and provides inhibitory regulation of another that produces efferent signals controlling thermoregulatory effectors. Within the limitations of the nanoinjection technique, our findings suggest that these functional groups of POA neurons are anatomically separate, occurring primarily in the MnPO and the MPO subregions of the POA, respectively. Each of the groups seems to have a specific role in the integration of multiple homeostatic sensory signals. Although this study has only demonstrated the importance of this local inhibitory circuit within the POA in the regulation of BAT thermogenesis and HR, the POA is involved in controlling other thermoregulatory functions such as cutaneous vasoconstriction, shivering and piloerection (Nagashima et al. 2000) and it will be of interest to determine if a similar local POA mechanism underlies the activation of these thermoregulatory effectors in response to stimulation of cutaneous cold receptors. Coordinated control of local circuits for the feedforward regulation of different thermoregulatory effectors would seem necessary for the POA to orchestrate the integrated responses that maintain thermal homeostasis.