Dopaminergic input from the posterior hypothalamus to the raphe pallidus area inhibits brown adipose tissue thermogenesis

Abstract

Systemic administration of dopamine (DA) receptor agonists leads to falls in body temperature. However, the central thermoregulatory pathways modulated by DA have not been fully elucidated. Here we identified a source and site of action contributing to DA’s hypothermic action by inhibition of brown adipose tissue (BAT) thermogenesis. Nanoinjection of the type 2 and type 3 DA receptor (D2R/D3R) agonist, 7-hydroxy-N,N-di-n-propyl-2-aminotetralin (7-OH-DPAT), in the rostral raphe pallidus area (rRPa) inhibits the sympathetic activation of BAT evoked by cold exposure or by direct activation of N-methyl-d-aspartate (NMDA) receptors in the rRPa. Blockade of D2R/D3R in the rRPa with nanoinjection of SB-277011A increases BAT thermogenesis, consistent with a tonic release of DA in the rRPa contributing to inhibition of BAT thermogenesis. Accordingly, D2Rs are expressed in cold-activated and serotonergic neurons in the rRPa, and anatomical tracing studies revealed that neurons in the posterior hypothalamus (PH) are a source of dopaminergic input to the rRPa. Disinhibitory activation of PH neurons with nanoinjection of gabazine inhibits BAT thermogenesis, which is reduced by pretreatment of the rRPa with SB-277011A. In conclusion, the rRPa, the site of sympathetic premotor neurons for BAT, receives a tonically active, dopaminergic input from the PH that suppresses BAT thermogenesis.

INTRODUCTION

Adipose tissue has a crucial role in metabolism, buffering fuel substrate, producing adipokines, and generating heat (1). The contribution of brown adipose tissue (BAT) thermogenesis to basal energy expenditure has been pursued as a potential therapeutic target to treat dyslipidemia, hyperglycemia, and obesity (2–4). To increase our understanding of the central nervous system (CNS) pathways and neurotransmitters influencing the sympathetic regulation of BAT energy expenditure, the present study examined the dopaminergic influence in the rostral raphe pallidus area (rRPa), the area containing the sympathetic premotor neurons for BAT (5).

Systemic administration of dopamine (DA) receptor subtype 2 and 3 (D2R/D3R) agonists causes hypothermia by inhibiting BAT thermogenesis (6) and by reducing the body temperature at which muscular shivering occurs (7). In contrast, the loss of dopaminergic neurons in models of Parkinson’s disease results in an increased protein content of tyrosine hydroxylase (TH) and uncoupling protein 1 (UCP1) in the BAT (8). These data suggest that the overall effect of the dopaminergic regulation of CNS thermoregulatory circuits would include a reduced energy expenditure through inhibition of BAT thermogenesis and shivering, and it may help to explain some of the thermoregulatory dysfunction seen in Parkinson’s disease (e.g., Parkinsonism-hyperpyrexia syndrome) (9).

Interestingly, treatment with a D2R/D3R agonist reduces core temperature and locomotor activity in wild-type and D3R knockout mice (10) but not in D2R knockout mice (11). These data provide a robust demonstration that the hypothermic action of DA is mediated via the D2R rather than the D3R. However, the site of action of DA within the central thermoregulatory network, as well as the identity of the relevant DA cell population, remains undefined. In this context, dense labeling of DA fibers can be found at the rRPa (12). The rRPa is a critical thermoregulatory area for rodents (13) and humans (14), since the rRPa contains sympathetic premotor neurons (SPmNs) that provide the excitatory drive to spinal sympathetic preganglionic neurons controlling BAT thermogenesis (15). The rRPa also harbors the sympathetic premotor neurons controlling cutaneous vasoconstriction (16) and the somatic premotor neurons for shivering (17). The activity of the BAT SPmNs in the rRPa is modulated by the balance of their inhibitory inputs [GABAergic (18), serotonergic (19, 20), glycinergic (21), and muscarinic (22)] and excitatory inputs [glutamatergic (23) and orexinergic (24)]. In this study, we expand the neurotransmitter influences on the activity of BAT SPmNs in the rRPa to include a tonically active, DA input from a subset of catecholaminergic neurons in the posterior hypothalamus (PH) that inhibits BAT thermogenesis.

MATERIALS AND METHODS

Ethics Statement

This study was carried out in strict accordance with the regulations detailed in the National Institutes of Health Guide for the Care and Use of Laboratory Animals and was approved by the Animal Care and Use Committee of the Oregon Health and Science University.

Animals

The data were obtained from 31 male Sprague-Dawley rats (330–450 g) obtained from Charles River Laboratories (Indianapolis, IN) that were housed with ad libitum access to food and water in a vivarium with a subthermoneutral temperature of 22–23°C and a 12:12-h light/dark cycle.

Anatomical Studies

For anatomical tract tracing studies and for the identification of DA receptor-expressing neurons in the rRPa, seven rats were used. In three rats, to identify the source of DA projections to the rRPa, we injected the retrograde tracer, cholera toxin subunit B (CTb), in the rRPa in conjuction with labeling for tyrosine hydroxylase (TH) immunoreactivity to identify DA neurons. In four other rats, the anterograde tracer, biotinylated dextran amine (BDA), was injected in the posterior hypothalamus (PH), the only brain region found to contain DA neurons double labeled for both CTb and TH. These rats were housed in an environmental chamber held at 23°C for 7 days to allow for BDA transport and to avoid the stress of being moved to a novel environment. On the seventh day, rats were cold exposed (10°C, attained within 15 min of changing the temperature setting of the environmental chamber) for 2 h before euthanasia. Brain sections through the rRPa were subsequently labeled immunohistochemically for the DA receptor subtype 2 (D2R) and tryptophan hydroxylase (TPH) to identify serotonergic neurons, BDA, and cold-evoked c-fos.

Adult male rats were anesthetized with 3% isoflurane in 100% O₂ and stereotaxically injected with CTb (1 mg/mL, 120 nL) conjugated to Alexa 594 into the rRPa (coordinates: 2.8–3.0 mm caudal to lambda, 0.0 mm lateral to the midline, −9.2 to 9.8 mm below the dural surface) or with BDA in the PH (coordinates: −5.5 to −6.0 mm caudal to the bregma, 0.3 mm lateral to the midline, −7.5 to −8.5 mm below the dural surface). The pipette was left in place for 10 min. Rats were treated with antibiotics (40,000 U/kg penicillin G im) and analgesics (0.05 mg/kg buprenorphine im). After 7 days, BDA-injected rats were exposed to skin cooling to achieve a core temperature of 35°C. Rats were then deeply anesthetized with pentobarbital (80 mg/kg ip) and transcardially perfused with saline followed by 4% paraformaldehyde. The brains were postfixed in 4% paraformaldehyde for 1–2 h and equilibrated overnight in PBS with 20% sucrose and 0.01% sodium azide. Serial coronal sections (30 µm) were cut with a freezing-stage microtome, collected sequentially in six sets, and stored in PBS with 0.01% sodium azide at 4°C.

Sections containing the PH and the rRPa were preincubated in an antibody dilution solution (ADS: 500 mL of PBS, 0.3% Triton X-100, 1.25 g of carrageenan, 100 mg of NaN₃, 5 mL of normal donkey serum) for 2 h and incubated overnight at room temperature with the primary antibodies for TH (1:1,000, rabbit-anti-TH, Millipore, AB152, RRID AB_390204), CTb (1:10,000, goat-anti-CTb, List Biological Laboratories, No. 703), D2R (rabbit-anti-D2R, 1:200, Millipore, AB5084P, RRID AB_2094980), TPH (mouse-anti-TPH, 1:500, Sigma, T0678, RRID AB_261587), and/or c-fos (rabbit-anti-c-Fos, 1:10k, Calbiochem, PC38, RRID AB_2106755). After three washes in PBS containing 0.3% Triton X-100 (TPBS, 20 mM), the tissue was incubated for 1 h in ADS containing the proper secondary antibody (donkey-anti-goat, 1:500, Invitrogen, A11055, RRID AB_2534102; donkey-anti-rabbit, 1:500, Invitrogen; A21206, RRID AB_2535792; donkey-anti-mouse, 1:500, Invitrogen, A21202, RRID AB_141607), or for BDA labeling, the tissue was incubated in streptavidin-Alexa 594 (1:500, Invitrogen, S32356) and then rinsed with PBS. After fluorescent labeling, the tissue was washed in PBS and mounted onto coated slides and coverslipped with antifade mounting medium (Pro-Long Gold, Invitrogen).

Tissue Analysis

The neuroanatomical designations of structures in histological sections are based on the stereotaxic rat brain atlas of Paxinos and Watson (25). Photomicrographs of brain sections and labeled neurons were taken with a camera attached to an Olympus BX51 fluorescence microscope. The photomicrographs were assembled into a plate using Adobe Photoshop to adjust contrast and brightness without altering the original colors.

Surgical Procedures

The rats were anesthetized with isoflurane (3% in 100% O₂) and transitioned to intravenous urethane (0.75 g/kg) and α-chloralose (60 mg/kg) after cannulation of the femoral artery to record mean arterial pressure (MAP) and heart rate (HR) and the femoral vein for drug delivery. The trachea was cannulated for artificial ventilation with a tidal volume of ∼1 mL/100 g body wt, 60 cycles/min with 100% O₂ after paralysis with d-tubocurarine (0.6 mg iv initial dose, supplemented with 0.3 mg/h) to maintain the resting end-expired CO₂ (Exp. CO₂) at 3.5%– 4.5% (Capnogard; Novametrix Medical Systems, Wallingford, CT). To control skin temperature, the shaved trunk was wrapped with a water-perfused blanket. To monitor the body temperature, thermocouples (Physitemp Instruments, Clifton, NJ) providing the signal to a TC-2000 receiver (Sable Systems, Las Vegas, NV) were placed on the abdominal skin temperature (T_SKIN), 5 cm into the rectum [core body temperature (T_CORE)], and within the left interscapular BAT [BAT temperature, (T_BAT)]. The rats were placed in a stereotaxic frame with the incisor bar positioned −4 mm below the interaural line and a spinal clamp on the T8 vertebra.

Postganglionic BAT sympathetic nerve activity (SNA) was recorded under mineral oil with a bipolar hook electrode from the central cut end of a nerve bundle isolated from the ventral surface of the right interscapular BAT pad. Complete inhibition of BAT SNA by treatment with the ganglionic blocker, hexamethonium (30 mg/kg), verified the postganglionic autonomic nature of the recorded nerves in some experiments. Nerve activity was filtered (1–300 Hz) and amplified (10,000–50,000×) with a CyberAmp 380 (Axon Instruments, UnionCity, CA). To obtain a continuous measure (4-s bins) of BAT SNA amplitude, the root mean square (rms) amplitude of the BAT SNA was calculated (Spike 2, CED) 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 SNA values are expressed as percentage of baseline (BL), where the BL value was taken as the mean BAT SNA amplitude during a 30-s period of minimum BAT SNA recorded when the rat was in a warm condition and basal BAT SNA was absent. All physiological variables were digitized [Micro 1401 MKII; Cambridge Electronic Design (CED), Cambridge, UK] and recorded to a hard drive for subsequent analysis (Spike 2, CED).

A burr hole was drilled in the interparietal bone for nanoinjection of drugs into the vicinity of the rRPa (coordinates: 2.8– 3.0 mm caudal to lambda, 0.0 mm lateral to the midline, −9.2 to 9.8 mm below the dural surface). Burr holes were also drilled in the right and left parietal bones for bilateral nanoinjection of drugs into the PH (coordinates: −5.5 to −6.0 mm caudal to the bregma, 0.3 mm lateral to the midline, −7.5 to −8.5 mm below the dural surface).

Nanoinjections

The nanoinjection pipettes (20–40 µm tip diameter) were stereotaxically positioned into the PH and the rRPa for the nanoinjection of drugs (60 nL, estimated using a calibrated microscope reticule to observe the displacement of the fluid meniscus in the glass pipette); this was accomplished with a pressure injection apparatus (Toohey, Fairfield, NJ).

All drugs for nanoinjections were obtained from Sigma-Aldrich (St. Louis, MO) and diluted in sterile phosphate-buffered saline: N-methyl-d-aspartate (NMDA) at 0.2 mM, (±)-7-Hydroxy-DPAT hydrobromide (D2/D3 dopamine receptor agonist) at 8 mM (the minimum dose to effectively inhibit BAT SNA in preliminary experiments), SB-277011-A (dopamine receptor antagonist) at 1 mM (the lowest concentration capable of increasing BAT SNA in preliminary experiments), gabazine (selectively antagonizes GABA-induced Cl⁻ currents) at 1 mM, bicuculline methiodide (Bic) (selective antagonist at GABA_A receptors) at 500 µM, and saclofen (selective antagonist at GABA_B receptors) at 25 mM (22). Nanoinjection of the vehicle into the PH and the rRPa had no effects on either BAT thermogenic or cardiovascular variables.

Fluorescent polystyrene microspheres (FluoSpheres, F8797, F8801, or F8803; Molecular Probes, Eugene, OR) were included in the injectate (1:100 dilution of FluoSpheres in the injectate) to show the localization of the nanoinjections. After the physiological recordings, rats were perfused (5% paraformaldehyde) transcardially, and their brains were removed, postfixed (2–12 h) in 5% formaldehyde for (2–12 h), and sectioned on a microtome (60-µm coronal sections). The sections were mounted on slides, and nanoinjection sites were localized and photographed.

Statistical Analyses

The control values of the variables were the mean values during the 120 s before the injection. Pretreatment control conditions were either a warm T_SKIN (36.2–38.5°C) and T_CORE (35.8–37.6°C), which resulted in low levels of pretreatment BAT SNA, or a cool T_SKIN (34.5 ± 0.4°C) and T_CORE (35.8 ± 0.4°C), which resulted in a modestly elevated level of pretreatment BAT SNA. The amplitudes of evoked responses were calculated from the mean levels of the variables during the 30-s period of the maximum response (either increase or decrease) occurring within 5 min of the treatment. For the rats that received two serial injections of bicuculline in the rRPa followed by an injection of saline or 7-hydroxy-N,N-di-n-propyl-2-aminotetrali (7-OH-DPAT) in the rRPa, the data from BAT SNA were analyzed by calculating the area under curve (AUC) from the time of injection of bicuculline in the rRPa to 25 min after the injection of bicuculline in the rRPa. The AUC data were transformed to the logarithmic scale. Data from the other variables were analyzed by comparing the changes from baseline at 15 min after the injection of bicuculline (the time of plateau for most of the variables analyzed) in the saline versus 7-OH-DPAT trials. Group data are given as means ± SE and compared with paired Student’s t tests using statistical software (GraphPad Prism or SPSS), with P < 0.05 considered significant.

RESULTS

Activation of D2/D3R in the rRPa Inhibits BAT Thermogenesis Evoked by Skin Cooling or by Activation of NMDA Receptors in the rRPa

The rRPa, a major thermoregulatory area (26), receives a dense dopaminergic input (12). To determine if D2R/D3R in the rRPa play a role in regulating BAT SNA and BAT thermogenesis, we injected the D2/D3R agonist 7-OH-DPAT into the rRPa during cold-evoked activation of BAT. Nanoinjection of 7-OH-DPAT (Fig. 1A) in the rRPa produced a marked inhibition of cold-evoked BAT SNA (−697.8 ± 341.0% BL; P = 0.01), T_BAT (−1.6 ± 0.6°C; P = 0.005), Exp. CO₂ (−0.8 ± 0.2%; P = 0.0008), and MAP (−17.2 ± 11.8 mmHg; P = 0.03). No effect was observed on T_CORE (P = 0.1) or HR (P = 0.1) (Fig. 1A).

Figure 1.

Nanoinjection of 7-hydroxy-N,N-di-n-propyl-2-aminotetralin (7-OH-DPAT) into the rostral raphe pallidus area (rRPa) attenuates the increase in brown adipose tissue (BAT) sympathetic nerve activity (SNA), BAT temperature (T_BAT), expired CO₂ (Exp. CO₂), and mean arterial pressure (MAP) evoked by cold exposure or nanoinjection of N-methyl-d-aspartate (NMDA) in the rRPa. A: representative example of the effect of 7-OH-DPAT injection into the rRPa (vertical line) on the recorded parameters during cold exposure. B: whisker dot plots for the physiological variables before (pre-7-OH-DPAT) and after (post-7-OH-DPAT) rRPa nanoinjection (n = 5). *P < 0.05. C: locations of the nanoinjection sites represented by open squares on a partial coronal section schematic drawing at approximately −11.8 mm caudal to the bregma and representative photomicrograph showing the fluorescent beads marking the injection site. D: representative example of the effect of 7-OH-DPAT injection into the rRPa (vertical line) on the recorded parameters during responses evoked by NMDA injection in the rRPa. E: plots of individual paired data (n = 5 rats) for the physiological variables between pre-NMDA and post-NMDA nanoinjections in saline or 7-OH-DPAT pretreated trials. Data shown as means ± SE. *P < 0.05, paired t test. F: locations of the nanoinjection sites represented by open circles (7-OH-DPAT) and asterisks (NMDA) on a partial coronal section schematic drawing and representative photomicrograph showing the fluorescent beads marking the injection sites. BL, baseline; bpm, beats per minute; HR, heart rate; Py, pyramidal tract; T_CORE, core temperature.

We also found that activation of D2R/D3R in the rRPa inhibits the increases in BAT SNA and BAT thermogenesis evoked by activation of NMDA receptors in the rRPa. Nanoinjection of 7-OH-DPAT into the rRPa reduced the peak BAT SNA and BAT temperature responses to nanoinjections of NMDA in the rRPa by −487.4 ± 150.3% BL (P = 0.01) and −0.6 ± 0.2°C (P = 0.03), respectively. The Exp. CO₂ (P = 0.8), T_CORE (P = 0.8), MAP (P = 0.2), and HR (P = 0.2) were not affected by nanoinjection of 7-OH-DPAT in the rRPa (Fig. 1B).

The Antagonism of D2R/D3R in the rRPa Elicits BAT Thermogenesis

Having established that activation of D2R/D3R in the rRPa inhibits BAT activity, likely by reducing the discharge of the BAT SPmN located in the rRPa, it was important to determine whether the endogenous, BAT sympathoinhibitory dopaminergic input to the rRPa has an ongoing level of activity. Nanoinjections of the D2/D3R antagonist, SB-277011A, in the rRPa evoked a marked and rapid increase in BAT SNA (+400 ± 201% BL; P = 0.006), T_BAT (+0.6 ± 0.7°C; P = 0.049), and Exp. CO₂ (+0.6 ± 0.7%; P = 0.02). No significant changes were observed in T_CORE (P = 0.4), MAP (P = 0.3), or HR (P = 0.05) (Fig. 2). These data indicate that, under the conditions of our experiment, the dopaminergic neurons that project into the rRPa are tonically active and their release of DA in the rRPa results in a reduced activity of BAT SPmNs in the rRPa.

Figure 2.

Nanoinjection of a dopamine D2 receptor antagonist, SB-277011A, into the rostral raphe pallidus area (rRPa) increased brown adipose tissue (BAT) sympathetic nerve activity (SNA), BAT temperature (T_BAT), and expired CO₂ (Exp. CO₂). A: representative example of SB-277011A injection into the rRPa (vertical dashed line) on the recorded parameters. B: whisker dot plots for the physiological variables before (pre-SB-277011A) and after (post-SB-277011A) rRPa nanoinjection (n = 5 rats). Data shown as means ± SE. *P < 0.05, paired t test. C: locations of the nanoinjection sites represented by asterisks (*) on a partial coronal section schematic drawing at approximately −11.8 mm caudal to the bregma and representative photomicrograph showing the fluorescent beads marking the injection site. BL, baseline; bpm, beats per minute; HR, heart rate; MAP, mean arterial pressure; Py, pyramidal tract; T_CORE, core temperature.

The D2R Is Expressed in Serotonergic Neurons and in Cold-Activated Neurons in the rRPa

To provide anatomical evidence on whether serotonergic neurons are among the rRPa neurons that are sensitive to a DA input to the rRPa, we performed immunohistochemical labeling for the D2R (Fig. 3A) and TPH (Fig. 3B) in brainstem sections through the rRPa. The rRPa contained neurons that were double labeled for the D2R and TPH and neurons that were labeled with only the D2R (Fig. 3C). These results suggest that a D2R/D3R agonist applied to the rRPa would act on D2Rs on both serotonergic and nonserotonergic neurons in the rRPa.

Figure 3.

Neurons in the rostral ventromedial medulla, including serotonergic neurons in the rostral raphe pallidus area (rRPa), contain D2 dopamine receptors. A–C: photomicrographs showing a coronal section through the rRPa immunohistochemically labeled for D2 dopamine receptors (D2R, red, A) and tryptophan hydroxylase (TPH, green, B). C is a merged file of A and B (arrowheads indicate double-labeled neurons). D–F: photomicrographs showing coronal sections through the rRPa immunohistochemically labeled for D2R (red, D) and c-Fos (green, E). C illustrates double-labeled neurons (arrowheads). Py, pyramidal tract.

We also performed immunohistochemical labeling for c-fos and the D2R in histological sections through the rRPa of rats exposed to a cold ambient temperature. Our finding that the rRPa contains neurons double labeled for the D2R and c-fos (Fig. 3F) indicates that some of the rRPa neurons that respond to local injection of D2R/D3R agonists and/or antagonists are among those that are activated by cold exposure. Since BAT SPmNs are activated by cold exposure, it is highly likely that some BAT SPmNs express the D2R.

Inhibition of BAT Thermogenesis by a D2R/D3R Agonist Requires Activation of GABA_A Receptors, but not GABA_B Receptors, in the rRPa

To determine whether the activation of GABA receptors in the rRPa is required for the sympathoinhibitory effect of activating the D2R in the rRPa, we pretreated the rRPa with either a GABA_A or a GABA_B antagonist before nanoinjecting 7-OH-DPAT into the rRPa. As previously shown (18, 22), nanoinjections of bicuculline in the rRPa increased BAT SNA, T_BAT, Exp. CO₂, T_CORE, HR, and MAP (Fig. 4A). A within-subject randomized trial design was used to compare two serial responses evoked by nanoinjections of bicuculline in the rRPa followed by a nanoinjection of saline or 7-OH-DPAT in the rRPa. Nanoinjection of 7-OH-DPAT attenuated the bicuculline-evoked increases in BAT SNA (P = 0.02), T_BAT (P = 0.046), Exp. CO₂ (P = 0.03), and HR (P = 0.03) in comparison with control saline trials (Fig. 4, A and B), whereas the changes in T_CORE (P = 0.17) and MAP (P = 0.14) were not different between the trials (Fig. 4, A and B). These results indicate that activation of GABA_A receptor in the rRPa is necessary for the complete inhibition of BAT SNA and BAT thermogenesis following nanoinjection of 7-OH-DPAT in the rRPa.

Figure 4.

Nanoinjection of 7-hydroxy-N,N-di-n-propyl-2-aminotetralin (7-OH-DPAT) into the rostral raphe pallidus area (rRPa) attenuates the increase in brown adipose tissue (BAT) sympathetic nerve activity (SNA), BAT temperature (T_BAT), expired CO₂ (Exp. CO₂), and heart rate (HR) evoked by nanoinjection of bicuculline (Bic) in the rRPa. A: representative example of the effect of 7-OH-DPAT injection into the rRPa on the recorded parameters during Bic-evoked responses. B: group data represented by means ± SE and plots of paired individual data from animals that received Bic and saline (filled black circles) or Bic and 7-OH-DPAT (open circles) in the rRPa. For BAT SNA, the area under curve of the log transformed data was quantified from the time of injection of Bic until 25 min after the injection. The quantifications of the changes in the measured variables other than BAT SNA were calculated at the 15 min time point after Bic injection (a time point at which the variables had largely plateaued). *P < 0.05, paired t test. C: locations of the nanoinjection sites represented by open circles on a partial coronal section schematic drawing at approximately −11.8 mm caudal to the bregma and representative photomicrograph showing the fluorescent beads marking the injection site (n = 5). BL, baseline; bpm, beats per minute; MAP, mean arterial pressure; Py, pyramidal tract; 7-OH-DPAT, 7-hydroxy-N,N-di-n-propyl-2-aminotetralin; T_CORE, core temperature.

Nanoinjection of the selective GABA_B receptor (GABA_BR) antagonist, saclofen, in the rRPa also increased BAT SNA, T_BAT, Exp. CO₂, T_CORE, HR, and MAP (Fig. 5), as previously shown (22). Nanoinjection of 7-OH-DPAT reversed the saclofen in rRPa-evoked rises in BAT SNA (−698.0 ± 388% BL; P = 0.008), T_BAT (−1.2 ± 0.7°C; P = 0.009), Exp. CO₂ (−0.9 ± 0.4%; P = 0.004), MAP (−11 ± 4 mmHg; P = 0.002), and HR (−48 ± 24 beats/min; P = 0.005) (Fig. 5). Nanoinjection of 7-OH-DPAT in the rRPa did not affect the saclofen-induced increase in T_CORE (P = 0.5). These data suggest that although there is a tonic BAT sympathoinhibitory effect due to an ongoing activation of GABA_BR in the rRPa under the conditions of our experiment (22), activation of GABA_BR is not required for the BAT sympathoinhibitory effect of 7-OH-DPAT in the rRPa.

Figure 5.

Activation of dopamine receptors in the rostral raphe pallidus area (rRPa) reverses the increases in brown adipose tissue (BAT) sympathetic nerve activity (SNA), BAT temperature (T_BAT), expired CO₂ (Exp. CO₂), mean arterial pressure (MAP), and heart rate (HR) evoked by blockade of GABA_B receptors in the rRPa. A: nanoinjection of saclofen in the rRPa increased BAT SNA, T_BAT, Exp. CO₂, and HR, subsequent nanoinjection of 7-OH-DPAT in the same site rapidly reversed these responses and decreased MAP. B: whisker dot plots for the physiological variables before and after the nanoinjection of 7-OH-DPAT (n = 5 rats). Data shown as means ± SE. *P < 0.05, paired t test. C: locations of the nanoinjection sites represented by squares (7-OH-DPAT) and asterisks (saclofen) on a partial coronal section schematic drawing and representative photomicrograph showing the fluorescent beads marking the injection sites. BL, baseline; bpm, beats per minute; HR, heart rate; MAP, mean arterial pressure; Py, pyramidal tract; 7-OH-DPAT, 7-hydroxy-N,N-di-n-propyl-2-aminotetralin; T_CORE, core temperature.

The rRPa Receives Dopaminergic and Nondopaminergic Projections from the PH

To identify the source of the dopaminergic input to the rRPa, we nanoinjected the retrograde tracer, CTb, in the rRPa and examined sections throughout the brain for neurons that were double labeled for CTb and TH. The PH was the only brain site containing DA neurons labeled for both CTb and TH (Fig. 6, A and C). Double-labeled neurons in the PH were distributed bilaterally between the left and right fasciculus retroflexus at rostral-caudal levels between −4.7 and −5.0 mm caudal to the bregma (Figs. 6, B and D). About 32% of the CTb-positive neurons found in this region of the PH were double labeled for CTb and TH (Fig. 6D).

Figure 6.

Dopaminergic and nondopaminergic neurons in the posterior hypothalamus project to the rostral raphe pallidus area (rRPa). A: schematic drawing of a coronal section through the level of the posterior hypothalamus at which rRPa-projecting dopaminergic neurons are found. B: photomicrograph of partial coronal section through the posterior hypothalamus illustrating cholera toxin subunit b (CTb)-immunoreactive (ir) retrogradely labeled neurons (red) following nanoinjection of CTb in the rRPa and tyrosine hydroxylase (TH)-ir neurons (green). C: higher magnification of the area denoted by the white square in B. D: scatter plots of CTb-ir and TH-ir neurons at three levels of the posterior hypothalamus from three rats that had received CTb injections in the rRPa (n = 3). Photomicrographs of partial coronal sections through the rRPa immunohistochemically labeled for biotinylated dextran amine (BDA, green, E) following BDA injection in the posterior hypothalamus, dopamine D2 receptor (D2R, red, F). G: a higher magnification of the merged photomicrographs (BDA-ir, green and D2R-ir, red) representing the area denoted by the dashed white box in E and F. fr, fasciculus retroflexus; Py, pyramidal tract.

To confirm the anatomical connection between TH-positive neurons in the PH and the rRPa, we injected the anterograde tracer, BDA, into the PH and explored the rRPa for close appositions of BDA-immunoreactive (ir) terminals and D2R-ir neurons. BDA-positive fibers and terminal boutons were found in the rRPa, and many were closely apposed to the processes of D2R-ir neurons in the rRPa (Fig. 6E).

Disinhibition of the PH Inhibits BAT Thermogenesis and Is Dependent on the Activation of D2/D3R in the rRPa

We determined the effect on cold-evoked BAT thermogenic and cardiovascular variables of activating neurons in the PH region containing the dopaminergic neurons that project to the rRPa. Nanoinjection of gabazine, a selective, competitive GABA_A receptor antagonist, in the PH reduced the skin cooling-evoked level of BAT SNA (−739 ± 311% BL; P = 0.049), T_BAT (−1.0 ± 0.1°C; P = 0.003), Exp. CO₂ (−0.3 ± 0.1%; P = 0.035), and HR (−43 ± 12 beats/min; P = 0.017) (Fig. 7). Following recovery from the first gabazine nanoinjection, a second nanoinjection of gabazine in the same site produced similar changes that did not differ from the first trial: BAT SNA (−789 ± 371% BL; P = 0.52 compared with the first trial), T_BAT (−1.0 ± 0.1°C; P = 0.64 compared with the first trial), Exp. CO₂ (−0.3 ± 0.1%; P = 0.39 compared with the first trial), and HR (−51 ± 18 beats/min; P = 0.67 compared with the first trial) (Fig. 7). Similarly, in a separate group of rats, nanoinjection of gabazine in the PH reduced the skin cooling-evoked level of BAT SNA (−373.3 ± 62% BL; P = 0.009), T_BAT (−0.5 ± 0.1°C; P = 0.006) Exp. CO₂ (−0.4 ± 0.1%; P = 0.02), and HR (−28 ± 8 beats/min; P = 0.04) (Fig. 8). Subsequent nanoinjection of SB-277011 in the rRPa prevented the second PH gabazine-evoked reductions in cold-evoked BAT SNA (+57 ± 31% BL; P = 0.17), T_BAT (+0.3 ± 0.1°C; P = 0.02), Exp. CO₂ (+0.2 ± 0.1%; P = 0.07), and HR (0 ± 9 beats/min; P = 0.5) (Fig. 8), such that the responses to the second (post-SB-27011) PH gabazine trial differed significantly from the first PH gabazine trial: BAT SNA (P = 0.009), T_BAT (P = 0.002), Exp. CO₂ (P = 0.008), and HR (P = 0.018) (Fig. 8B).

Figure 7.

Nanoinjection of gabazine into the posterior hypothalamus (PH) decreased brown adipose tissue (BAT) sympathetic nerve activity (SNA), BAT temperature (T_BAT), expired CO₂ (Exp. CO₂), core temperature (T_CORE), and heart rate (HR). A: representative example of the effect of gabazine injection into PH (dashed vertical lines) on the recorded parameters during cold exposure. Repeated injections of gabazine into the same PH site produced remarkably consistent responses. B: plots of paired individual data for the responses of the physiological variables to injection of gabazine in the PH for each of the two repeated trials (n = 4 rats). *P < 0.05, paired t test. C: locations of the nanoinjection sites in PH (open circles, left) at −4.5 mm and −5.0 mm caudal to the bregma and representative photomicrographs showing the fluorescent beads marking the injection sites. BL, baseline; bpm, beats per minute; fr, fasciculus retroflexus; MAP, mean arterial pressure; mp, mamillary peduncle; PH, posterior hypothalamus; 3 V, third ventricle; SN, substantia nigra.

Figure 8.

The decrease in brown adipose tissue (BAT) sympathetic nerve activity (SNA), BAT temperature (T_BAT), expired CO₂ (Exp. CO₂), and heart rate (HR) evoked by nanoinjection of gabazine into the posterior hypothalamus (PH) is prevented by nanoinjection of the dopamine receptor antagonist, SB-277011A, in the rostral raphe pallidus area (rRPa). A: representative example of the effect of gabazine injection into PH (solid vertical lines) before and after injection of SB-277011A in the rRPa (dashed vertical line) on the recorded parameters during cold exposure. B: plots of paired individual data for the responses of the physiological variables to injection of gabazine in the PH before and after SB-277011A injection in the rRPa (n = 4 rats). *P < 0.05, paired t test. C: locations of the nanoinjection sites in the PH (open circles, left) at −4.5 mm and −5.0 mm caudal to the bregma and in the rRPa (open squares, right) on partial coronal section schematic drawing at approximately −11.8 mm caudal to the bregma (rRPa) and representative photomicrographs showing the fluorescent beads marking the injection sites. BL, baseline; bpm, beats per minute; fr, fasciculus retroflexus; MAP, mean arterial pressure; mp, mamillary peduncle; Py, pyramidal tract; 7-OH-DPAT, 7-hydroxy-N,N-di-n-propyl-2-aminotetralin; T_CORE, core temperature; 3 V, third ventricle; SN, substantia nigra.

DISCUSSION

DA agonists elicit a hypothermic effect in humans and in rodents (6, 7, 27, 28). Systemic administration of DA agonists reduces core body temperature by inhibiting metabolic thermogenesis from muscle shivering (7) and from BAT (6), rather than by increasing heat loss (27). In the present study, we identified the rRPa as a brain site at which DA receptor agonists can cause hypothermia. The rRPa, which contains the premotor neurons for shivering and for BAT thermogenesis, is a critical node in the central thermoregulatory circuitry that controls the levels of sympathetic and somatic motor nerve activities for BAT and shivering thermogenesis, respectively. We found that local nanoinjection of the D2/D3R agonist, 7-OH-DPAT, in the rRPa caused a significant reduction in BAT SNA and BAT thermogenesis, indicating that there are BAT thermogensis-promoting neurons within the rRPa, potentially the BAT SPmNs, that express the D2R or the D3R. Our finding of D2R-ir in serotonergic and cold-activated neurons in the rRPa indicates that the effects of the DA input to the rRPa are likely to be mediated by the D2R. The coupling of the D2R to Gi protein, which results in neuronal inhibition (29), is consistent with 7-OH-DPAT eliciting hypothermia through a D2R-mediated inhibition of the BAT SPmNs in the rRPa. In our study, we identified TH-ir neurons in the PH as the single source of DA input to the rRPa. Since nanoinjection of a D2/D3R antagonist in the rRPa induced an increase in BAT thermogenesis, these dopaminergic neurons in the PH are tonically active under the conditions of our experiment and exert an ongoing restraint on the level of BAT thermogenesis and BAT energy expenditure. Overall, we have identified dopaminergic neurons in the PH as a novel source of ongoing BAT sympathoinhibition through their release of DA onto the D2R in the rRPa. These data complement our previous descriptions of BAT sympathoinhibitory inputs to the rRPa, including those mediated by GABA_A receptors (18), GABA_B receptors (Fig. 4; see Ref. 22), glycine_A receptors (21), and muscarinic cholinergic receptors (22).

Our finding that 7-OH-DPAT in the rRPa inhibits the BAT thermogenesis induced by NMDA in the rRPa is consistent with a direct inhibitory effect of the D2R on BAT SPmNs in the rRPa, rather than by a decrease in glutamate release via D3R autoreceptors activations (30, 31) on glutamatergic terminals in the rRPa. This conclusion is further supported by the coupling of the D2R to Gi protein, which results in neuronal inhibition (29) and could counter the depolarization of SPmNs by NMDA.

Nanoinjection of bicuculline, a GABA_AR antagonist, in the rRPa increases BAT SNA and BAT thermogenesis (18), and this activation is reversed by blockade of excitatory amino acid receptors (32), indicating that the activity of BAT SPmNs in the rRPa represents a balance of the levels of their tonically active GABAergic inputs that restrain the activity of BAT SPmNs and their tonically active excitatory inputs that drive BAT SPmN activity. We found that blockade of D2R/D3R in the rRPa with SB-277011A led to an increase in BAT SNA, consistent with an ongoing release of DA onto inhibitory D2R in the rRPa. We observed that nanoinjection of 7-OH-DPAT in the rRPa only partially inhibited the BAT thermogenesis elicited by the blockade of GABA_AR in the rRPa, indicating that activation of GABA_AR is required for DA in the rRPa to completely block BAT thermogenesis. This result suggests that without the hyperpolarizing effect of the tonical activation of GABA_AR in the rRPa, activation of the D2R in the rRPa, and the consequent stimulation of Gi protein in D2R-expressing neurons, is not sufficient to completely suppress the activity of BAT SPmNs resulting from their tonically active excitatory inputs. Since the D2R is an inhibitory receptor, it is unlikely that its inhibitory actions on BAT SNA result from the D2R expressed on local GABAergic interneurons in the rRPa; a conclusion supported by our finding that D2Rs are expressed by TPH-ir neurons in the rRPa. It would be of interest to determine whether BAT SPmNs express both D2Rs and GABA_ARs.

Presynaptic GABA_BRs are metabotropic receptors that inhibit the release of neurotransmitters, whereas postsynaptic GABA_BRs generate slow inhibitory postsynaptic potentials (reviewed in Ref. 33). The blockade of GABA_BRs with nanoinjection of saclofen in the rRPa induced an immediate activation of BAT SNA and BAT thermogenesis (22), which was reversed by activation of the D2R in the rRPa. The activation of BAT SNA following saclofen in the rRPa indicates that, under the conditions of our experiments, there is a tonic release of GABA onto GABA_BR expressed by rRPa neurons within the excitatory pathway controlling the activity of BAT SPmNs in the rRPa. Such GABA_BR activation could inhibit BAT SPmN activity by either, or both, presynaptically reducing glutamate release or postsynaptically hyperpolarizing BAT sympathoexcitatory neurons. The effectiveness of 7-OH-DPAT in the rRPa in reversing the BAT sympathoexcitatory effect of saclofen suggests that, in contrast to the tonic release of GABA onto GABA_AR in the rRPa, the tonic release of GABA onto GABA_BR produces only a small inhibitory effect that can be replaced by the D2R-mediated activation of Gi protein in rRPa neurons within the excitatory pathway controlling the activity of BAT SPmNs.

Previous anatomical studies have shown anterogradely labeled fibers projecting from the caudal PH to the rRPa (34, 35). We expanded these data to show that there is a population of TH-expressing neurons within the population of PH neurons that project to the rRPa. This population of DA cells projecting to the rRPa also extends caudally beyond the posterior hypothalamus and into the rostral midbrain. This subset of DA neurons is generally found ventral and medial to the fasciculus retroflexus and is notable for having small perikarya (Fig. 5, A–C), a typical characteristic of TH-ir neurons in this region (35, 36) with TH-ir also present in the nuclei, similar to a previous description in ventral tegmental area neurons (37). Neurons in the PH have been implicated in pulmonary hypertension (38) and in arousal (39), but the physiological and/or thermoregulatory role of the PH input to the rRPa, including the DA neurons, remains to be identified. DA neurons in the substantia nigra express transient receptor potential vanilloid-type 3 (TRPV3) and type 4 (TRPV4), respectively, and their discharge is sensitive to local temperature (40). Although TRPV3 gene expression in the PH is increased following menthol application on the skin (41), the phenotype and projections of these neurons remain unknown.

Previous studies using transections of the neuraxis have demonstrated a prepontine descending tonically active pathway that suppresses BAT SNA (42) and BAT thermogenesis (43, 44). The locations of the neuronal populations in these previous studies suggest that they are distinct from the dopaminergic pathway described in the current study. Nonetheless, given that blockade of dopamine receptors in the rRPa increased BAT SNA and BAT thermogenesis, elimination of this descending dopaminergic pathway likely would have contributed to the increase in BAT SNA and BAT thermogenesis following the transections in previous studies.

Blockade of GABA_AR in the PH inhibited the BAT SNA and BAT thermogenesis evoked by skin cooling. Since subsequent nanoinjection of SB-277011A in the rRPa reversed this inhibition and prevented an inhibition of BAT SNA in response to a further nanoinjection of gabazine in the PH, we conclude that at least a portion of the inhibition of BAT SNA in response to disinhibition of PH neurons is mediated by the population of DA neurons in the PH that projects to the rRPa. The source(s) of the active GABAergic inhibition of the BAT sympathoinhibitory neurons in the PH area remains to be determined. The disinhibitory activation of PH neurons with local nanoinjections of gabazine does not selectively target the DA neurons projecting to the rRPa, and two-thirds of the PH neurons that project to the rRPa were not TH-ir. Our experiments did not determine whether these nonDA, PH neurons projecting to the region of the rRPa have a role in thermoregulation or whether the PH DA neurons projecting to the rRPa have collateral projections that are of relevance to thermoregulation.

Perspectives and Significance

In summary, our results are consistent with tonically active DA neurons in the PH that project to D2R-expressing BAT SPmNs in the rRPa to contribute, through activation of an intracellular Gi protein pathway, to the ongoing inhibitory regulation of the activity of BAT SPmNs and in turn, via spinal BAT sympathetic preganglionic neurons and BAT sympathetic ganglion cells, to the control of BAT thermogenesis and BAT energy expenditure. The neural circuitry described in these studies may also affect other thermoregulatory functions, such as cutaneous vasoconstriction and shivering, that are also mediated by neurons in the rRPa. This is particularly intriguing since neurons in the rRPa contribute to the rhythmic firing of motor circuits involved in shivering thermogenesis (17). Removal of the DA input to the rRPa from neurons in the PH could lower the thermal threshold for shivering and thereby bias the spinal motoneuron circuits such that they are more suseptible to the descending tremor-driving inputs activated in Parkinson’s disease. The dopaminergic drive from the PH to SPmNs in the rRPa is an exciting and novel pathway that could play a significant role in the induction of hypothermic states and other physiological and pathophysiological states. Further exploration of the physiological circumstances that modulate the PH-rRPa DA axis and of the modulatory inputs to the PH neurons projecting to the rRPa will yield an improved understanding of the functional role of DA in neural circuits involved in thermoregulation, metabolism, and, perhaps, motor function.

GRANTS

This work was supported by the National Institutes of Health Grant No. R01 DK112198 (to C.J.M.).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

E.P.S.d.S.F, S.F.M., and C.J.M. conceived and designed research; E.P.S.d.S.F, C.M.D.M., E.V., and C.J.M. performed experiments; E.P.S.d.S.F., C.M.D.M., E.V., and C.J.M. analyzed data; E.P.S.d.S., C.M.D.M., S.F.M., and C.J.M. interpreted results of experiments; E.P.S.d.S.F., C.M.D.M., and C.J.M. prepared figures; E.P.S.d.S.F and C.M.D.M. drafted manuscript; E.P.S.d.S.F., C.M.D.M., S.F.M., and C.J.M. edited and revised manuscript; E.P.S.d.S.F., C.M.D.M., S.F.M., and C.J.M. approved final version of manuscript.