Introduction

Homeotherms, including mammals, maintain core body temperature (CBT) within a narrow range, an essential requirement for survival. The preoptic area of the hypothalamus plays an important role in CBT regulation by integrating peripheral thermal information and by sending efferent signals that control thermal effector organs (1–3). Via projections to the dorsomedial hypothalamus and rostral raphe pallidus, preoptic neurons control thermogenesis as well as heat loss mechanisms (4). The preoptic area also orchestrates the fever response, an important mechanism for fighting infection or inflammatory disease (1, 5). Recent studies have identified populations of preoptic neurons that integrate peripheral thermal information and control the core body temperature (CBT) and play an important role in the fever response to endotoxin (6–9).

Neurotensin (Nts) is a 13 aminoacid peptide found in the central nervous system (CNS) as well as in the gastrointestinal tract (10). Nts-producing neurons and their projections are widely distributed in the brain, which may explain the wide variety of effects of this peptide (11). Pharmacological approaches have revealed a role of Nts in analgesia, arousal, blood pressure, feeding, reward, sleep, the stress response and thermoregulation (12–15). Recently, studies have started to unravel the role played by specific populations of neurotensinergic neurons in feeding (16, 17) and drinking behaviors (18) as well as in sleep (19, 20) and appetitive and reinforcing aspects of motivated behaviors (21).

When infused intracerebroventricularly or in the preoptic area Nts induces a 4-5 °C hypothermia (22–24). The effect was attributed to activation of heat loss mechanisms (25). The medial preoptic area (MPO) expresses high densities of both neurotensinergic neurons and neurotensin receptors (11, 26, 27). A previous study has identified a population of MPO Nts neurons that express VGAT and modulate sexual behavior via projections to the ventral tegmental area (21). The present study has investigated the cellular characteristics of MPO Nts neurons, their influence on the activity of nearby MPO neurons and on CBT as well as the effect of VGAT deletion on these actions.

Results

MPO^Nts;hChR2 neurons are GABAergic and their optogenetic stimulation decreases the firing rate of MPO^PACAP neurons by increasing the frequency of IPSCs

Nts-cre mice received bilateral MPO injections of AAV5-EF1a-double floxed-hChR2(H134R)-EYFP-WPRE-HGHpA (see Methods) to express the excitatory opsin ChR2-EYFP in MPO Nts neurons. We refer to these mice as MPO^Nts;hChR2. Patch-clamp recordings revealed that MPO^Nts;hChR2 neurons were spontaneously active and the majority of them (16 out of 22 neurons studied) fired mostly doublets or triplets of action potentials (Fig 1A) while the rest fired single action potentials. The average firing rate at 36°C was 4.29±2.45 Hz (n=22). Intrinsic properties of PO/AH neurons were investigated using injections of square current pulses delivered in whole-cell configuration. The average capacitance of the neurons tested was 17.86±2.48 pF (n=22). All neurons tested displayed a low threshold spike (LTS) upon the end of a hyperpolarizing current injection (Fig 1B). Injection of positive currents generated bursts of doublets or triplets of action potentials (Fig 1B). To verify the presence of Nts transcripts in MPO^Nts;hChR2 we have carried out scRT/PCR in these neurons. We could detect Nts transcripts in 7 out of 10 MPO^Nts;hChR2 studied (Fig 1C). We have also tested the expression of the neuronal markers VGAT(Slc32a1) and Vglut2. VGAT was detected in 6 out of 7 Nts-expressing neurons (Fig 1C) while none of these cells expressed Vglut2 (Fig1 Suppl) suggesting that the majority of MPO^Nts;hChR2 are GABAergic.

Figure 1.

Spot illumination of MPO^Nts;hChR2 neurons instantly depolarized them and increased their firing rate in all neurons tested (not shown). Recordings from non-labeled MPO neurons revealed that spot illumination of nearby MPO^Nts;hChR2 neurons resulted in a robust decrease in their firing rate (9 out of 30 neurons studied) or no effect in the others (21 out of 30 neurons studied). Fig 2A illustrates the decrease in firing rate of a MPO neuron (from 2.2 Hz to 0.9 Hz) induced by optogenetic stimulation of a nearby MPO^Nts;hChR2 neuron. The effect was associated with a robust increase in the frequency of IPSPs from 0.5 Hz to 21.8 Hz (Fig 2A, inset). Overall, optogenetic stimulation of nearby MPO^Nts;hChR2 neurons reduced the firing rate of the recorded MPO neurons from 5.10±1.32 Hz to 1.39±0.36 Hz (n=9). The control value for the firing rate was calculated as the average value during the 5 min period preceding the optogenetic stimulation. Upon ending the photostimulation a transient increase in firing rate was observed (Fig 2A). We therefore studied the effect of photostimulation during incubation with the GABA-A receptor antagonist gabazine (5 µM). In the presence of the antagonist the same neurons displayed the opposite effect, an increase in firing rate in response to photostimulation. Their firing rate was 7.89±2.61 Hz (n=9) in the presence of gabazine and increased to 9.57±2.83 Hz (n=9) during photostimulation (Fig 2A, B). Since we have previously found that exogenous Nts applied locally increased the firing activity of MPO neurons by activating NtsR1 (24) we have tested the effect of photostimulation in the presence of both gabazine (5 µM) and the NtsR1 antagonist SR48692 (100 nM). The antagonist blocked the effect of photostimulation indicating that the increase in firing rate depended on activation of NtsR1. MPO thermoregulatory neurons are a subgroup of the PACAP-expressing MPO population (7, 9). We have studied whether the MPO neurons excited by photostimulation (in the presence of gabazine) express PACAP transcripts. We have detected PACAP transcripts in 6 out of 8 neurons tested (Fig 2C).

Figure 2.

To further characterize the mechanisms involved in the modulation of the firing rates of MPO^PACAP neurons in response to optogenetic stimulation of MPO^Nts;hChR2 we carried out additional experiments in voltage-clamp mode. In the presence of the glutamate receptor blockers CNQX (10 µM) and AP-5 (50 µM), as well as of the NtsR1antagonist SR48692 (100 nM), photostimulation MPO^Nts;hChR2 neurons induced a robust increase in the IPSCs frequency (Fig 3A). These results indicate that GABA release was not dependent neither upon excitatory synaptic transmission nor upon activation of NtsR1, i.e. it is synaptically released by MPO^Nts;hChR2 neurons. We have also found that brief photostimulation (10s or less) resulted in increased frequency of IPSCs only (Fig 3B, C) while longer photostimulation (30s or more) additionally activated an inward current (Fig 3B). The inward current was isolated in the presence of extracellular gabazine (5 µM) and averaged 9.26±2.39 pA (n=10). The inward current was abolished by bath application of NtsR1 antagonist SR48692 (100 nM) (Fig 3B, D). We have carried out a set of experiments in MPO^Nts;hChR2 from female mice and recorded similar results. Optogenetic stimulation resulted in a robust increase in the frequency of IPSCs followed, with a delay of 20-40 s, by an inward current that averaged 11.3±3.4 pA (n=5) (Suppl Fig2). All the optogenetically-evoked responses were blocked by incubation with gabazine (5 µM) and SR48692 (100 nM) (Fig 3 Suppl).

Figure 3.

Finally, we have recorded from MPO^Nts;hChR2 neurons and stimulated, using spot illumination, a different MPO^Nts;hChR2 neuron to question the presence of reciprocal connections. None out of 8 MPO^Nts;hChR2 neurons studied presented either IPSCs or an inward current evoked by optogenetic stimulation of other MPO^Nts;hChR2 neurons.

Thermoregulatory preoptic neurons project to the dorsomedial hypothalamus, arcuate, paraventricular thalamus and to the rostral raphe pallidus to control thermoeffector mechanisms (8, 9, 28). We have examined the brains of MPO^Nts;eYFP mice in coronal slices to identify MPO projections to these sites, however no fluorescent signal was detected suggesting that MPO^Nts;eYFP modulate CBT by acting on MPO^PACAP neurons.

Optogenetic stimulation of MPO^Nts;ChR2 neurons in vivo induces hyperthermia. In contrast, optogenetic stimulation of MPO^{Nts;VGAT-/-;ChR2}neurons induces hypothermia

Optogenetic activation of Nts neurons in the MPO using MPO^Nts;ChR2 mice resulted in hyperthermia of up to 1.22±0.35 °C when compared with control MPO^Nts;eYFP mice (n=6 mice in each condition) (Fig 4A, B). To assess the role played by GABA release from MPO Nts neurons in the observed hyperthermia we activated Nts neurons in MPO^{Nts;VGAT-/-;ChR2} mice (see Methods). Surprisingly, a decrease in CBT of 1.44±0.29 °C was recorded relative to control MPO^{Nts;VGAT-/-;eYFP} mice (Fig 4C).

Figure 4.

Similar results were obtained also in female MPO^Nts;ChR2 mice and MPO^{Nts;VGAT-/-;ChR2} mice. Intra-MPO optogenetic stimulation resulted in a hyperthermia of 1.40±0.62 °C in MPO^Nts;ChR2 females and in a hypothermia of 1.63±0.22 °C in MPO^{Nts;VGAT-/-;ChR2} females (Fig4 Suppl).

Optogenetic stimulation of MPO^{Nts; VGAT-/-;ChR2} neurons increases the firing rate of MPO^PACAP neurons by activating an inward current

The basal electrophysiological properties of MPO^{Nts; VGAT-/-;ChR2} neurons were similar with those of MPO^Nts;ChR2 neurons. The neurons were spontaneously active, with an average firing rate at 36 °C of 4.79±2.45 Hz (n=26). Recordings from non-labeled MPO neurons revealed that spot illumination of nearby MPO^{Nts; VGAT-/-;ChR2} neurons increased their firing rate (12 out of 26 neurons studied) or had no effect in the others (14 out of 26 neurons). Fig 5A illustrates the increase in firing rate of a MPO neuron (from 1.6 Hz to 3.9 Hz) induced by optogenetic stimulation of a nearby MPO^Nts;ChR2 neuron, an effect which was associated with an apparent depolarization. Overall, optogenetic stimulation of nearby MPO^{Nts;VGAT-/-;ChR2} neurons increased the firing rate of the recorded MPO neurons from 5.68±1.66 Hz to 12.20±3.65 Hz (n=12) (Fig 5B). This effect was blocked by preincubation with the NtsR1 antagonist SR48692 (100 nM) (Fig 5A,B). In voltage-clamp experiments photostimulation activated an inward current which was abolished by bath application of NtsR1 antagonist SR48692 (100 nM) (Fig 5C, D). Using s.c. RT/PCR assays we have detected PACAP transcripts in 8 out of 11 neurons excited by photostimulation tested.

Figure 5.

We have studied the presence of VGAT and Nts transcripts in MPO slices from Nts^VGAT-/- and in control Nts^VGAT+/+ mice (Fig 5 Suppl). Nts and VGAT transcripts overlapped in control tissue while there was no colocalization in tissue from Nts^VGAT-/- mice (Fig 5 Suppl).

Characterization of the fever response, the heat exposure and the cold exposure responses as well as of the circadian CBT profile in Nts^VGAT-/-mice

Since MPO^{Nts; VGAT-/-;ChR2} mice displayed drastically different CBT responses to optogenetic stimulation relative to those of MPO^Nts;ChR2 mice, suggesting that GABA release by Nts neurons played an important role in thermoregulation, we decided to further characterize thermoregulatory responses in the Nts^VGAT-/-mice.

The LPS-induced fever was significantly different in Nts^VGAT-/- mice when compared with that of wild-type littermates Nts^VGAT+/+ (Fig 6A). During the intermediate and late phases of fever the CBT was significantly lower in Nts^VGAT-/-mice (Fig 6A).

Figure 6.

During heat exposure the CBT increased at a faster rate in Nts^VGAT-/- mice than in Nts^VGAT+/+ mice (Fig 6B). The Nts^VGAT-/- mice and Nts^VGAT+/+ mice reached 41.5 °C within 1.33±0.04 h and 1.53±0.04 h, respectively (Kruskal-Wallis, P=2.14x10^-4, n=9 mice in each group). The Nts^VGAT-/- mice displayed significantly higher CBT (by ∼0.5 °C) after 45 min of heat exposure (Fig 6B).

The circadian CBT profile of Nts^VGAT-/- mice was also altered relative to that of Nts^VGAT+/+ mice: the active phase started ∼1h earlier and ended ∼3h later (Fig 6C).

During a 3h exposure in a cold room at 4 °C the two mouse lines displayed strikingly different CBT responses (Fig 6D). The control mice displayed a gradual decrease in CBT which stabilized at 33-34 °C by third hour of the cold exposure. In contrast, Nts^VGAT-/- mice displayed a hyperthermia of 37.4 °C at the beginning of the experiment followed by a slow decline that reached ∼35 °C at the end of the three hours. Upon return to vivarium the CBT of controls returned to 37 °C while the Nts^VGAT-/-mice displayed a transient hyperthermia to 39 °C.

Activation of central Nts^hM3D(Gq) neurons induces potent hypothermia

Previous studies have reported the intracerebroventricular or intra MPO infusions of exogenous Nts induce potent hypothermia (23, 24). Since our results indicated that activation of MPO Nts neuron does not mimic this effect we hypothesized that Nts release by Nts neurons in other brain regions may be required to induce hypothermia. In order to activate all central Nts neurons we have generated Nts^hM3D(Gq) mice (see Methods) and employed chemogenetic stimulation. Activation of central Nts^hM3D(Gq) neurons induced a potent hypothermia of 4.8±0.6 °C relative to Nts-cre mice (control) (Fig 7A), value comparable with that obtained with MPO infusions of exogenous Nts (24). To test the hypothesis that the main locus of action is the MPO and to determine the possible role of NtsR1 and NtsR2 in this effect we have performed the chemogenetic activation of Nts^hM3D(Gq) in the presence of NtsR1 and/or NtsR2 antagonists in the MPO (Fig7B). The antagonists (or aCSF as control) were infused via a bilateral guide cannula 1.5 hours prior to CNO injection (i.p. 1 mg/kg). The NtsR1 antagonist SR48692 (300 nM, 100nl, blue trace) decreased the chemogenetically induced hypothermia by 41 % relative to control (Fig 7B). Higher concentrations of SR48692 (600 nM and 1.5 μM) resulted in similar reductions in the hypothermic response of 43 and 31%, respectively (n=6 mice in each condition). Following intra-MPO injections of NTSR2 antagonist NTRC 824 (200 nM, 100nl, black trace) the CNO-induced hypothermia was decreased by 32% relative to control (Fig 7B). When the two doses of antagonists were infused together the chemogenetically-induced hypothermia decreased by 76% (Fig 7B, red trace). These results indicate that the MPO accounts for most of the hypothermia induced by Nts^hM3D(Gq) neurons and that both receptor subtypes play an important role.

Figure 7.

Chemogenetic activation of NTS^hM3D(Gq) neurons and projections in the MPO results in potent excitation of MPO^PACAP

In order to understand the mechanisms by which chemogenetic activation of NTS^hM3D(Gq) neurons induced hypothermia, we have studied the effect of CNO on the activity of MPO non-labeled neurons in acute slices from NTS^hM3D(Gq) mice. Bath application of CNO excited 9 out 20 neurons studied and had no effect in the others. Fig 7C illustrates such a response in a non-labeled MPO neuron. Interestingly prior to depolarization an increase in the frequency of sIPSPs (arrow) was recorded. The average firing rate increased from 4.15±2.95 Hz to 10.63±5.83 Hz (n=9, ANOVA, F(1/16)=8.99 P=8.52x10^-3)) in response to CNO (3 µM) (Fig 7D). S.c. RT/PCR assays identified PACAP transcripts in 7 of the 9 neurons excited by CNO tested. In voltage-clamp experiments bath application of CNO (3 µM) activated an inward current and potently increased the frequencies and amplitudes of both sEPSCs and sIPSCs (Fig7 E). The frequencies of sEPSCs and sIPSCs increased from 2.76±1.16 Hz to 9.14±5.36 Hz and 1.98±1.43 Hz to 8.84±6.38 Hz, respectively (n=6)(Fig 7F,H). The amplitudes of sIPSCs and sEPSCs increased from 13.61±4.51 pA to 29.60±5.32 pA and 13.02±3.99 pA to 22.12±5.53 pA, respectively (n=6)(Fig 7G,I). The NtsR1 antagonist SR48692 (100 nM) abolished the inward current activated by CNO (Fig 7E, middle trace) and significantly reduced the effect on the frequency of sIPSCs (Fig 7H). The NtsR2 antagonist NTRC 824 (100 nM) did not change the inward current activated by CNO but significantly decreased the amplitude of sEPSCs (Fig 7G).

Discussion

In this study we have characterized a novel population of MPO neurons that express Nts (MPO^Nts). These neurons are GABAergic as indicated by detection of VGAT transcripts as well as by the activation of gabazine-sensitive IPSCs in postsynaptic neurons in response to spot illumination of MPO^Nts;hChR2 neurons. Interestingly, this population had a tendency to burst in response to depolarization, a characteristic of neurosecretory neurons (29, 30). In response to optogenetic stimulation MPO^Nts;hChR2 neurons released both GABA and Nts as indicated by pharmacological experiments. Postsynaptically, increased frequency of IPSCs was detected instantaneously upon photostimulation while an inward current was activated after 20-40 seconds, indicating a differential timecourse of action. In view of its slower timecourse of action, Nts may act to limit and/or shorten the inhibitory effect of GABA on the activity of the postsynaptic neurons following increased burst of activity of MPO^Nts neurons. Nevertheless, the net effect during photostimulation was a decrease in the firing rate of the postsynaptic neuron, an increase in firing rate being unmasked only upon ending photostimulation or in the presence of gabazine. Recordings from MPO^{Nts;VGAT-/-;hChR2} neurons confirmed the activation of an inward current in postsynaptic neurons in response to photostimulation.

In terms of local neuronal networks, MPO^Nts neurons did not present reciprocal connections instead they appeared to synapse onto MPO^PACAP neurons. PACAP has been identified as a marker of preoptic thermoregulatory neurons activated in a warm environment (9) and also during torpor (28). Activation of MPO^PACAP neurons induces hypothermia (9, 28). However, PACAP is expressed in several subpopulations of neurons, representing a large proportion of the entire preoptic population (7, 28). Recent studies have identified a subpopulation of the MPO^PACAP neurons that expresses the EP3 prostanoid receptors, is activated in warm environment and inhibited in a cold environment (8). This population plays a critical role in the initiation of the fever response (6, 8). Modulating the activity of preoptic EP3 neurons up or down induces hypothermia or hyperthermia, respectively (8). It is possible that MPO^Nts neurons overlap with a previously hypothesized population of inhibitory neurons that receive peripheral cold input and project to EP3 neurons to trigger thermogenesis and inhibit heat loss(1, 2).

In view of the potent hypothermic properties of Nts when applied peripherally, centrally or in the preoptic area it was surprising to find that photostimulation of MPO^Nts;hChR2 neurons in vivo resulted in a mild hyperthermia, instead. In contrast, photostimulation of MPO^{Nts;VGAT-/-;hChR2} neurons, which caused a net excitation of MPO^PACAP neurons, resulted in hypothermia, supporting the idea that their excitation induces hypothermia (9, 28).

Numerous studies have revealed that, in the preoptic area, GABA exerts complex effects on CBT depending on the environmental context and the specific neuronal subpopulation involved (31–34). Nts^VGAT-/- mice displayed significantly different thermoregulatory characteristics suggesting that GABA release from neurotensinergic neurons plays an important role in the respective responses. The fever response to endotoxin, when compared with the control, was characterized by lower CBT during the intermediate and late phases of the fever suggesting that GABA signaling is involved in the hyperthermic mechanisms activated. This finding is in line with the previous reports that preoptic GABA can induce thermogenesis (32, 33) however it is in disagreement with observations that the induction of fever is associated with decreased synaptic inhibition in the preoptic area (35, 36). It is possible that both mechanisms influence CBT during fever with the former playing a prevalent role in the intermediate and later phases.

In a warm environment preoptic release of GABA decreases and exogenous GABA-A antagonists applied locally have hyperthermic effects (32), thus Nts^VGAT-/- would be expected to have a higher CBT during a heat challenge. Indeed, we observed that during heat exposure the CBT of Nts^VGAT-/- mice increased significantly faster than in the controls suggesting that in a warm environment GABA release from neurotensinergic neurons is involved in the control of heat loss mechanisms.

During cold exposure Nts^VGAT-/- mice displayed higher CBT relative to controls. Previous studies have revealed increased endogenous release of GABA in the preoptic area during cold exposure and hypothermic effects of exogenous GABA-A antagonist in a cold environment (32). Thus, a decreased release of GABA from Nts^VGAT-/- neurons would be expected to induce a larger hypothermia during cold exposure, the opposite of the observed results. However, while an increase in preoptic GABA concentration induces hyperthermia, at all ambient temperatures, a widespread central increase in GABA is hypothermic (31, 37), therefore it is possible that VGAT deletion in Nts neurons outside of the MPO may impact CBT during cold exposure and/or that the respective regions are differentially recruited during distinct thermoregulatory mechanisms.

Finally, Nts^VGAT-/- mice have entered the “up” phase of the circadian CBT rhythm earlier and ended it later when compared to the control, supporting a role of neurotensinergic neurons, possibly in the suprachiasmatic nucleus, in the control of circadian rhythm (38, 39).

To question whether stimulation of all central neurotensinergic neurons induces hypothermia we have generated Nts^hM3D(Gq) mice. Chemogenetic activation of central Nts^hM3D(Gq) neurons resulted in a robust hypothermia, comparable to that induced by central infusion of exogenous neurotensin. The response was largely blocked by NtsR1 and NtsR2 antagonists injected in the MPO, suggesting that the locus of action is the MPO. Taken together our results suggest that stimulation of neurotensinergic neurons in the MPO is not sufficient to induce hypothermia although this region represents the locus where the hypothermia is triggered.

Experiments at the cellular level showed that locally applied CNO in slices from Nts^hM3D(Gq) mice had a net excitatory effect in nearby MPO PACAP neurons. In addition to the actions observed during photostimulation of MPO^Nts;hChR2, namely increased sIPSCs frequency and an inward current, CNO also increased the frequency and amplitudes of sEPSCs recorded in MPO^PACAP neurons. This may be due to activation of glutamatergic projections from brain regions that contain glutamatergic Nts^hM3D(Gq) neurons. Such neurons have been identified in the posterior thalamus and in the ventrolateral periaqueductal gray (19, 20). It is also possibIe that some of the parabrachial nucleus neurons’ glutamatergic projections (40) are also neurotensinegic. It is likely that the chemogenetic stimulation of MPO Nts^hM3D(Gq) neurons and projections elevates the local Nts concentration to a level that activates the low affinity NtsR2. These receptors are expressed in MPO astrocytes and upon activation stimulate glutamate release from nearby synaptic terminals as observed when exogenous Nts is applied locally (24). A schematic model of the network and of the cellular mechanisms involved in the CBT modulation by neurotensinergic neurons is presented in Fig 8.

Figure 8.

In summary, this study characterizes the cellular properties of MPO^Nts neurons and their influence on CBT and provides insights in the cellular mechanisms at the MPO level that result in hypothermia.

Materials and Methods

Animals

Experiments on animals were carried out in accordance with the National Institute of Health Guide for the care and use of Laboratory animals (1996 (7th ed.) Washington DC: National Research Council, National Academies Press). The protocols were reviewed and approved by the Institutional Animal Care and Use Committee of the Scintillon Institute. The standards are set forth by the American Association for the Accreditation of Laboratory Animal Care (AAALAC) and in the Animal Welfare Act. The work was designed in such a way to minimize the number of animals used as well as their suffering.

The Nts-cre driver line (B6;129-Ntstm1(cre)Mgmj/J; stock no: 017525) (23) was purchased from Jackson Laboratory (Bar Harbor, ME, USA). The Nts-cre driver line was crossed with the Slc32a1tm1Lowl (also referred to as VGATflox/flox) (Jackson Laboratory; stock no: 012897) to generate mice that have the GABA vesicular transporter VGAT (solute carrier family 32 member 1), knocked-down in Nts neurons. Heterozygous double transgenic mice Nts+/- ;Vgat+/-were crossed to obtain Nts+/+;Vgat-/-. These mice are referred to as Nts^VGAT-/-. We have also crossed the Nts-cre driver line with the B6N;129-Tg(CAG-CHRM3*,-mCitrine)1Ute/J line (Jackson Laboratory; stock no: 026220) to generate mice that express the excitatory designer receptor hM3D(Gq) in Nts-neurons in all brain regions. This new line was named Nts^hM3D(Gq). In chemogenetics experiments, to activate hM3D(Gq), mice received an i.p. injection of CNO (0.7-0.8 mg/kg).

Transgenic animals and AAV injections

The Nts-cre homozygous mice (4 to 6 months old) received bilateral stereotaxic injections (200 nL at a rate of 0.1 μL/min) of AAV-EF1a-double floxed-hChR2(H134R)-eYFP-WPRE-HGHpA (Addgene #20298; 2.1 x 10¹³ virus molecules/ml) or AAV5-EF1a-DIO-eYFP (Addgene #27056; 1.3 x 10¹³ virus molecules/ml) in the MPO to express the opsin Channelrhodopsin2 (hChR2) or eYFP (control) in MPO Nts neurons. We refer to these mice as MPO^Nts;hChR2 and MPO^Nts;eYFP, respectively. To ensure that only cre-expressing neurons were targeted we have carried out control injections of the two viral vectors in wild-type C57/Bl6 mice (3 mice for each) and confirmed that no fluorescent signal was detected in the brain of these mice. We have carried out RNAscope for Nts (see below) to determine the colocalization with eYFP (Fig 9) in 3 MPO^Nts;hChR2 mice. We have studied colocalization from 6 randomly chosen field of view from 3 slices from 3 different animals and found colocalization in 82% (355 out of 433) of the Nts positive neuros. Only 4 out of 359 eYFP neurons appeared to be Nts-negative (1.1%). Together with the detection of Nts transcripts in the labeled neurons (see Results section) we conclude that we have specifically targeted Nts neurons using these viral vectors.

Figure 9.

Finally, we have performed bilateral stereotaxic injections (200 nL at a rate of 0.1 μL/min) of the AAV5-EF1a-double floxed-hChR2(H134R)-eYFP-WPRE-HGHpA (Addgene #20298; 2.1 x 10¹³ virus molecules/ml) or AAV5-EF1a-DIO-eYFP (Addgene #27056; 1.3 x 10¹³ virus molecules/ml) in the MPO of Nts^VGAT-/- to express ChR2 or eYFP, in MPO Nts^VGAT-/- neurons. We refer to these mice as MPO^{Nts;VGAT-/-;hChR2} and MPO^{Nts;VGAT-/-;eYFP}, respectively.

All experiments were carried out in male mice. Key sets of experiments were also performed in female mice as specified in the text.

Estrous cycle monitoring

Female mice have a 4–6 day estrous cycle that consists of four stages – proestrus, estrus, diestrus I/metestrus, and diestrus/II. Mice were habituated to handling and vaginal smears prior to the behavioral experiments. Mice were swabbed daily before 9a.m. and experiments started after 10 am. The stage of the cycle was determined using vaginal cytology. The smears were stained with hemalum-eosin and analyzed microscopically. Experiments in females were carried out in diestrus days since during this stage the baseline CBT is less variable for each animal and among different animals. The days of diestrus are characterized by the presence of leukocytes and the absence of keratinized epithelial cells with no nuclei.

Telemetry and MPO Injections

For CBT measurements the mice (4 to 6 months old) were anesthetized with isoflurane (induction 3%-5%, maintenance 1%-1.5%) and radio telemetry devices (Anipill, BodyCap, Hérouville Saint-Clair, France) were surgically implanted into the peritoneal cavity. For MPO injections, mice were stereotaxically implanted with a bilateral guide cannula (27 Ga) as described in our previous studies (41). Coordinates for cannula implants in the MPO were: from Bregma: 0.05 mm, 0.35 mm, and −0.35 mm lateral, and ventral 4.75 mm (42).The ambient temperature was maintained at ∼28 ± 0.5°C in a 12:12-hour light-dark cycle-controlled room (lights on 8:00 am, ZT0). All substances injected were dissolved in sterile artificial cerebrospinal fluid (aCSF). For MPO injections, mice were placed in a stereotaxic frame and the injector (33 Ga) was lowered inside the cannula. A volume of 100 nL (rate 0.1 μL/min) was delivered using an injector connected to a microsyringe (0.25 μL). After injections the animal was returned to the home cage. Injections were carried out at 10 am local time, during the “subjective light period.”

Slice Preparation

Coronal tissue slices containing the MPO were prepared from mice 4-6 months old MPO^Nts;hChR2, MPO^{Nts;VGAT-/-;hChR2} or Nts^hM3D(Gq) mice. The slice preparation was as previously described (41). The slice used in our recordings corresponded to the sections located from 0.15 mm to -0.05 mm from Bregma in the mouse brain atlas (42).

Whole-cell patch-clamp recording

Whole-cell patch-clamp clamp was performed as described in our previous studies (41). The aCSF contained (in mM) the following: 130 NaCl, 3.5 KCl, 1.25 NaH₂PO₄, 24 NaHCO₃, 2 CaCl₂, 1 MgSO₄, and 10 glucose, osmolarity of 300–305 mOsm, equilibrated with 95% O₂ and 5% CO₂, pH 7.4. Other salts and agents were added to this saline. Whole-cell recordings were carried out using a K⁺ pipette solution containing (in mM) 130 K-gluconate, 5 KCl, 10 HEPES, 2 MgCl₂, 0.5 EGTA, 2 ATP and 1 GTP (pH 7.3). The electrode resistance after back-filling was 2–4 MΩ. All voltages were corrected for the liquid junction potential (−13 mV). Data were acquired with a MultiClamp 700B amplifier (Molecular Devices, Sunnyvale, CA, USA) digitized using a Digidata 1550 interface and the Pclamp10.6 software package. The sampling rate for the continuous recordings of spontaneous activity was 50 kHz. The cell capacitance was determined and compensated using the Multiclamp Commander software.

The recording chamber was constantly perfused with extracellular solution (2–3 mL·min⁻¹). The antagonists were bath-applied. The bath temperature was maintained at 36–37°C by using an inline heater and a TC-344B temperature controller (Warner Instruments, Hamden, CT, USA).

Synaptic activity was quantified and analyzed statistically as described previously (41, 43). Synaptic events were detected and quantified (amplitude, kinetics, frequency) off-line using a peak detection program (Mini Analysis program, Synaptosoft, Decatur, NJ, USA). Events were detected from randomly selected recording stretches of 2 min before and during incubation with pharmacological agent. Cumulative distributions of the measured parameters (inter-event interval, amplitude, rise time, time constant of decay) were compared statistically using with the Kolmogorov-Smirnov two-sample test (K-S test, P<0.05) using the Mini Analysis program. The averages for the measured parameters (frequency, amplitude, rise time, time constant of decay) for each experiment were obtained using the Mini Analysis program. Event frequency was calculated by dividing the number of events by the duration (in seconds) of the analyzed recording stretch.

Spot illumination of MPO^Nts;hChR2 neurons was carried out using a Polygon300 illumination system (Mightex, Toronto, Canada) that allows the control of the size, shape, intensity and position of the illuminated spot as described previously. In a field of view there were 1-4 fluorescent neurons. The position and size of the spot were controlled using the PolyScan2 software (Mightex). The intensity of the light at the bottom of the recording chamber was measured using a photosensor (ThorLabs, Newton, NJ, USA) and ranged from 1.2 to 1.4 mW/mm^-2. The light pulses were 50 ms long and delivered at 10Hz for durations of 20 to 120 s.

Optogenetic stimulation in vivo

A dual fiber-optic cannula (100µm diameter, NA 0.22, 0.7 mm pitch,) (Doric Lens, Quebec, Canada) was implanted as described above for the guide cannula. Light pulses were applied using high-power photodiodes, a digital and analog I/O control module and Polygon 300 software (Mightex, Toronto, Canada) via dual fiber cables (Doric Lens, Quebec, Canada). The intensity of the light at the end of the optic cannula was measured using a fiber photodiode power sensor (ThorLabs, Newton, NJ, USA) and ranged from 3.9 to 4.3 mW/mm^-2. The light pulses were 50 ms long and delivered at 10Hz for 30 s periods followed by 10s recovery breaks. The total duration of light stimulation was 60 min.

Chemicals

Agonists and antagonists were purchased from Tocris (Ellisville, MO, USA). The other chemicals were from Sigma (Carlsbad, CA, USA).

Cell Harvesting, Reverse Transcription and PCR

MPO neurons in slices were patch-clamped and then harvested into the patch pipette by applying negative pressure as previously described (41, 43). The content of the pipette was expelled in a PCR tube. dNTPs (0.5 mM), 50 ng random primers (Invitrogen) and H₂O were added to each cell to a volume of 16 μl. The samples were incubated at 65°C for 5 min and then put on ice for 3 min. First strand buffer (Invitrogen), DTT (5 mM, Invitrogen), RNaseOUT (40 U, Invitrogen) and SuperScriptIII (200 U, Invitrogen) were added to each sample to a volume of 20 μl followed by incubation at room temperature for 5 min, at 50°C for 50 min and then at 75°C for 15 min. After reverse transcription samples were immediately put on ice. 1 μl of RNAse H was added to samples and kept at 37°C for 20 min. PCR assays were carried out using the pairs of primers listed in Suppl Table 1. The PCR products were verified using Sanger sequencing. Two rounds of PCR, using nested primers, were used to detect Nts and Vglut2 transcripts. For PACAP and VGAT a second round of PCR was not necessary since it did not yield additional positives when compared with the first round. Two negative controls were routinely carried out. The first one was amplified from a harvested cell without reverse-transcription while the second was represented by a RT/PCR of the “pipette tip” (e.g. obtained when a successful gigaseal was not achieved and the cytoplasm was not harvested by suction). Since all single MPO cells tested were negative to Vglut2 a second positive control was carried out. Ventromedial hypothalamic neurons, a predominantly glutamatergic population was tested. The positivity rate was 90% (9 of 10 neurons tested) (Suppl Fig1B).

RNAscope assays

Tissue was processed using in situ hybridization to detect mRNA for Nts, VGAT (Slc32a1). 40-μm thick coronal sections fixed in 4% PFA for 15 min at 4°C, dehydrated in serial concentrations of ethanol (50%–100%), and processed according to the protocol provided in the RNAscope kit (ACDbio Cat# 320293). Sections were hybridized with the following mixed probes; Nts (Mm-Nts, Cat. 420441) and VGAT (Mm-Slc32a1-C2, Cat. 319191-C2) for 2 h at 40°C.

Endotoxin induced fever, heat and cold exposure tests

To induce a fever response we injected i.p. a dose of 0.1 mg/kg lipopolysaccharides (LPS) dissolved in 0.3 ml sterile saline. To study the CBT change to heat exposure the mice were placed, in their home cages, in a 37 °C incubator. Their CBT was monitored continuously by telemetry, and an animal was removed from the incubator once it reached a CBT of 41.5 °C. For cold exposure tests the mice were placed in a cold room (4 °C) for 3h. All the experiments were started at 10 a.m. (ZT2).

Data quantification and statistical analysis

The values reported are presented as mean ± standard deviation (SD). We used power analyses (www.biomath.info) with values from our data (means and SDs) to calculate that our study was powered to detect a 0.7 °C change in CBT with at least >80% reliability for all transgenic models used in this study. The normality of the samples was tested using the Shapiro-Wilk and Anderson-Darling tests. Statistical significance of the results pooled from two groups was assessed with t-tests or Mann-Whitney test using Prism9 (GraphPad Software). One-way analysis of variance (ANOVA, Kruskal-Wallis) with Tukey’s post hoc test (P<0.05) was used for comparison of multiple groups. Cumulative distributions were compared with the Kolmogorov-Smirnov test (P<0.05). Data collected as time series were compared across time points by one-way ANOVA with repeated measures (P<0.05) (Prism4, GraphPad Software), followed by Mann Whitney U tests (P<0.05) for comparisons at each time point. The statistical value, the degrees of freedom and the p value are reported in the figure legends or, if data is not presented in a figure, the respective values are reported in the text. P values for the results of Tukey’s tests are presented in tables.

Key resources table

Acknowledgements

This is manuscript number 1073 from the The Scintillon Institute. This research was supported by the National Institutes of Health Grants NS094800 and NS124844 (I.V.T.).