Electrophysiological Properties of Genetically Identified Histaminergic Neurons

Abstract

Histaminergic neurons of the tuberomammillary nucleus (TMN) are important regulators of behavioral and homeostatic processes. Previous work suggested that histaminergic neurons exhibit a characteristic electrophysiological signature, allowing for their identification in brain slice preparations. However, these previous investigations focused on neurons in the ventral subregion of the TMN of rats. Consequently, it remains unclear whether such electrophysiological properties extend to mice, including other subregions of the TMN, and the potential for differences between males and females. To further characterize the electrophysiological properties of histaminergic neurons, we performed whole-cell patch-clamp recordings on transgenic mice expressing Cre recombinase in histidine decarboxylase (HDC)-expressing cells; the sole enzyme for histamine synthesis (Hdc-cre::tdTomato). Despite similarities with the electrophysiological properties reported in rats, we observed considerable variability in mouse HDC neuron passive membrane properties, action potential firing, and intrinsic subthreshold active membrane properties. Overall, the electrophysiological properties of HDC neurons appeared similar across subregions of the TMN, consistent with a lack of topographical organization in this nucleus. Moreover, we found no obvious sex differences in the electrical excitability of HDC neurons. However, our data reveal a diversity in the electrophysiological properties of genetically identified histaminergic neurons from mice not previously appreciated from rat studies. Thus, these data highlight the utility of mouse genetics to target the widespread histaminergic neuronal population within the TMN and support the idea that histaminergic neurons are a heterogeneous neuronal population.

INTRODUCTION

Histamine is a biogenic amine involved in multiple physiological functions, including the regulation of arousal, energy homeostasis, body temperature, and the immune response (Schwartz et al., 1991; Schneider et al., 2002; Haas et al., 2008). Within the central nervous system (CNS), histamine acts as a neurotransmitter, with histaminergic neuron somas being confined to the tuberomammillary nucleus (TMN) in the posterior hypothalamus (Panula et al., 1984; Takeda et al., 1984; Watanabe et al., 1984). Histaminergic neurons project extensively throughout the brain and spinal cord, with histamine exerting its physiological effects via activation of four G-protein coupled histamine receptors (H1–H4) (Ericson et al., 1987; Haas and Panula, 2003). The wide range of physiological functions influenced by histaminergic neurons, and their heavy innervation of the CNS, has led to concentrated efforts in understanding the regulation of histamine and histaminergic neurons in multiple research settings.

Identification of histaminergic neurons in brain slice preparations have traditionally been aided by their characteristic electrophysiological properties. Initial studies in rats reported that histaminergic neurons were spontaneously active, with a resting membrane potential of −50 mV, a distinctive A-like current (I_A – outward rectification), hyperpolarization-activated time-dependent inward rectification (I_h), and a prominent after-hyperpolarization following action potentials (Haas and Reiner, 1988). Subsequent studies have exploited these characteristics to identify histaminergic neurons in both rats (Greene et al., 1990; Kamondi and Reiner, 1991; Schonrock et al., 1991; Eriksson et al., 2001a, b; Parmentier et al., 2009; Yanovsky et al., 2011; Yin et al., 2019) and mice (Mochizuki et al., 2000; Schone et al., 2012).

Overall, limited comparisons have been made between the electrophysiological properties of histaminergic neurons from rats and mice, and it is currently unclear if species differences exist. In addition, the original characterization of histaminergic neurons was constrained to TMN neurons near the ventral surface of the brain (Haas and Reiner, 1988). The TMN is a diffuse nucleus, and is categorized into multiple sub-regions (Kohler et al., 1985; Ericson et al., 1987; Inagaki et al., 1990), which are more widespread in mice than rats (Karlstedt et al., 2001). Growing evidence suggests that histaminergic neurons are a functionally heterogeneous population (Arrang et al., 1991; Miklos and Kovacs, 2003; Giannoni et al., 2009; Blandina et al., 2012; Michael et al., 2020). Therefore, the identification of mouse histaminergic neurons based on electrophysiological properties reported from one subregion of the TMN in rats may not be entirely accurate.

The rapid increase and reliance on mouse genetics has allowed for the development of mouse models aiding in the targeting of specific neuronal populations. Histaminergic neurons are characterized by the expression of histidine decarboxylase (HDC), the sole enzyme required for histamine synthesis, which catalyzes the formation of histamine from the amino acid l-histidine (Taylor and Snyder, 1972; Green et al., 1987). Multiple HDC Cre-recombinase mouse models have been developed (GENSAT project: Gong et al., 2007; Yanovsky et al., 2012; Zecharia et al., 2012), allowing for identification of histaminergic neurons in brain slices when mice are bred to fluorescent reporter mice.

Fujita et al. (2017) utilized such transgenic mice and provided comprehensive electrophysiological data from genetically identified HDC neurons. Unfortunately, this report was also limited to HDC cells in the ventral TMN. The penetrance of the Cre recombination in this model was below 50%, further restricting the sample of HDC neurons used for characterization (Fujita et al., 2017). Therefore, the present study aimed to determine the electrophysiological properties of genetically identified histaminergic neurons from mice. In addition to contrasting these characteristics with that originally obtained by Haas and Reiner (1988) in rats, we also aimed to determine if HDC neurons are electrophysiologically similar across the subregions of the TMN, and between male and female mice.

EXPERIMENTAL PROCEDURES

Animals

To target genetically identified histaminergic neurons we used the Hdc-Cre mouse developed by Jeffrey Zigman, which has been previously validated (Yanovsky et al., 2012; Walker et al., 2013). The Hdc-Cre mouse was bred to a tdTomato reporter mouse (Ai14, JAX stock # 007914) obtained from The Jackson Laboratory (Maine), to allow for fluorescent labeling of Hdc-expressing cells. All mice were bred to have one copy of each transgene and were maintained on a C57BL/6J background.

Mice were maintained on a 12 h light-dark cycle in a temperature-controlled conventional facility (ARC at UT Southwestern) with free access to food and water. Mice were fed a standard chow diet (Envigo Teklad Global 2016 Diet, 16% protein 4% fat) and were sacrificed for electrophysiology 2–3 h after the beginning of the light period.

All experiments adhered to the guidelines established by the National Institute of Health Guide for the Care and Use of Laboratory Animals, and were approved by the University of Texas Southwestern Medical Center Institutional Animal Care and Use Committee (IACUC).

Electrophysiology

Whole-cell patch-clamp recordings were obtained from tdTomato fluorescently-labeled HDC neurons located in the TMN of Hdc-cre::tdTomato mice. Male and female adult mice were utilized and were between 8 and 14 weeks of age.

Briefly, mice were anesthetized with chloral hydrate, decapitated, and the brain rapidly removed and cut in a modified, sucrose-based, ice-cold artificial cerebrospinal fluid (aCSF) containing (in mM): 213 sucrose; 2.5 KCl; 5 MgCl₂; 1 CaCl₂, 1 NaH₂PO₄, 26 NaHCO₃, and 10 d-glucose. Posterior hypothalamic brain slices of 250 μm (coronal sections) containing the TMN were prepared on a Leica VT1000S vibratome. Slices were then incubated at 32 °C in standard aCSF containing (in mM): 126 NaCl; 2.8 KCl; 2.5 CaCl₂, 1.25 NaH₂PO₄, 26 NaHCO₃, 1.2 MgSO₄; and 10 d-glucose, and remained in this solution until slices were transferred to the recording chamber where they were continuously perfused with a reduced d-glucose concentration (5 mM) version of the standard heated aCSF.

HDC neurons in the TMN were visually identified using an Axioskop FS2 (Zeiss) microscope fitted with DIC optics, infrared videomicroscopy and florescence (Zeiss filter set 31 – excitation: BP 565/30, beamsplitter FT 585, Emission BP 620/60). Following electrophysiological recording, the subregion of the TMN that the cell was located was recorded. Different nomenclature describing the different subregions of the TMN are often used. In this study HDC neurons were classified into three sub-populations, the dorsal TMN, the ventral TMN (core) and the TMN bridge as described by Kohler et al. (1985). The dorsal TMN represents HDC cells clustered above and around the third ventricle, and the ventral TMN represents HDC cells located on the ventral surface of the brain (as illustrated in Franklin and Paxinos, 2007; Kohler et al., 1985). The dorsal TMN cells correspond to the E4 region described by Inagaki et al. (1990), and the ventral TMN correspond to the E1, E2 and E3 regions. The TMN bridge represents HDC cells that are scattered between the dorsal and ventral regions, corresponding to the E5 region described by Inagaki et al. (1990) or often referred to as the diffuse region by others (Kohler et al., 1985; Ericson et al., 1987). The location of the cell was manually recorded on an anatomical map adapted from the mouse brain atlas (Franklin and Paxinos, 2007).

Patch pipettes were pulled from thin-walled borosilicate glass (TW150F-4, World Precision Instruments) with resistances between 4 and 7MΩ when filled with an intracellular solution containing (in mM): 120 K-gluconate, 10 KCl, 1 NaCl, 1 MgCl₂, 1 CaCl₂, 5 EGTA, 10 HEPES, and 2 Mg₂ATP), pH adjusted with KOH. Whole-cell patch-clamp recordings were made using a Multiclamp 700B amplifier (Molecular Devices LLC, Sunnyvale, CA, USA). Current clamp data was digitized and filtered at 4 kHz. Signals were stored on a personal computer for analysis with pClamp10 software (Molecular Devices). Resting membrane potential is reported as the read-out from the amplifier, un-corrected for the liquid junction potential offset (approx. −8 mV).

Voltage clamp data was obtained for a small proportion of the HDC neurons recorded and was filtered at 2 kHz. Spontaneous excitatory post synaptic currents (EPSCs) were recorded using a holding potential between post −60 mV and −70 mV and inhibitory post synaptic currents (IPSCs) −15 mV. PSC frequency was measured using the Mini Analysis program (Synaptosoft, Inc.).

All current clamp and voltage clamp assessments were performed in standard aCSF solutions (described above), free from any inhibitors of synaptic transmission unless otherwise stated.

Quantification of membrane properties

After entering a whole-cell configuration, all HDC neurons were required to display a stable recording of at least (1–2 min) prior to the assessment of their biophysical properties. Resting membrane potential and firing frequency were obtained for all cells recorded. In any case where a HDC neuron only displayed one action potential every 10 seconds or more (≤ 0.1 Hz) the cell was considered silent and the firing frequency was not reported.

Current voltage relations were performed to determine the input resistance (MΩ) of the cell. This was achieved by measuring the change in membrane potential (mV) in response to a series of current injections (pA). Current voltage relations were also visually inspected for the presence of any sub threshold active conductance’s. Transient outward rectification (A-like current, I_A), hyperpolarisation-activated time-dependent inward rectification (I_h), and instantaneous inward rectification (I_IR) could be observed in HDC neurons and their presence/absence was noted for each cell.

Drugs

CNQX, AP5, and picrotoxin were used to confirm if EPSCs and IPSCs occurring on HDC neurons were mediated by glutamate and GABA release respectively. Tetrodotoxin (TTX) was used to synaptically isolate HDC neurons, cesium chloride was used to block I_h, and barium chloride to block I_IR. CNQX, AP5, TTX, and picrotoxin (all from Tocris) were made as concentrated stocks in distilled water, aliquoted, and stored at <4 °C. Cesium chloride and barium chloride (Sigma) were made as concentrated stocks in distilled water and stored at 4 °C. All drugs were diluted to the required concentration in aCSF immediately prior to use. All drugs were bath applied to the slice by a peristaltic pump connected to the main perfusion line.

Statistical analysis

All electrophysiological data was analyzed using Clampfit 10 (MDS Analytical Technologies) or mini analysis (Synaptosoft). All data are presented as means ± SEM. Two-tailed unpaired t-tests and two-way ANOVA were generated using GraphPad Prism 8 for comparisons between conditions. Chi square statistics were used to test differences in response type between groups. Parametric statistics were used when the data from both conditions adopted a Gaussian distribution, and nonparametric statistics were used when this assumption was not met.

RESULTS

Passive membrane properties and action potential firing in HDC neurons

Histaminergic neurons were identified via red fluorescence in Hdc-cre::tdTomato mice and were clearly discernable from surrounding neurons (Fig. 1A). HDC expressing neurons were targeted for electrophysiological recordings throughout the rostro-caudal axis of the TMN. Whole-cell patch-clamp recordings were performed on 141 HDC neurons from male mice. HDC neurons displayed an average resting membrane potential of −47.6 ± 0.5 mV (n = 141, Fig. 1B, C). Half (71/141) of the HDC neurons were silent. However, HDC neurons exhibited a firing frequency of 1.3 ± 0.1 Hz (n = 70) in those cells that were active (Fig. 1D). HDC neurons displayed an average input resistance of 749 ± 22MΩ (Fig. 1E). All HDC neurons targeted for electrophysiological recording were mapped across the rostral/caudal and medial/lateral axis of the TMN, with approximately equal numbers of cells from each subregion of the TMN (Fig. 1F).

Fig. 1.

Genetically identified HDC neurons in the TMN were targeted for electrophysiological recordings. (A) Brightfield illumination of a patch pipette targeting a Hdc-cre::tdTomato neuron, the same neuron illuminated with fluorescence (Zeiss filter set 31), and merge of the images. (B) Current clamp recording from a HDC neuron in the TMN of a male mouse. This example is representative of the average membrane potential and firing frequency displayed in HDC neurons from male mice. (C) Frequency histogram showing the distribution of membrane potential values in HDC neurons from male mice. Values ranged from −38.5 to −63.0 mV with a standard deviation of 5.4 mV. (D) Frequency histogram showing the distribution of firing frequencies in HDC neurons from male mice. Values ranged from 0.1 to 4.1 Hz with a standard deviation of 0.9 Hz. (E) Frequency histogram showing the distribution of input resistance values in HDC neurons of male mice. Values ranged from 362 to 1764 MΩ with a standard deviation of 260 MΩ. (F) HDC neurons patched in male mice were mapped across the rostral caudal and medial lateral axis of the TMN. Neurons were grouped into the three subregions of the TMN for further analysis of their electrophysiological properties. Note: Image modified from: Mouse Brain: In stereotaxic coordinates (3rd ed.). Franklin, K. B. J., & Paxinos, G. Copyright Elsevier (2007).

Intrinsic subthreshold active membrane properties

One of the most striking features of HDC neurons is the presence of a transient outward rectification, or A-like current (I_A) (Haas and Reiner, 1988; Haas et al., 1989). I_A was clearly visible in HDC neurons and could be observed as a prolonged time-course of recharging following hyperpolarizing current injections (Fig. 2A). The voltage dependence of the transient outward rectification could be observed by holding the HDC neurons at different membrane potentials. At hyperpolarized membrane potentials, I_A was observed as a delayed membrane response to depolarizing current injections (Fig. 2B).

Fig. 2.

Active conductance’s expressed in HDC neurons. (A) An obvious A-like current can be seen in HDC neurons, as demonstrated by a delayed return to a resting membrane potential following negative current injection. Arrow marks the strong transient outward rectification visible in HDC neurons. (B) Current clamp recordings of positive and negative current injections demonstrates the voltage dependence of the A-like current easily observed in HDC neurons. At more depolarized potentials, the A-like current is visible following negative current injection, indicated by the arrows. At more hyperpolarized potentials the A-like current is visible following positive current injections and manifests as a delayed membrane response to depolarizing currents. (C) An h current could be observed in HDC neurons, indicated by arrow. The I_h is visible as a “depolarizing sag” following hyperpolarizing current injections, particularly following larger negative current injections. (D) Current-clamp recording following −100 pA current injection. I_h (indicated with arrows) displayed sensitivity to Cesium chloride 1 mM following chemical isolation of the HDC neuron using TTX 500 nM. The block of I_h is clearly visible, indicated by removal of the “depolarizing sag” following CsCl application. (E) Instantaneous time-independent inward rectification could be observed in HDC neurons, indicated by the arrow. The membrane shows a non-linear response following equally spaced negative current injections. (F). Current-clamp recording following negative current injections (15 pA steps). I_IR(indicated with arrows) displayed sensitivity to barium chloride 100 μM following chemical isolation of the HDC neuron using TTX 500 nM. (G) The proportion of cells expressing each subthreshold active conductance is graphed. The majority of HDC neurons express an A-like current, approximately half express I_h, and relatively few express inward rectification.

Another characteristic feature of HDC neurons is the presence of a hyperpolarization-activated cyclic nucleotide-gated non-selective cation conductance, or anomalous inward rectification (I_h) (Haas and Reiner, 1988; Kamondi and Reiner, 1991). I_h was visible in HDC neurons and was observed as a slowly developing inward rectification, or ‘depolarizing sag’ in the membrane response following hyperpolarizing current injections, particularly following larger negative current injections (Fig. 2C). The presence of this current was further confirmed in synaptically isolated HDC neurons (in the presence of TTX 500 nM) by its sensitivity to Cs⁺ (CsCl 1 mM, n = 6/6) (Fig. 2D).

Finally, an instantaneous time-independent inward rectification (I_IR) was expressed in a small proportion of HDC neurons, evidenced as a nonlinear response in the membrane potential to negative current injections of equal increments. I_IR was activated at hyperpolarized membrane potentials and most pronounced in response to larger amplitude negative current injections (Fig. 2E). The presence of this current was further confirmed in synaptically isolated HDC neurons (in the presence of TTX 500 nM) by its sensitivity to Ba⁺ (BaCl 1 mM, n = 3/3), an inwardly rectifying potassium channel blocker (Fig. 2F).

Overall, HDC neurons displayed differential expression of these subthreshold active conductances. I_A was the most prominent, being expressed in almost all (94%, 133/141 cells) HDC neurons. I_h was expressed in just over half of the HDC neurons (57%, 80/141 cells), and I_IR was only observed in a small proportion of HDC neurons (11%, 16/141 cells) (Fig. 2G).

Comparisons between the different TMN subregions

HDC neurons were classified into three sub-populations based on their location in the posterior hypothalamus (TMN ventral/core, dorsal, and bridge). Electrophysiological properties of HDC neurons in each subregion were compared to assess for heterogeneity in electrical excitability. The resting membrane potential for HDC neurons was −46.9 ± 0.7 mV, −47.2 ± 0.7 mV and −49.0 ± 0.9 mV for the TMN core, dorsal TMN and TMN bridge, respectively, and was not significantly different between the three TMN regions (Kruskal–Wallis test = 3.25, p = 0.197, n = 141) (Fig. 3A). The proportion of cells that were active, and the action potential frequency at which they fired, were also similar between the three subregions of the TMN (TMN core: 56.6%, dorsal TMN: 52.1% and TMN bridge: 37.5% active) (firing rate TMN core: 1.2 ± 0.1 Hz, dorsal TMN: 1.0 ± 0.1 Hz and TMN bridge: 1.9 ± 0.3 Hz, Kruskal–Wallis test = 7.32, p = 0.0257, n = 70), although the action potential frequency in the TMN bridge was slightly higher than in the dorsal TMN (Dunn’s multiple comparison test, p = 0.0220) (Fig. 3B, C).

Fig. 3.

Electrical excitability and membrane properties were similar between the TMN subregions. (A) Membrane potential was not significantly different between the three TMN regions. (B) The proportion of cells that were active were not significantly different between the three TMN regions. (C) The firing frequency of active cells was not significantly different between the three TMN regions. (D) Input resistance was not significantly different between the three TMN regions. (E) The proportion of cells expressing each active conductance was not statistically different between the three TMN regions.

Input resistances of HDC neurons were similar between the TMN core, dorsal TMN and TMN bridge (727 ± 39 MΩ, 803 ± 35 MΩ, and 715 ± 39MΩ, respectively, Kruskal–Wallis test = 5.63, p = 0.0598, n = 141) (Fig. 3D). When the expression of active conductances were examined across the sub regions of the TMN, there was no difference in the proportion of cells expressing I_A, I_h or I_IR in these three subregions (Fisher’s exact test p = 0.5347, p = 0.7555 and p = 0.6209 respectively) (Fig. 3E). Overall, these data demonstrate similar electrophysiological properties of HDC neuron across the different subregions of the TMN.

Sex differences in HDC neurons

The sexual dimorphism of the hypothalamus, including the histaminergic system, is not completely understood. To better examine sex differences of TMN histaminergic neurons, 60 HDC neurons from female mice were targeted for electrophysiological recordings and compared with HDC neurons from males. HDC neurons were sampled from the entire TMN (Fig. 4A). Resting membrane potential was not significantly different in HDC neurons obtained from male or female mice (male: −47.6 ± 0.5 mV, female: −48.0 ± 0.7 mV, Mann–Whitney U = 4036, p = 0.6086) (Fig. 4B). The activity of the neurons were similar between the sexes with 49.6% of HDC neurons being active in males, and 36.7% being active in females (Fisher’s exact test p = 0.1215), along with no significant difference in firing rate in those cells that were active (male: 1.3 ± 0.1 Hz vs female: 1.3 ± 0.2 Hz, Mann–Whitney U = 717, p = 0.6312) (Fig. 4C, D). Neuronal input resistance was also similar between male and female mice (749 ± 22MΩ, 746 ± 42MΩ, respectively, Mann–Whitney U = 4046, p = 0.6272). These data suggest a lack of sex differences in the basal electrical excitability of histaminergic neurons.

Fig. 4.

Electrical excitability and membrane properties were similar between HDC neurons obtained from male and female mice. (A) Female HDC neurons patched were mapped across the rostral caudal and medial lateral axis of the TMN and grouped into the three subregions of the TMN, demonstrating a similarl distribution to that of the male HDC neurons recorded (as illustrated in Fig. 1E). Note: Image modified from: Mouse Brain: In stereotaxic coordinates (3rd ed.). Franklin, K. B. J., & Paxinos, G. Copyright Elsevier (2007). (B) Membrane potential of HDC neurons did not differ between male and female HDC neurons. (C) The proportion of HDC neurons that were active did not differ between the sexes. (D) The firing frequency of active HDC neurons did not differ between male and female cells. (E) The input resistance of HDC neurons did not differ between the sexes.

Extrinsic spontaneous synaptic transmission

In addition to the intrinsic membrane conductances contributing to HDC neuronal excitability we also investigated the characteristics of extrinsic spontaneous synaptic inputs to HDC neurons in both male and female mice. Both excitatory (EPSC) and inhibitory (IPSC) post synaptic currents were observed in HDC neurons. EPSCs were confirmed to be glutamatergic by their sensitivity to AMPA/kainite & NMDA antagonists (CNQX 10 μM and AP5 50 μM) (3/3 cells) (Fig. 5Ai, ii). IPSCs were confirmed to be GABAergic by their sensitivity to the GABA_A receptor antagonist picrotoxin (3/3 cells) (Fig. 5Bi, ii). The frequency of EPSCs in all HDC neurons was 2.3 ± 0.3 Hz (n = 17) and 2.7 ± 0.4 Hz (n = 22) from male and female mice respectively. The frequency of IPSCs was 0.6 ± 0.3 Hz (n = 15) and 0.5 ± 0.2 Hz (n = 15) from male and female mice. A two-way ANOVA revealed no significant effect of sex (F(1, 65) = 0.27, p = 0.6076), nor a significant interaction (F(1, 65) = 061, p = 0.4388). However, there was a significant effect of PSCs, indicating that HDC neurons receive significantly less IPSCs than EPSCs (F(1, 65) = 29.05, p < 0.0001) (Fig. 5C).

Fig. 5.

Synaptic inputs to HDC neurons. (Ai) Voltage clamp recording demonstrating the sensitivity of excitatory post synaptic currents (EPSCs) arriving at HDC neurons to CNQX and AP5, indicating that they were glutamatergic in nature. (Aii) Enlarged segment of trace above, demonstrating a characteristic EPSC waveform. (Bi) Voltage clamp recording demonstrating the sensitivity of inhibitory post synaptic currents (IPSCs) arriving at HDC neurons to picrotoxin, indicating that they were GABAergic in nature. (Bii) Enlarged segment of trace above, demonstrating a characteristic IPSC waveform. (C) There was no difference in excitatory, or inhibitory post synaptic currents received by HDC neurons of male and female mice. Overall though, significantly more excitatory currents were received by HDC neurons than inhibitory currents. (D) Pharmacological inhibition of synaptic transmission (action potential mediated and glutamatergic) influenced the activity of HDC neurons. (E) Whole-cell current clamp recording demonstrating excitation of a HDC neuron with inhibitors of glutamatergic signaling. // represents a discontinuity in the recording where current voltage relations were performed (not shown).

To determine if these extrinsic events were contributing to the resting membrane potential, we investigated the impact of blocking all action potential mediated synaptic transmission (TTX 500 nM) in a small sample of HDC neurons. Overall, there was no difference in resting membrane potential between control (−47.7 ± 1.3 mV) and TTX (−46.6 ± 1.8 mV) conditions (n = 16, Wilcoxon matched-pairs signed-rank test, W = 55.0, p = 0.1241). However, individual cells could be observed responding to TTX (Fig. 5D). TTX excited 5/16 HDC neurons depolarizing the membrane potential by an average 4.6 ± 0.5 mV, and inhibited 2/16 HDC neurons hyperpolarizing the membrane potential by an average −3.9 ± 0.8 mV. This suggests that in subpopulations of HDC neurons, the resting membrane potential is influenced by extrinsic factors.

Given that HDC neurons in our preparation mainly received glutamatergic inputs, we blocked AMPA/kainite and NMDA receptors (CNQX 10 μM and AP5 50 μM) in an additional group of HDC neurons. Once again, there was no change in membrane potential overall (control: − 48.9 ± 1.1 mV vs CNQX/AP5: −47.6 ± 1.2 mV; n = 15, paired t-test, t(14) = 1.25, p = 0.2333), but individual cells were sensitive to glutamate receptor antagonism (Fig. 5D). Blockade of AMPA/kainite and NMDA receptors excited 8/15 and inhibited 2/15 HDC neurons. The HDC neurons that were excited by glutamate receptor antagonism displayed both a depolarization of the membrane potential and an increase in action potential firing (membrane potential: − 50.0 ± 1.5 mV to −46.2 ± 1.3 mV, n = 8, paired t-test, t(7) = 3.79, p = 0.0068) (action potential firing: 0.22 ± 0.11 Hz to 0.86 ± 0.28 Hz, n = 8, Wilcoxon matched-pairs signed-rank test, W = 21.0, p = 0.0313). This suggests that a proportion of HDC neurons are under some form of pre-synaptic inhibitory tone. Blockade of glutamatergic signaling in the slice ceases this inhibitory influence, subsequently exciting a subpopulation of HDC neurons (Fig. 5D, E).

DISCUSSION

This study provides an assessment of the electrophysiological properties from over 200 genetically identified histaminergic neurons located throughout the TMN in mice. HDC neurons displayed a resting membrane potential of approximately −48 mV, a firing rate of 1.3 Hz in active cells, and pervasive expression of an A-like current (I_A). Minimal differences were observed between HDC neurons from male versus female mice, or between different sub-regions of the TMN. Despite many similarities with the electrophysiological signature of histaminergic neurons originally reported in rats (Haas and Reiner, 1988), we find that genetically identified HDC neurons in mice display greater variability in passive membrane properties, action potential firing, and intrinsic subthreshold active membrane properties. These data suggest that targeting histaminergic neurons based on their electrophysiological signature and anatomical location alone may fail to identify a large proportion of the histaminergic neurons, especially in mice.

Electrical excitability of HDC neurons

In the present study, assessment of HDC neurons resting membrane potential and firing frequency revealed a marked similarity to that originally reported in rats (Haas and Reiner, 1988). However, unlike the spontaneously active rat histaminergic neurons, only half of the HDC neurons in mice spontaneously generated action potentials. The input resistance was also considerably higher with a substantial range. These passive membrane properties and action potential spiking reveal considerable variability in the electrical excitability of genetically identified HDC neurons in mice, compared to that described in rats, and may represent a species difference. Consistent with this idea, the percentage of cells that were active in the present study, is more similar to that reported from the ventral region of the TMN of mice (Fujita et al., 2017).

Histaminergic neuron electrical activity is vastly different depending on behavioral state. These wake active neurons decrease their activity during slow wave sleep and cease firing completely during rapid eye movement sleep (Vanni-Mercier et al., 1984). Histaminergic neurons are also responsive to metabolically relevant signals such as orexin, melanocortin receptor agonism, and insulin induced hypoglycemia (Mochizuki et al., 2000; Miklos and Kovacs, 2003; Michael et al., 2020), and their activity is altered by fasting (Valdes et al., 2010; Wang et al., 2015). Therefore, it is likely that the activity of histaminergic neurons differs depending on the time of day (sleep/wake), the metabolic status of the animal, and preparation medium of the brain slices.

Synaptic inputs to HDC neurons

Additionally, HDC neuron electrical activity is influenced by presynaptic inputs. A significant afferent input to histaminergic neurons originates from the lateral hypothalamus (LH) and ventrolateral preoptic area (VLPO) (Sherin et al., 1996, 1998; Saper et al., 2005). Collectively, these hypothalamic nuclei contribute towards sleep/wake regulation through the ‘sleep switch’, or ‘flip flop model of sleep’ (Saper et al., 2001, 2005). With sleep onset, activation of the GABAergic VLPO neurons increases inhibitory inputs to both the HDC and glutamatergic orexin neurons, therefore favoring a more inhibitory tone to the TMN (Sherin et al., 1996, 1998). We found that all HDC neurons received EPSCs with an average frequency of 2.3–2.7 Hz (male vs female cells). However, HDC neurons received a very limited number of IPSCs, suggesting the TMN histaminergic neurons received a greater excitatory rather than inhibitory tone. Given that the mice in this study were awake prior to the brains being removed, and posterior hypothalamic slices do not contain the VLPO, it is not surprising that HDC neurons received an overall excitatory tone with limited inhibitory currents. Understanding the electrophysiological properties of HDC neurons during different sleep/wake or metabolic states (with fasting or obesity) warrants further investigation.

To determine if excitatory postsynaptic events contributed to the resting membrane potential of HDC neurons we pharmacologically inhibited synaptic transmission. Blockade of both action potential mediated synaptic transmission and excitatory glutamatergic neurotransmission unexpectedly excited a large proportion of the HDC neurons, revealing an inhibitory tone to these neurons in basal conditions. Given that the HDC neurons in our preparation received an extremely limited GABAergic input, and excitation was initiated upon antagonism of the ionotropic glutamate receptors, suggest that this sub-population of HDC neurons are under the influence of neuromodulators other than, or in addition to, traditional fast neurotransmitters. Histaminergic neurons are well known to be regulated by the autoinhibitory histamine 3 receptor (H3R) (Arrang et al., 1983; Garbarg et al., 1989; Morisset et al., 2000; Haas and Panula, 2003). In some subregions of the TMN all HDC neurons express the H3R (Giannoni et al., 2009; De Luca et al., 2016). Blockade of excitatory synaptic transmission may decrease histamine release to other areas of the TMN and thus reduce H3R mediated autoinhibition. Histaminergic neurons can also be inhibited by nociception (Eriksson et al., 2000; Bajic et al., 2004), galanin (Schonrock et al., 1991), adenosine (Oishi et al., 2008), and the dopamine 4 receptor (Yanovsky et al., 2011). While interesting to speculate, future experiments will be required to fully decipher the nature of this inhibitory tone.

Intrinsic subthreshold active membrane properties

HDC neurons displayed differential expression of a range of subthreshold active conductances. The most striking, and most prevalent, was the expression of a transient outward rectification, or A-like current (I_A). Expression of I_A is a defining feature of histaminergic neurons and has been extensively characterized (Haas and Reiner, 1988; Haas et al., 1989; Greene et al., 1990). Such voltage gated potassium channels contribute significantly to the electrical excitability of neurons, regulating the membrane potential response to influence a variety of factors determining firing frequency (Connor and Stevens, 1971; Rudy et al., 2009). The expression of this conductance in the majority of histaminergic neurons, in our and other studies, is therefore consistent between rats and mice (Haas and Reiner, 1988; Fujita et al., 2017), and likely contributes largely to the firing rate of these neurons.

Hyperpolarization-activated cyclic nucleotide-gated non-selective cation conductance (I_h) has also been considered a part of the electrophysiological signature of histaminergic neurons (Haas and Reiner, 1988; Haas et al., 1989) and has been characterized previously (Kamondi and Reiner, 1991). Expression of I_h is evidenced as a depolarizing voltage sag following hyperpolarizing current injections, most prominent in response to larger negative current injections (Robinson and Siegelbaum, 2003; Biel et al., 2009). The presence of this conductance influences the propagation and integration of dendritic currents, and can function to stabilize the membrane potential by dampening both excitatory and inhibitory inputs (Biel et al., 2009). I_h is also considered a pacemaker current as it contributes towards the generation of rhythmic oscillations in some neuronal populations (McCormick and Bal, 1997; Luthi and McCormick, 1998), however, it is not expected to regulate histaminergic neuron activity in this way (Kamondi and Reiner, 1991). Although we observed I_h expression in a large proportion of HDC neurons (57%), it was not as ubiquitously expressed as in the rat (Haas and Reiner, 1988). Our results are consistent with others using the same HDC-cre mouse model (De Luca et al., 2016), and suggest this conductance is not a reliable indicator of histaminergic neurons in mice.

We also observed expression of a classical anomalous inward rectifier (K_IR) in a limited number of HDC neurons. Although K_IR has not been reported in any of the previous studies investigating subthreshold active membrane properties of HDC neurons (Haas and Reiner, 1988; Haas et al., 1989; Greene et al., 1990; Kamondi and Reiner, 1991; Fujita et al., 2017) the presence of such a conductance has been implicated in the mechanism of action of nociception on histaminergic neurons (Eriksson et al., 2000; Bajic et al., 2004). In central neurons, K_IR channels are important regulators of neuronal excitability and are critical for the maintenance of the resting membrane potential (Hibino et al., 2010). Functionally, K_IR in HDC neurons may help to maintain the membrane potential within a limited range and may protect the cells against becoming overly hyperpolarized. Given that this conductance was only expressed in such a small proportion of HDC neurons, it may represent a functionally distinct class of HDC neurons.

Overall, while the active conductance profile of genetically identified histaminergic neurons was in many cases congruent with that originally described in rats (Haas and Reiner, 1988), we report a greater variability in expression of such subthreshold active membrane properties in mice. Strictly adhering to the reported electrophysiological signature of histaminergic neurons based on expression of I_A and I_h fails to account for up to 45% of the HDC cells in mice.

Heterogeneity of HDC neurons

Previous studies have defined the electrophysiological properties of HDC neurons, however, all reports were based on HDC neurons in the ventral (core) region of the TMN (Haas and Reiner, 1988; Haas et al., 1989; Fujita et al., 2017). Given that HDC neurons are a heterogenous population, it is possible that their electrophysiological properties may differ across the different subregions of the TMN. In our study, we found limited to no evidence of electrophysiological diversity across the ventral, dorsal and bridge regions of the TMN, arguing against subpopulations based on the subregion of the TMN. Anatomical tracing studies have also demonstrated a lack of a topographical organization of histaminergic neurons (Kohler et al., 1985; Ericson et al., 1987; Inagaki et al., 1990). Nevertheless, we observed variability in the electrophysiological properties of HDC neurons. Sub-populations of HDC neurons have previously been demonstrated based on their sensitivity to GABA, CB1R agonists, acidification, insulin-induced hypoglycemia, MC4R agonism, glycine, and H3R agonism (Blandina et al., 2012; Michael et al., 2020). When considered along with anatomical tracing studies, HDC neuron’s differential responsiveness to various stimuli, the diverse physiological roles they regulate, and the variability in electrical excitability of HDC neurons, suggests that histaminergic neurons form intermingled subpopulations of neurons. Moreover, these findings imply that independent functions may be regulated by subsets of histaminergic neurons.

Sex differences

Previous studies have suggested that the central histaminergic system displays sex differences. This includes enhanced histamine metabolism, and greater expression of histaminergic receptors (H1R and H2Rs) in females (Prell et al., 1991; Ghi et al., 1999; Yoshizawa et al., 2009). Sex differences have also been reported in the age related decline in histamine release (Ferretti et al., 1998), and females have been shown to be more sensitive to the effects of H1R antagonism and dietary histidine (the precursor for histamine) supplementation (Easton et al., 2004; Kasaoka et al., 2005). In contrast to these observations, we report a lack of sex differences in the electrophysiological properties of histaminergic neurons. All measures of neuronal electrical excitability examined were comparable between male and female HDC neurons. This suggests that any sex differences in central histaminergic signaling likely exist downstream of the neurons (e.g. histamine receptors, or its metabolism), or in the responsiveness of these neurons to external inputs (e.g. neurotransmitters), rather than due to any differences in basal activity of the neurons themselves. However, one factor potentially limiting our ability to detect sex differences is that the female mice used were not ovariectomized, and the stage of the estrus cycle was not noted.

In summary, we report the electrophysiological profiles of genetically identified histaminergic neurons across the whole TMN region. Our data suggest that the electrophysiological properties and presence of certain currents (electrophysiological signature), may not as reliably identify histaminergic neurons in the mouse, compared to that reported in the rat. Sex differences in HDC neuron electrical excitability were not apparent, and we did not find differences based on the subregions of the TMN. However, variability in HDC neurons electrical excitability, including expression of subthreshold active membrane properties, support the possibility of functionally distinct populations of HDC neurons, albeit without a distinct topographical organization. Histaminergic neurons are involved in the regulation of numerous physiological functions including sleep and wakefulness, energy homeostasis, nociception, temperature regulation and locomotion. Recognizing fundamental differences in HDC neurons may be relevant for the functional outputs of such neurons.

Abbreviations:

HDC

histidine decarboxylase

I_h

hyperpolarisation-activated time-dependent inward rectification

I_IR

instantaneous inward rectification

I_A

A-like current (transient outward rectification)

LH

lateral hypothalamus

K_IR

inwardly rectifying potassium channel

TMN

tuberomammillary nucleus

TTX

tetrodotoxin

VLPO

ventrolateral preoptic area