Visual Abstract
Visual abstract was created in BioRender. Michael, N. (2026) https://BioRender.com/92wahel.
Abstract
Galanin-expressing neurons in the ventrolateral preoptic area (VLPOgalanin) are active during sleep and play an important role in regulating non-rapid eye movement (NREM) sleep. It is generally believed that VLPOgalanin neurons promote sleep via inhibitory actions in arousal-promoting regions of the brain. Histaminergic neurons are a population of wake-active neurons that receive strong projections from the sleep-active VLPOgalanin neurons. However, the ability of galanin to influence the activity of histaminergic neurons has received limited attention. Here, using whole-cell patch-clamp electrophysiological recordings from genetically identified histaminergic neurons in male mice, we explore the mechanisms by which galanin influences histaminergic neuron electrical excitability. Our results reveal that galanin is a powerful inhibitor of histaminergic neuron activity and demonstrate that the inhibitory effects of galanin are mediated by galanin receptor 1 (GALR1) and the subsequent opening of G-protein-coupled inwardly rectifying (GIRK) and large conductance calcium-activated potassium (BK) channels. Furthermore, we identify that histaminergic neurons highly express Galr1 mRNA and show that the GALR1-mediated hyperpolarization of histaminergic neurons is largely independent of action potential-dependent synaptic transmission or fast excitatory or inhibitory neurotransmitters. Together, these results suggest that direct postsynaptic activation of GALR1 expressed on histaminergic neurons mediates the inhibitory effects of galanin on these neurons. This data also supports the notion that the sleep-promoting effects of VLPOgalanin neuron activation may occur via the ability of galanin to inhibit the arousal-promoting histaminergic neurons.
Significance Statement
The “flip-flop switch” model of sleep and wakefulness proposes that sleep-active galanin-expressing neurons in the ventrolateral preoptic area (VLPO) promote sleep via inhibitory actions on arousal-promoting neurons, such as the histaminergic neurons. However, the molecular mechanisms surrounding this theory are lacking. Here, we report that histaminergic neurons are strongly inhibited by galanin, an effect that occurs via galanin receptor 1 (GALR1)-mediated opening of potassium channels. The GALR1-induced inhibition persisted in blockers of synaptic transmission and Galr1 mRNA was expressed in histaminergic neurons. Together, these results suggest that galanin inhibits histaminergic neurons via GALR1 expressed on histaminergic neurons and supports the notion that galanin-expressing VLPO neurons could silence the wake-active histaminergic neurons to promote sleep.
Introduction
Sleep is a highly conserved behavioral state regulated by complex, interconnected, neural circuitry. The hypothalamus is one important brain structure recognized for regulating sleep and wakefulness (Saper et al., 2005). It contains sleep-promoting and wake-promoting subregions, and disrupted neuronal function in these regions leads to alterations in the sleep/wake cycle (von Economo, 1930; Chemelli et al., 1999; Lin et al., 1999; Saper et al., 2001; Anaclet et al., 2009). At present, the exact mechanisms regulating sleep and wakefulness are incompletely defined, and additional work characterizing these circuits may provide important insights for the development of novel pharmacological strategies for the treatment of sleep disorders (Mignot et al., 2002).
Thirty years ago, seminal work identified a population of neurons located in the ventrolateral preoptic area (VLPO) that were activated by sleep (Sherin et al., 1996). These neurons were found to send monosynaptic projections to the tuberomammillary nucleus (TMN), an important component of the ascending arousal system (Sherin et al., 1996). Subsequent work from the same group demonstrated that these sleep-active VLPO neurons release the inhibitory neurotransmitters GABA and galanin, suggesting that they may have the ability to silence arousal-promoting neurons (Sherin et al., 1998). Since this time, the understanding of the role of the VLPO in promoting sleep has been extensively recognized; however, the molecular mechanisms allowing VLPO neurons to inhibit wake-active neuronal populations have not been fully characterized.
Within the VLPO, galanin is considered the most robust cellular marker identifying sleep-active neurons (Gaus et al., 2002). Optogenetic stimulation of VLPOgalanin neurons promotes sleep and increases non-rapid eye movement (NREM) sleep (Kroeger et al., 2018). Similarly, loss of galanin neurons in this region is associated with altered sleep homeostasis (Ma et al., 2019). It is generally assumed that some of the sleep-promoting effects of VLPOgalanin neuron activation are due to their ability to inhibit the wake-promoting histaminergic neurons of the TMN (Sherin et al., 1996; Saper et al., 2005). However, the mechanisms by which galanin influences histaminergic neuron activity remain poorly defined.
Histamine release experiments have demonstrated that galanin suppresses potassium-evoked histamine release in the hypothalamus and the hippocampus of rats (Arrang et al., 1991). Such effects were shown to be insensitive to tetrodotoxin (TTX), suggesting that galanin receptors are likely expressed on histaminergic neurons themselves (Arrang et al., 1991). Electrophysiological experiments performed on putative histaminergic neurons have also demonstrated that galanin reduces the firing frequency of TMN neurons (Schonrock et al., 1991); however, the mechanisms behind this effect have not been explored. Importantly, the galanin receptor 1 (GALR1) is highly expressed in the hypothalamus (Parker et al., 1995; Waters and Krause, 2000), including in the TMN (Sun et al., 2004), suggesting that galanin may influence histaminergic neuron excitability via activation of GALR1.
In this study, we aimed to characterize the effects of galanin on the electrical excitability of genetically identified histaminergic neurons and aimed to identify the ionic and molecular mechanisms by which galanin inhibits histaminergic neuron activity. Our results indicate that galanin strongly inhibits histaminergic neurons via direct action at GALR1 expressed on histaminergic neurons.
Materials and Methods
Animals
All animal work and experimentation adhered to the standards outlined in the Canadian Guide for the Care and Use of Laboratory Animals and received prior approval from the Animal Care Committee of Université Laval (CPAUL). Male mice were housed in ventilated cages, with a 12 h light/dark cycle (lights on from 0600 to 1800) and constant ambient temperature of 23 ± 1°C. Mice were provided with ad libitum access to water and a standard chow diet (Harlan Teklad, 2918).
To target genetically identified histaminergic neurons for electrophysiological recordings, mice expressing a Cre recombinase under the control of the histidine decarboxylase (Hdc) promoter (Yanovsky et al., 2012; Walker et al., 2013) were crossed to a tdTomato reporter mouse (Ai14, JAX stock # 007914) and were maintained on a C57BL/6J background. This allowed fluorescent labeling of all cells expressing Hdc, the sole enzyme required for histamine synthesis. Validation of this Hdc-Cre mouse has demonstrated high specificity, with 91–98% of the Cre (tdTomato) labelled cells in the different TMN subregions coexpressing Hdc immunoreactivity (Walker et al., 2013). C57BL/6J (JAX stock # 000664) mice were obtained from The Jackson Laboratory (Maine) for the fluorescent in situ hybridization experiments.
Electrophysiology
Slice preparation
Animals were euthanized between 8 and 12 weeks of age, ∼3 h after the beginning of the light period for whole-cell patch-clamp electrophysiology experiments as described previously (Michael et al., 2020a,b). Briefly, animals were anesthetized using isoflurane, decapitated, and their brains rapidly removed and maintained in a modified, sucrose-based, ice-cold artificial cerebrospinal fluid (aCSF). The aCSF contained the following (in mM): 213 sucrose; 2.5 KCl; 5 MgCl2; 1 CaCl2, 1 NaH2PO4, 26 NaHCO3, and 10 d-glucose and was continuously perfused with 95% O2 and 5% CO2.
Coronal hypothalamic sections (250 µm) containing the TMN were prepared using a Leica VT1000S Vibratome and incubated (∼32°C) for at least 40 min in a standard aCSF containing the following (in mM): 126 NaCl; 2.8 KCl; 2.5 CaCl2, 1.25 NaH2PO4, 26 NaHCO3, 1.2 MgSO4; and 10 d-glucose. Once slices were transferred to the recording chamber, they were continuously perfused with a reduced d-glucose concentration (5 mM) version of the standard aCSF, heated to ∼32°C via an in-line heater (Warner Instruments).
Whole-cell patch-clamp recordings
Whole-cell patch-clamp recordings were established from fluorescently labeled HDC neurons using an Axio Examiner fixed stage microscope (Zeiss) fitted with infrared videomicroscopy and fluorescence. Current-clamp recordings were performed using an Axopatch 200B amplifier and Digidata 1550B digitizer (Molecular Devices). Data was filtered at 4 kHz and stored on a personal computer for offline analysis with pClamp10 software (Molecular Devices). Patch pipettes were pulled from thin-walled glass capillaries (TW150F-4, World Precision Instruments) providing resistances of ∼5 MΩ when filled with an intracellular solution containing the following (in mM): 120 K-gluconate, 10 KCl, 1 NaCl, 1 MgCl2, 1 CaCl2, 5 EGTA, 10 HEPES, and 2 Mg2ATP, adjusted for pH and osmolality with KOH and sucrose. All neurons were allowed time to settle before standard measures of electrical excitability were assessed.
Drugs
All drugs were prepared as concentrated stock solutions in distilled water, except for picrotoxin which was dissolved in ethanol, aliquoted, and stored <4°C. Immediately before use, stocks were diluted to the required concentration in aCSF and delivered to the recording bath by a peristaltic perfusion system (PPS2, Multi Channel Systems). Galanin (1–29; rat, mouse) and selective galanin receptor agonists targeting the GALR1, GALR2, and GALR3 (M617, M1145, and Spexin, respectively) were obtained from Tocris (Bio-Techne). Galanin was used at 100 nM, a concentration previously shown to reliably produce inhibitory effects in the TMN and in other neuronal populations (Schonrock et al., 1991; Ma et al., 2001). For mechanistic studies, receptor-selective agonists were used at concentrations >100-fold above their Ki values to ensure penetration of the brain slice and receptor activation. While M617 is 25 times more selective for GALR1 than GALR2, its use at 500 nM to target GALR1 may also activate GALR2 (Ki 5.71 nM; Lundstrom et al., 2005). However, the use of M1145 at 200 nM to target GALR2, and Spexin at 1 µM to target GALR3, are well below the Ki values of GALR1 (587 nM and no activity, respectively; Runesson et al., 2009; Kim et al., 2014), ensuring an ability to differentiate the contribution of each galanin receptor subtype to the responses observed. On all occasions, a total of 20 ml of galanin (or galanin receptor agonist) was delivered at a rate of ∼4 ml/min. Tetrodotoxin (TTX; from Tocris and Hello Bio) was used to synaptically isolate HDC neurons while picrotoxin, CNQX disodium salt, and d-AP5 were used to block GABAergic and glutamatergic transmission (also from Tocris). Tetraethylammonium chloride (TEA) and tertiapin-Q (from Tocris), along with tolbutamide and barium chloride (BaCl2; from Sigma), were used to explore the involvement of potassium channels in the responses observed.
Statistical analysis
All data are presented as mean ± SEM. Comparisons between conditions were assessed by two-tailed paired t tests generated using GraphPad Prism 10. Parametric statistics were used for data adopting a Gaussian distribution and nonparametric statistics were used when this assumption was not met.
Fluorescent in situ hybridization
Brain preparation
Fluorescent in situ hybridization was used to explore and quantify Galr1 expression in histaminergic (Hdc-expressing) neurons. C57BL/6J mice were deeply anesthetized with isoflurane and transcardially perfused with ice-cold saline (0.9%), followed by cold paraformaldehyde (PFA, 4%) for 5 min. The brains were extracted and maintained in 4% PFA for 7 d, after which they were transferred to a solution containing PFA (4%) and sucrose (10%) for 1 d, and transferred to a solution containing PFA (4%) and sucrose (20%) for 1 d until they sank to the bottom of the vial. The brains were frozen in crushed dry ice for 10 min and were then maintained at −80°C until they were cut at 25 µm with a sliding microtome (HM 440E, Microm Laborgeräte). The coronal brain slices were stored at −20°C in a cryoprotectant solution containing sodium phosphate buffer (50 mM), ethylene glycol (30%), and glycerol (20%). Posterior hypothalamic slices containing the TMN (approx. bregma −2.06; −2.46 and −2.80 mm) were selected and washed three times for 7 min with PBS. The slices were mounted on Superfrost Plus microscope slides (Thermo Fisher Scientific) and dried overnight under a chemical hood.
Hybridization
The fluorescent in situ hybridization experiments were performed using the RNAscope Multiplex Fluorescent Reagent Kit v2 (Advanced Cell Diagnostics, ACD) as described previously (Srour et al., 2023). Briefly, slides were washed for 5 min in PBS and incubated at 60°C for 30 min in an ACD HybEz II hybridization system. The samples were then postfixed in PFA (4%) for 15 min, dehydrated in EtOH (50%, 70%, and 2 × 100%) 5 min each, and treated with H2O2 for 10 min to block endogenous peroxidase activity. Slides were consecutively immerged in 1× target retrieval reagent (ACD) for 5 min in a steamer, rinsed with distilled water, and dried out in 100% EtOH for 3 min. An ImmEdge pen (Vector Laboratories) was used to draw a hydrophobic barrier around tissue sections and samples were incubated with Protease III (ACD) for 30 min at 40°C. After this step, slides were rinsed with distilled water and incubated with a probe mix containing Mm-Hdc-C2 (490471-C2) and Mm-Galr1 (448821) or Mm-Galr2 (448831) or Mm-Galr3 (448841) for 2 h at 40°C. Then, at room temperature (RT), the slides were cleaned with 1× wash buffer (ACD) and kept overnight in 5× saline sodium citrate buffer (SSC). On the second experimental day, all incubation steps were performed at 40°C. Briefly, slides were incubated with AMP1 reagent for 30 min, AMP2 for 30 min, AMP3 for 15 min, HRP-C1 for 15 min, Opal 520 fluorophore for 30 min (1:1,500; Akoya Biosciences), HRP blocker for 15 min, HRP-C2 for 15 min, and Opal 620 fluorophore for 30 min (1:1,500; Akoya Biosciences). Finally, slides were stained 30 s with DAPI at RT and coverslipped with the ProLong Gold Antifade Mountant (Thermo Fisher Scientific).
Analysis of the hybridization signals
Brain slices were digitized using an AxioScan.Z1 slide scanning microscope fitted with a Colibri 7 LED light source and filter sets (Zeiss). Manual cell counting of histaminergic (Hdc-expressing) neurons that were positive for Galr1, Galr2, or Galr3 was performed by two independent investigators. A total of nine brain slices from three male mice (three slices/mouse) were used for each RNAscope experiment. Using ZEN Blue software (Zeiss), colocalization of Galr1 was considered positive when five or more Galr1 probe signals (dots) per histaminergic neuron were detected. The data are presented as the percentage of Galr1 positive Hdc-expressing cells.
Results
Galanin strongly inhibits histaminergic neurons
While one previous study reported that galanin inhibits the firing frequency of putative histaminergic neurons, identified based on their electrophysiological signature (active cells displaying inward rectification and transient outward rectification; Schonrock et al., 1991), we have recently demonstrated greater diversity in the electrophysiological properties of histaminergic neurons than previously thought (Michael et al., 2020b). Therefore, we characterized the effects of galanin on genetically identified histaminergic neurons using a Hdc-Cre mouse model (Walker et al., 2013). Whole-cell current-clamp recordings were established from histaminergic neurons (expressing Hdc) located across all subregions of the TMN, as done previously (Michael et al., 2020a,b). Bath application of galanin (100 nM) strongly inhibited 81% (21/26) of histaminergic neurons (Fig. 1Ai–iii).
Galanin-induced inhibition of histaminergic neurons. Ai, Whole-cell current-clamp recording demonstrating the inhibitory effect of galanin on histaminergic neuron activity, including a return to baseline activity levels during the wash period (return to aCSF). The small gap in the trace represents the time when current–voltage relationships were performed. Aii, Hyperpolarization of the membrane potential in histaminergic neurons that were inhibited by galanin. Aiii, Decrease in the firing rate of histaminergic neurons that were inhibited by galanin. Bi, Input resistance decreased in the majority (16/21) of the galanin-inhibited histaminergic neurons. Bii, Current–voltage (IV) relationships of a galanin-inhibited histaminergic neuron that displayed a decrease in input resistance. Membrane potential changes are in response to successive positive and negative current injections in steps of 22 or 20 pA, respectively. ***p < 0.001 and ****p < 0.0001.
The galanin-induced inhibition was associated with a significant hyperpolarization of the membrane potential from −46.4 ± 1.3 mV in control conditions to −51.0 ± 1.4 mV following the administration of galanin (paired t test, t(20) = 8.59, n = 21, p < 0.0001; Fig. 1Aii). The inhibitory effect of galanin was also characterized by a significant decrease in firing frequency (control: 1.37 ± 0.41 Hz vs 0.25 ± 0.15 Hz in the presence of galanin; Wilcoxon matched-pairs signed-rank test, W = −78, n = 21, p = 0.0005; Fig. 1Aiii). The onset of galanin's inhibitory effects was observed within 2 min (1.97 ± 0.20 min), with maximal effects typically observed at approximately 5 and a half minutes (5.52 ± 0.34 min) after being exposed to galanin. All histaminergic neurons that were inhibited by galanin displayed signs of recovery during the wash period (return to aCSF; Fig. 1Ai), of ∼7–9 min in most cases. Galanin-inhibited histaminergic neurons (n = 21) returned to a resting membrane potential of −47.3 (±1.3) mV and a firing frequency of 0.85 (±0.33) Hz during the wash, which was similar to baseline conditions.
The galanin-induced inhibition was associated with a decrease in input resistance in the majority (16/21) of the cells (Fig. 1Bi,ii). In these cells, input resistance decreased from 810 ± 129 MΩ in control conditions to 663 ± 113 MΩ in the presence of galanin (Wilcoxon matched-pairs signed-rank test, W = −136, n = 16, p < 0.0001) and was associated with a mean reversal potential of −74.1 ± 6.9 mV which is similar to galanin-induced effects reported in other hypothalamic neurons (Papas and Bourque, 1997). In the remaining galanin-inhibited histaminergic neurons, the galanin-induced inhibition was associated with a significant increase in input resistance (control: 410 ± 57 MΩ vs galanin: 479 ± 58 MΩ; paired t test, t(4) = 4.851, n = 5, p = 0.0083) and a reversal potential of −11.8 ± 10.2 mV.
In the remaining 19% (5/26) of histaminergic neurons that did not respond to galanin, there was no difference in resting membrane potential (control: −46.9 ± 2.4 mV vs galanin: −47.0 ± 2.6 mV; paired t test, t(4) = 0.1693, n = 5, p = 0.8738), firing frequency (control: 1.08 ± 0.77 Hz vs galanin: 0.75 ± 0.47 Hz; paired t test, t(4) = 1.023, n = 5, p = 0.3640), or input resistance (control: 729 ± 143 MΩ vs galanin: 709 ± 134 MΩ; paired t test, t(4) = 0.7172, n = 5, p = 0.5129). Together, these results suggest that multiple mechanisms may mediate the galanin-induced inhibition of histaminergic neurons.
Histaminergic neurons strongly express GALR1 mRNA
As galanin can bind to three types of galanin receptors, we aimed to determine which receptor subtype was mediating the galanin-induced inhibition of histaminergic neurons. As previous work suggested that GALR1 is expressed in the TMN (Sun et al., 2004), we examined if histaminergic neurons express Galr1 using RNAscope fluorescent in situ hybridization. Galr1 was identified in the posterior hypothalamus, consistent with previous reports (Sun et al., 2004; Fig. 2Ai). Moreover, Galr1 was strongly expressed in the TMN where it was coexpressed in Hdc-expressing (histaminergic) neurons (Fig. 2Ai,ii). Coexpression analysis revealed that 77.7% (±1.8) of Hdc cells coexpressed Galr1 demonstrating that histaminergic neurons highly express GALR1 (Fig. 2Aiii). This also suggested that direct activation of GALR1 may mediate the galanin-induced inhibition of histaminergic neurons.
GALR1 is the dominant galanin receptor subtype expressed in histaminergic neurons. Ai, Representative fluorescent in situ hybridization image of Hdc and Galr1 expression within the posterior hypothalamus showing coexpression of Hdc and Galr1. Two cells identified by white arrows are shown in higher magnification in Aii, demonstrating the high expression of Galr1 in histaminergic neurons. Aiii, 77.7% of histaminergic neurons expressed Galr1. Bi, Representative fluorescent in situ hybridization image of Hdc and Galr2 expression within the posterior hypothalamus showing almost no coexpression of Hdc and Galr2. Two cells identified by white arrows are shown in higher magnification in Bii, demonstrating that histaminergic neurons rarely express Galr2. Biii, 0.3% of histaminergic neurons expressed Galr2. Ci, Representative fluorescent in situ hybridization image of Hdc and Galr3 expression within the posterior hypothalamus showing minimal coexpression of Hdc and Galr3. Two cells identified by white arrows are shown in higher magnification in Cii, demonstrating the expression of Galr3 in some histaminergic neurons. Ciii, 32.3% of histaminergic neurons expressed Galr3 (n = 3 mice for all experiments). Each dot represents an average of two counters for one coronal brain slice.
As Galr2 and Galr3 are also expressed within the hypothalamus (Fathi et al., 1997; Waters and Krause, 2000), we also examined if histaminergic neurons express these galanin receptor subtypes. Consistent with previous reports, Galr2 was observed within the hypothalamus; however, it did not colocalize (0.3 ± 0.2%) with Hdc-expressing neurons (Fig. 2Bi–iii). In contrast, Galr3 did colocalize with Hdc-expressing neurons (Fig. 2Ci,ii), although to a much lesser extent than Galr1. Coexpression analysis demonstrated that 32.3% (±2.6) of Hdc cells coexpressed Galr3 (Fig. 2Ciii). These results raise the potential for galanin to also influence histaminergic neuron activity via GALR3.
GALR1 agonism mimics the inhibitory effects of galanin on the activity of histaminergic neurons
As GALR1 was strongly expressed in histaminergic neurons, we used the selective GALR1 agonist M617 to investigate its involvement in the galanin-induced inhibition of histaminergic neurons. Bath application of M617 (500 nM) mimicked the galanin-induced inhibitory effect with 73% (16/22) of histaminergic neurons being inhibited by the GALR1 agonist (Fig. 3Ai–iii). The M617-induced inhibition was associated with a significant hyperpolarization of the membrane potential from −44.7 ± 1.4 mV in control conditions to −48.6 ± 1.4 mV in M617 (paired t test, t(15) = 11.90, n = 16, p < 0.0001; Fig. 3Aii). The inhibitory effect of M617 was also characterized by a significant decrease in firing frequency (control: 0.79 ± 0.28 Hz vs M617: 0.16 ± 0.09 Hz; Wilcoxon matched-pairs signed-rank test, W = −45, n = 16, p = 0.0039; Fig. 3Aiii). While the inhibitory effect of M617 on histaminergic neurons largely replicated what was seen with galanin, it had a slightly more rapid onset, occurring ∼1 min (1.11 ± 0.11 min) after entering the slice chamber and maximal effects were observed after ∼3 min (3.27 ± 0.27 min). All but one of the histaminergic neurons that were inhibited by M617 displayed signs of recovery during the wash (return to aCSF) of ∼5.5–7 min (Fig. 3Ai), and one of the cells could not be assessed as it died during the wash period. M617-inhibited histaminergic neurons (n = 15) returned to a resting membrane potential of −45.6 (±1.4) mV and a firing frequency of 0.65 (±0.22) Hz during the wash, which was similar to that observed in baseline conditions.
Galanin receptor 1 agonist (M617)-induced inhibition of histaminergic neurons. Ai, Whole-cell current-clamp recording demonstrating the inhibitory effect of M617 on histaminergic neuron activity, including a return of activity during the wash period (return to aCSF) to levels like that observed in baseline conditions. The small gap in the trace represents the time when current–voltage relationships were performed. Aii, Hyperpolarization of the membrane potential in histaminergic neurons that were inhibited by M617. Aiii, Decrease in the firing rate of histaminergic neurons that were inhibited by M617. Bi, Input resistance decreased in the majority (12/16) of the M617-inhibited histaminergic neurons. Bii, Current–voltage (I–V) relationships of a M617-inhibited histaminergic neuron that displayed a decrease in input resistance. Membrane potential changes are in response to successive positive and negative current injections in steps of 34pA. **p < 0.05 and ****p < 0.0001.
Like galanin, the M617-induced inhibition was associated with a decrease in input resistance in the majority (12/16) of the cells (Fig. 3Bi,ii). In these cells, input resistance decreased from 681 ± 69 MΩ in control conditions to 603 ± 65 MΩ in the presence of M617 (paired t test, t(11) = 9.712, n = 12, p < 0.0001) and was associated with a mean reversal potential of −76.9 ± 6.0 mV. In the remaining M617-inhibited histaminergic neurons, the M617-induced inhibition was associated with a significant increase in input resistance (control: 591 ± 60 MΩ vs M617: 798 ± 119 MΩ; paired t test, t(3) = 3.307, n = 4, p = 0.0455) and a reversal potential of −29.7 ± 2.1 mV. In the six histaminergic neurons that were not inhibited by M617, one displayed an excitatory response and the remaining five displayed no significant differences in resting membrane potential (paired t test, t(4) = 1.004, n = 5, p = 0.3720), firing frequency (paired t test, t(4) = 1.259, n = 5, p = 0.2765), or input resistance (paired t test, t(4) = 0.5401, n = 5, p = 0.6178) between control and M617 conditions. Together, these results demonstrate that GALR1 activation strongly inhibits the electrical excitability of histaminergic neurons in a manner similar to galanin.
GALR2 and GALR3 agonism do not influence the activity of histaminergic neurons
As galanin 2 receptor (Galr2) and galanin 3 receptor (Galr3) mRNA are also expressed within the hypothalamus (Fathi et al., 1997; Waters and Krause, 2000), and as our results suggested some histaminergic neurons express Galr3, we explored whether galanin could potentially influence histaminergic neuron activity via these galanin receptor subtypes. Bath application of M1145 (200 nM), a selective GALR2 agonist failed to influence histaminergic neuron activity (Fig. 4A) in any of the neurons (0/12) tested. Bath application of M1145 had no effect on resting membrane potential (control: −47.8 ± 1.9 mV vs M1145: −47.8 ± 2.0 mV; paired t test, t(11) = 0.04736, n = 12, p = 0.9631), firing frequency (control: 0.84 ± 0.23 Hz vs M1145: 0.89 ± 0.23 Hz; paired t test, t(11) = 0.8840, n = 12, p = 0.3956), or input resistance (control: 683 ± 53 MΩ vs M1145: 702 ± 51 MΩ; Wilcoxon matched-pairs signed-rank test, W = 39, n = 12, p = 0.1333).
Galanin receptors 2 and 3 do not influence histaminergic neuron activity. A, Whole-cell current-clamp recording demonstrating no change in histaminergic neuron activity following exposure to M1145, a selective GALR2 agonist. B, Whole-cell current-clamp recording demonstrating no change in histaminergic neuron activity following exposure to Spexin, a potent GALR2/GALR3 agonist.
Seeing as GALR2 activation had no impact on histaminergic neuron activity, we used Spexin, a potent GALR2/GALR3 agonist, to determine any potential contribution of GALR3 to the galanin-induced inhibition we had observed. Bath application of Spexin (1 µM) failed to influence histaminergic neuron activity (Fig. 4B) in any of the neurons (0/14) tested. Histaminergic neurons displayed no change in resting membrane potential (control: −44.5 ± 1.6 mV vs Spexin: −44.3 ± 1.6 mV; paired t test, t(13) = 0.4327, n = 14, p = 0.6723), firing frequency (control: 1.35 ± 0.36 Hz vs Spexin: 1.37 ± 0.40 Hz; Wilcoxon matched-pairs signed-rank test, W = 5.000, n = 14, p = 0.8203), or input resistance (control: 563 ± 38 MΩ vs Spexin: 575 ± 38 MΩ; paired t test, t(13) = 0.6527, n = 14, p = 0.5253) in response to Spexin. These data suggest that neither GALR2 nor GALR3 influences histaminergic neuron activity and that the galanin-induced inhibition of histaminergic neuron electrical excitability is entirely mediated by GALR1.
GALR1 inhibits histaminergic neurons via the opening of multiple types of potassium channels
The GALR1 is a G-protein-coupled receptor (GPCR) that signals through the Gαi/o pathway (Habert-Ortoli et al., 1994; Lang et al., 2015). Its activation is associated with the inhibition of adenylate cyclase (AC) and the opening of G-protein-gated inwardly rectifying potassium (GIRK) channels (Habert-Ortoli et al., 1994; Smith et al., 1998). In our experiments, the galanin and M617-induced inhibition of histaminergic neurons was largely associated with a decrease in input resistance associated with a hyperpolarized reversal potential. To confirm if the GALR1-induced inhibition of histaminergic neurons occurred via opening of inwardly rectifying potassium (Kir) channels, we utilized the Kir channel blocker Ba2+ (Hibino et al., 2010). As expected, prior administration of BaCl2 (100 µM) completely blocked the GALR1-induced inhibition of histaminergic neurons. Under these conditions, none (0/10) of the histaminergic neurons responded to M617 (500 nM; Fig. 5A). In the presence of BaCl2, M617 had no effect on membrane potential (BaCl2: −44.1 ± 1.1 mV vs M617 in BaCl2: −43.6 ± 1.3 mV; paired t test, t(9) = 1.876, n = 10, p = 0.0934), firing frequency (BaCl2: 2.42 ± 0.36 Hz vs M617 in BaCl2: 2.36 ± 0.36 Hz; paired t test, t(9) = 0.8730, n = 10, p = 0.4054), or input resistance (BaCl2: 700 ± 48 MΩ vs M617 in BaCl2: 700 ± 58 MΩ), paired t test, t(9) = 0.01851, n = 10, p = 0.9856). These results suggest that galanin inhibits histaminergic neurons via opening of Ba2+ sensitive potassium channels.
Blockade of G-protein-coupled inwardly-rectifying potassium channels or calcium-activated potassium channels reduces the galanin 1 receptor-induced inhibition of histaminergic neurons. A, Whole-cell current-clamp recording demonstrating that the galanin receptor 1 agonist M617 did not influence histaminergic neuron activity in the presence of BaCl2. B, Whole-cell current-clamp recording demonstrating that the GIRK channel blocker tertiapin-Q reduced the inhibitory effect of M617 on histaminergic neurons. C, Whole-cell current-clamp recording demonstrating that the M617-induced inhibition of histaminergic neurons persisted after the blockade of KATP channels with tolbutamide. D, Whole-cell current-clamp recording demonstrating that the BK channel blocker TEA reduced the inhibitory effect of M617 on histaminergic neurons. E, Comparisons of the response of histaminergic neurons to M617 in control conditions and after blockade of the different potassium channels tested. F, The M617-induced membrane hyperpolarization of histaminergic neurons was reduced by blockade of GIRK or BK channels. TQ, tertiapin-Q and Tolb, tolbutamide.
As Ba2+ blocks all Kir channels and can block other types of potassium channels, we investigated the effects of other pharmacological agents that target specific types of Kir channels. First, we used tertiapin-Q (100 nM), a high affinity blocker of GIRK channels (Kir3.1/3.4). In the presence of tertiapin-Q, the proportion of histaminergic neurons that were unresponsive to M617 increased to 36% (4/11 cells), and the inhibitory effects in those cells that were inhibited by M617 (64%, 7/11 cells) appeared reduced (Fig. 5B). While the inhibitory effects of M617 in the presence of tertiapin-Q were still associated with a significant hyperpolarization of the membrane potential (tertiapin-Q: −41.3 ± 1.0 mV vs M617 in tertiapin-Q: −43.0 ± 1.3 mV; paired t test, t(6) = 2.700, n = 7, p = 0.0356), M617 no longer completely silenced histaminergic neurons (Fig. 5B) and was only associated with a trend toward a decrease in firing (tertiapin-Q: 0.62 ± 0.19 Hz vs M617 in tertiapin-Q: 0.17 ± 0.10 Hz; Wilcoxon matched-pairs signed-rank test, W = −15.00, n = 7, p = 0.0625). These results suggest that GIRK channels contribute to the inhibitory effects of M617 on histaminergic neurons but also suggest that other mechanisms may be involved.
In the pancreas, galanin-induced inhibition is mediated by ATP-sensitive potassium channels (KATP; de Weille et al., 1988; Dunne et al., 1989). These channels, formed by Kir6.1/6.2 subunits, are also expressed in the hypothalamus. To test their involvement in galanin receptor-mediated inhibition, we used the KATP channel blocker tolbutamide (200 µM) during recordings. In the presence of tolbutamide, the M617-induced inhibition persisted (Fig. 5C). Under these conditions, M617 inhibited 70% (7/10) of histaminergic neurons, like when M617 was used alone. This inhibition was associated with a significant hyperpolarization of the membrane potential (tolbutamide: −41.5 ± 2.0 mV vs M617 in tolbutamide: −44. 3 ± 2.7 mV; paired t test, t(6) = 2.780, n = 7, p = 0.0320) and a significant reduction in the firing frequency (tolbutamide: 0.53 ± 0.23 Hz vs M617 in tolbutamide: 0.03 ± 0.03 Hz; Wilcoxon matched-pairs signed-rank test, W = −21.00, n = 7, p = 0.0312). These results suggest that KATP channels are not involved in the M617-induced inhibition of histaminergic neurons.
Galanin-induced inhibitory effects have also previously been linked to large conductance calcium-activated potassium (BK) channels (Pieribone et al., 1995). Although GALR1 is not typically associated with BK channel activation, we tested this possibility using tetraethylammonium (TEA 10 mM). In the presence of TEA, M617 only inhibited 55% (6/11) of histaminergic neurons. In those cells that were visibly inhibited by M617 in the presence of TEA (Fig. 5D), the M617-induced membrane hyperpolarization was abolished (TEA: −44.4 ± 1.5 mV vs M617 in TEA: −45.3 ± 2.1 mV; paired t test, t(5) = 1.040, n = 6, p = 0.3462), although there was a trend toward a decreased firing rate (TEA: 0.41 ± 0.12 Hz vs M617 in TEA: 0.08 ± 0.07 Hz; Wilcoxon matched-pairs signed-rank test, W = −15.00, n = 6, p = 0.0625). These results suggest that BK channels contribute to the inhibitory effect of M617.
Despite a reduction in the percentage of histaminergic neurons that were inhibited by M617 following blockade of GIRK or BK channels, with tertiapin-Q and TEA, respectively, neither of these compounds eliminated the inhibitory effects like Ba2+ did (Fig. 5E). Importantly, both tertiapin-Q and TEA significantly attenuated the magnitude of the M617-induced membrane hyperpolarization compared with control conditions (M617 alone; Fig. 5F). Together, these results suggest that galanin inhibits histaminergic neurons via GALR1-mediated opening of GIRK channels and activation of BK channels, but not through other Kir family channels such as KATP.
GALR1-induced inhibition of histaminergic neurons persists in inhibitors of synaptic transmission
To determine if the effects of GALR1 activation required action potential-dependent transmitter release, we exposed histaminergic neurons to M617 in the presence of tetrodotoxin (TTX 500 nM), a voltage-gated sodium channel blocker. Under these conditions M617 continued to hyperpolarize histaminergic neurons (Fig. 6A). In TTX, the M617-induced inhibitory effect was associated with a significant hyperpolarization of the membrane potential (TTX: −45.3 ± 2.5 mV vs M617: −50.6 ± 2.3 mV; paired t test, t(12) = 5.485, n = 13, p = 0.0001) and a significant reduction in neuronal input resistance (TTX: 646 ± 58 MΩ vs M617: 560 ± 41 MΩ; Wilcoxon matched-pairs signed-rank test, W = −81.0, n = 13, p = 0.0024) which was associated with a mean reversal potential of −72.8 ± 16.4 mV. This suggests that direct activation of GALR1 expressed on histaminergic neurons predominantly mediates the hyperpolarizing effect of galanin on histaminergic neuron electrical excitability. Interestingly, the percentage of histaminergic neurons responding to M617 in the presence of TTX (57%, 13/23 neurons) was slightly lower than that seen with M617 alone (73%).
GALR1-induced hyperpolarization of histaminergic neurons persists in the presence of inhibitors of synaptic transmission. A, Whole-cell current-clamp recording demonstrating the hyperpolarizing effect of M617 on histaminergic neuron activity following blockade of action potential mediated synaptic transmission with TTX. B, Whole-cell current-clamp recording demonstrating the inhibitory effect of M617 on histaminergic neuron activity following glutamatergic and GABAergic signaling with d-AP5, CNQX, and picrotoxin.
To further explore any role of fast transmitter release to the GALR1-induced inhibition, we exposed histaminergic neurons to M617 following blockade of glutamatergic and GABAergic signaling with d-AP5 (50 µM), picrotoxin (50 µM), and CNQX (10 µM). In the presence of these inhibitors of synaptic transmission, M617 continued to inhibit 61% (11/18) of histaminergic neurons (Fig. 6B) and was associated with a significant hyperpolarization of the membrane potential (synaptic inhibitors: −44.5 ± 1.4 mV vs M617: −48.5 ± 1.6 mV; paired t test, t(10) = 8.566, n = 11, p < 0.0001) and a significant decrease in firing frequency (synaptic inhibitors: 1.07 ± 0.33 Hz vs M617: 0.00 ± 0.00 Hz; Wilcoxon matched-pairs signed-rank test, W = −45.00, n = 11, p = 0.0039). The inhibitory effect of M617 was also associated with a decrease in input resistance (synaptic inhibitors: 683 ± 58 MΩ vs M617: 571 ± 52 MΩ; paired t test, t(10) = 7.262, n = 11, p < 0.0001) and a reversal potential of −80.9 ± 4.6 mV. Together, these results demonstrate that the GALR1-mediated inhibition of histaminergic neurons is largely independent of action potential-dependent synaptic transmission, or fast excitatory or inhibitory neurotransmitters, suggesting that the effect is mediated by direct activation of GALR1 expressed on histaminergic neurons.
Discussion
Using electrophysiological recordings established from genetically identified histaminergic (Hdc-expressing) neurons, we demonstrate that galanin strongly inhibits histaminergic neurons via activation of GALR1. We show that this GALR1-induced inhibition is mediated by GIRK and BK channels and is largely independent of action potential-mediated synaptic transmission or glutamatergic or GABAergic signaling. Moreover, histaminergic neurons strongly express Galr1. Together, these data represent the first demonstration that galanin inhibits histaminergic neuron activity via direct activation of GALR1 expressed on histaminergic neurons.
While multiple lines of evidence suggest that histaminergic neurons are inhibited by galanin or galanin-producing neurons (Arrang et al., 1991; Schonrock et al., 1991; Sherin et al., 1998), very few studies have directly explored this. Galanin has been demonstrated to reduce the firing of putative histaminergic neurons in the ventral subregion of the TMN (Schonrock et al., 1991), but more thorough investigations including the mechanisms involved are lacking. Here we report strong inhibitory effects of galanin on histaminergic neurons. This includes membrane hyperpolarization and inhibition of firing, which in most cases completely silenced histaminergic neurons. Given that all three galanin receptor subtypes are expressed in the hypothalamus (Waters and Krause, 2000), and all can interact with inhibitory Gi-type G-proteins (Lang et al., 2007), we systematically activated each receptor, demonstrating that only GALR1 activation recapitulates the galanin-induced inhibition. Moreover, we demonstrated Galr1 in almost 80% of histaminergic neurons, similar to the percentage of histaminergic neurons that responded to galanin or GALR1 agonism.
Our data demonstrate that GALR1-induced inhibition of histaminergic neurons is primarily mediated by direct postsynaptic activation of GALR1 expressed on these cells, consistent with prior evidence that galanin suppresses histamine release in a TTX-insensitive manner (Arrang et al., 1991). Pharmacological blockade of Kir channels, implicated GIRK channels in this response, in agreement with extensive literature showing that GALR1 engages G-protein signaling pathways to activate GIRK channels in multiple types of neurons (Mazarati et al., 2006; Webling et al., 2012; Constantin and Wray, 2016; Bai et al., 2018). Interestingly, TEA, a potent blocker of BK channels (Latorre et al., 1989), also attenuated GALR1-induced inhibition. Although GALR1 signals exclusively through Gi/o-type G-proteins, it can also promote Gβγ-mediated Ca2+ release from intracellular stores (Wang et al., 1998), a mechanism shared by other Gi-coupled receptors (Pfeil et al., 2020). Thus, GALR1-mediated activation of multiple types of potassium channels likely contributes to the powerful ability of galanin to silence histaminergic neurons.
Despite the GALR1-mediated hyperpolarization persisting after the blockade of voltage-gated sodium channels, and after blockade of glutamatergic and GABAergic signaling, the percentage of hyperpolarized cells decreased slightly in these conditions, raising the possibility of a small presynaptic contribution to the overall galanin-induced inhibition of histaminergic neurons. Additionally, we observed that the inhibitory effect of galanin, or GALR1 activation alone, was associated with an increase in input resistance in a small proportion of cells (24–25%). The corresponding reversal potentials indicate that this effect may result from inhibition of a nonselective cation conductance, consistent with previous suggestions of GALR1 signaling in the locus ceruleus (Bai et al., 2018). Because GALR1 activation is not generally thought to directly regulate nonselective cation channels, this effect may instead reflect reduced excitatory synaptic drive. In this regard, wake-active orexin neurons in the lateral hypothalamus provide excitatory inputs to histaminergic neurons via both orexin and glutamate release (Schone et al., 2012, 2014) and themselves express the inhibitory GALR1 (Laque et al., 2013). Thus, a dual action of galanin on orexin and histaminergic neurons may act synergistically to produce the strong inhibitory effect of galanin observed on histaminergic neurons.
Although we found that activation of GALR1 alone recapitulated the inhibitory effects of galanin, we also detected Galr3 expression in some histaminergic neurons. Despite no direct postsynaptic effects of GALR3 activation, it is possible that functional GALR3 are expressed on presynaptic terminals of histaminergic neurons. Such a localization would align with reports of presynaptic galanin actions mediated by GALR2/GALR3 in other brain regions (Kozoriz et al., 2006). Whether GALR3 contributes to the regulation of transmitter release from histaminergic neurons remains to be established.
Histaminergic neurons are described as wake-active, as their firing pattern is tightly associated with the waking state (Vanni-Mercier et al., 1984, 2003; Takahashi et al., 2006). This wake-selective activity pattern is also correlated with brain histamine levels, highest during waking, compared with NREM and REM sleep (Strecker et al., 2002; Dong et al., 2023). The “flip-flop switch” model of sleep and wakefulness proposes that inputs from the VLPO and lateral hypothalamus regulate histaminergic neuron activity (Saper et al., 2001, 2010). In agreement with this model, our data suggests that galanin inputs to histaminergic neurons, known to arise from the VLPO (Sherin et al., 1996, 1998), likely represent an important stimulus synchronizing histaminergic neuronal activity with behavioral state.
Galanin-expressing neurons are widely distributed throughout the central nervous system, including in other hypothalamic nuclei such as the arcuate nucleus, lateral hypothalamus, and paraventricular hypothalamus (Skofitsch and Jacobowitz, 1985; Melander et al., 1986). This raises the potential for additional sources of galaninergic inputs to histaminergic neurons. However, most of these galanin-expressing populations project to distinct brain regions and cell types which has not presently included to the TMN or histaminergic neurons (Holets et al., 1988; Merchenthaler et al., 1993; Leibowitz et al., 1998; Laque et al., 2013; Weinshenker and Holmes, 2016). In rodents, subpopulations of histaminergic neurons have been reported to coexpress galanin which may allow for local suppression of histamine release (Kohler et al., 1986; Staines et al., 1986; Kukko-Lukjanov and Panula, 2003). Nonetheless, confirmed afferents to histaminergic neurons remain limited (Sherin et al., 1998; Saper et al., 2005), with the VLPO currently representing the most likely major source of galaninergic innervation.
Since the discovery of VLPO sleep-active neurons, much work has further defined the role and neurotransmitter phenotypes of sleep-active neurons in preoptic (POA) subnuclei (Sherin et al., 1998; Zhang et al., 2015; Chung et al., 2017). As a result, POA neurons expressing galanin have emerged as key players modulating sleep. In particular, VLPOgalanin neuron activation has been shown to promote sleep and increase NREM sleep (Kroeger et al., 2018). Similarly, lateral preoptic (LPO) area galanin neurons have been shown to be essential for sleep homeostasis, the recovery sleep that occurs following sleep deprivation (Ma et al., 2019). Furthermore, these galanin neurons have been suggested to be responsible for promoting NREM sleep associated with α2 adrenergic agonist-induced sedation (Zhang et al., 2015).
While VLPOgalanin neuron promotion and maintenance of sleep is well recognized, theories surrounding how these neurons increase NREM are based on their proposed ability to inhibit wake-active neurons (Sherin et al., 1996, 1998; Saper et al., 2010). However, the mechanisms allowing VLPOgalanin neurons to inhibit the wake-active histaminergic neurons have not previously been characterized. Our findings identify a pivotal mechanism by which VLPOgalanin neurons could silence histaminergic neurons and are consistent with the idea that sleep-active VLPOgalanin neurons promote sleep via their inhibition of wake-active neurons (Kroeger et al., 2018). Importantly, other populations of wake-active neurons, including those in the dorsal raphe nucleus, locus ceruleus, and lateral hypothalamus, receive projections from VLPOgalanin neurons (Sherin et al., 1998; Lu et al., 2002; Kroeger et al., 2018), express galanin receptors (Parker et al., 1995; Laque et al., 2013), and are inhibited by galanin (Seutin et al., 1989; Xu et al., 1998; Goforth et al., 2014). Together, these findings suggest that a coordinated suppression of multiple populations of wake-active neurons, including the histaminergic neurons, likely mediates the sleep-promoting effect of VLPOgalanin neuron activation.
Elucidating the cellular and molecular mechanisms by which VLPOgalanin neurons promote sleep may inform pharmacological strategies to increase or consolidate sleep in individuals experiencing sleep disturbances. Indeed, lesions of the VLPO reduce sleep duration and promote sleep fragmentation in animal models (Lu et al., 2000), whereas targeted activation of VLPOgalanin neurons induces sleep in a rodent model of insomnia (Kroeger et al., 2018). Consistent with these preclinical findings, human studies report that loss of galanin neurons in the intermediate nucleus (the human equivalent of the VLPO) is associated with sleep fragmentation (Lim et al., 2014). These converging lines of evidence highlight the importance of VLPOgalanin neurons and their downstream targets for sleep regulation, as well as their potential relevance to sleep-related disorders.
In the present study, only male mice were used. However, the galanin system is influenced by sex hormones. In the hypothalamus, the expression of galanin fluctuates with the estrous cycle and is regulated by gonadal steroids (Gabriel et al., 1990; Merchenthaler et al., 1991). Likewise, Galr1 expression varies across the estrous cycle and is modulated by estradiol (Faure-Virelizier et al., 1998; Mitchell et al., 2004). It therefore remains to be determined if histaminergic neurons in female mice respond to galanin in the same way as described here.
In conclusion, the current study demonstrates that galanin represents a strong inhibitory stimulus to histaminergic neurons. We show for the first time that histaminergic neurons strongly express Galr1 and that the galanin-induced inhibition of histaminergic neurons largely occurs via direct postsynaptic activation of GALR1. Moreover, these effects are associated with the opening of multiple types of potassium channels. Together, these findings support the idea that some of the sleep-promoting effects of VLPOgalanin neuron activation occur via their ability to inhibit wake-active neurons, including histaminergic neurons.
Footnotes
The authors declare no competing financial interests.
This work was supported by funding from the Canada Research Chairs Program (to N.J.M. and A.C.), the Sleep Research Society Small Grants program (to N.J.M.), the Foundation of the Québec Heart and Lung Institute (to N.J.M.), the Canadian Foundation for Innovation (to N.J.M. and A.C.), and the Canadian Institutes of Health Research (to N.J.M.). N.J.M. was supported by a Sentinel North Partnered Research Chair in Sleep Pharmacometabolism (Canada First Research Excellence Fund). N.J.M. and A.C. were supported by a Fonds de Recherche du Quebéc – Santé (FRQS) Research Scholar J1 award. We thank Dr. Jeffrey M. Zigman (University of Texas Southwestern Medical Center) for his generous gift of the Hdc-Cre mice used in this work.
This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International license, which permits unrestricted use, distribution and reproduction in any medium provided that the original work is properly attributed.
References
In this issue
