High-Salt Diet Increases Suprachiasmatic Neuronal Excitability Through Endothelin Receptor Type B Signaling

Abstract

Circadian rhythms are 24-h oscillations in behavioral and biological processes such as blood pressure and sodium excretion. Endothelin B (ET_B) receptor has been connected to the molecular clock in peripheral tissues and plays a key role in the regulation of sodium excretion, especially in response to a high-salt diet. However, little is known about the role of ET_B in the primary circadian pacemaker in the brain, the suprachiasmatic nucleus (SCN), despite recent reports showing its enrichment in SCN astrocytes. In this study, we tested the hypothesis that high-salt diet (4.0% NaCl) impacts the circadian system via the ET_B receptor at the behavioral, molecular, and physiological levels in C57BL/6 mice. Two weeks of high-salt diet feeding changed the organization of nighttime wheel-running activity, as well as increased the SCN expression of ET_B mRNA determined by fluorescence in situ hybridization at night. Neuronal excitability determined using loose-patch electrophysiology was also elevated at night. This high-salt diet-induced increase in SCN activity was ameliorated by ex vivo bath application of an ET_B antagonist and could be mimicked with acute treatment of endothelin-3. Finally, we found that the excitatory effects of endothelin-3 were blocked with co-application of an N-methyl-D-aspartate (NMDA) receptor antagonist, suggesting that glutamate mediates endothelin-induced neuronal excitability in the SCN. Together, our data demonstrate the presence of functional ET_B receptors in SCN astrocytes and point to a novel role for endothelin signaling in mediating neuronal responses to a dietary sodium intake.

Graphical Abstract

Open in new tabDownload slide

Introduction

Circadian rhythms are 24-h oscillations in behavioral and physiological processes such as blood pressure and sodium excretion.^1,2 At the cellular level, circadian rhythms are controlled by a transcriptional translational feedback loop of core clock genes, which drive time-of-day changes in expression of numerous, tissue-specific, clock-controlled genes.³ While these local circadian “clocks” exist throughout all tissues, the central pacemaker—the suprachiasmatic nucleus (SCN) of the hypothalamus—serves as the primary clock for the body. The SCN is responsible for synchronizing peripheral tissues and the rhythms they drive with external time cues such as light as well as other secondary clocks. Time-of-day information is communicated by the SCN to local clocks through daily rhythms in spontaneous neuronal activity, with high activity during the day and low activity at night.⁴ Recent work has demonstrated that this canonical SCN output is affected by acute salt loading through signaling from the organum vasculosum of the lamina terminalis (OVLT),⁵ yet the effect of chronic consumption of a high-salt diet (HSD) is still unknown. It is well known that homeostatic control of thirst and acute drinking behavior is driven by the subfornical organ (SFO), which induces a rapid increase in water intake in response to hyperosmolarity (for review see Zimmerman⁶). However, the recent emergence of the SCN-OVLT axis as a driver of daily rhythms in fluid intake, independent of plasma osmolarity,^5,7 has highlighted the SCN response to HSD as a remaining gap in our current understanding.

The endothelin system plays an important role in cardiovascular regulation and fluid-electrolyte homeostasis.⁸ This system is made up of 2 G-protein–coupled receptors and 3 structurally similar ligands,⁸ with many of these components present in the hypothalamus.⁹ Endothelin B (ET_B) receptors in the kidney play a key role in the excretion of salt and regulation of blood pressure, particularly in response to an HSD.¹⁰ Specifically, endothelin-1 (ET-1) functions as an autocrine factor to stimulate ET_B receptors in response to an HSD to inhibit the reabsorption of sodium. Loss of ETB receptor function impairs the ability of the kidney to excrete salt and so the resulting salt-dependent hypertension is necessary for maintaining balance between intake and excretion.⁹ Recent evidence from our group and others suggests a connection between the molecular clock and the endothelin system in peripheral tissues.^2,11 Interestingly, single-cell RNA sequencing of the SCN has revealed that Ednrb encoding ET_B is expressed in SCN astrocytes. Moreover, SCN endothelial cells rhythmically express clock genes as well as Edn3 (encoding endothelin-3 or ET-3).¹² This result suggests that the endothelin system has a physiological role in the SCN, and although several studies have examined endothelin in other hypothalamic nuclei,^13,14 the functional role of the endothelin system in the SCN remains unexplored. Therefore, the goal of this study was to determine the potential role for ET_B in the SCN, and specifically, the SCN response to an HSD.

Materials and Methods

Animals

Male and female, 2-month-old C57BL6/J mice were bred in house or purchased from Jax (Strain #:000664) and then group housed (unless otherwise noted) in a 12:12 light/dark (LD) cycle with food and water provided ad libitum. For the day/night and pharmacological recordings, mice were acclimated to a reverse LD cycle ≥2 weeks before feeding interventions. Mice were fed either an HSD (4% NaCl, Teklad Diets, #92034) or normal salt diet (NSD, 0.49% NaCl) for 2 weeks prior to experiments and were euthanized with cervical dislocation and rapid decapitation. The 2-week period was chosen to allow animals to adjust the many factors that regulate fluid electrolyte balance and is typical of that used in rodent studies.^15–18 All animal procedures were approved by the University of Alabama at Birmingham Institutional Animal Care and Use Committee in accordance with the NIH Guide for the Care and Use of Laboratory Animals.

Behavior

To avoid the confounding effects of the estrous cycle on wheel-running activity such as an inconsistent “siesta” period,^19,20 only male mice were used for this experiment. Male mice aged 3-4 months were single-housed in individual wheel-running cages, and wheel-running activity was recorded and analyzed with ClockLab software (Actimetrics, Wilmette, IL). Mice assigned to the HSD group were switched to HSD when first placed in wheel-running cages, while NSD mice continued on normal chow. Behavior was analyzed across days 11-19 after start of diet to allow time for acclimation to wheel-running cages. Actograms of activity in 6-min bins were used to determine the time of activity onset and offset as well as alpha length as previously described.²¹ Average activity profiles in 1-h bins were used to compare activity between groups across full 24-h cycle. Individual activity profiles in 30-min bins were used to determine time of “siesta” and length of the first and second bouts of activity per night. Time of “siesta” was defined as the 30-min period of lowest activity between Zeitgeber time (ZT) 18-23 (where ZT 12 refers to lights off).²² The start and the end of each activity bout were determined as the first 30-min bin that crossed above or below 50% maximum activity, respectively (see Figure 1C) and were used to calculate the first and second bout duration.²³ In mice without a second peak in activity or where the second peak did not cross this threshold, the second bout duration was considered zero.

Electrophysiology

Male and female mice were fed HSD or NSD for 2 weeks ad lib starting at 2.5 months of age. Fresh brain slices were prepared using previously published methods (as in Paul et al.²⁴). Briefly, brains were submerged in ice-cold oxygenated sucrose saline (in m m): 250 sucrose, 26 NaHCO₃, 1.25 Na₂-HPO₄-7H₂O, 10 glucose, 2.5 MgCl₂, and 3.5 KCl, and coronal SCN–containing sections (200 µm) sections were taken on a vibroslicer (Campden 7000SMZ, World Precision Instruments). Slices were rested for ∼20 min at RT in an oxygenated mixture of 50% sucrose saline and 50% extracellular solution (in m m): 130 NaCl, 20 NaHCO₃, 1 Na₂-HPO₄-7H₂O, 10 glucose, 1.3 MgSO₄-7H₂O, 3.5 KCl, and 2.5 CaCl₂. Slices were transferred to a heated (34°C ± 0.5°C) recording chamber and continuously perfused (2 mL/min) with oxygenated extracellular solution for the duration of the recording window. For pharmacological experiments, treatments were bath-applied 30 min before the start of recordings and remained in bath throughout the 4-h window. Individual cells were visualized using differential interference contrast (DIC) on an Axio Examiner microscope (Carl Zeiss Inc.). Targeted loose-patch recordings were made from SCN neurons during the day (ZT 4-8) or night (ZT 14-18) using 3-5 mΩ resistance glass pipettes filled with internal solution (in m m): 150 NaCl, 10 4-(2-hydroxyethyl) piperazine-1-ethanesulfonic acid (HEPES), 10 glucose, 2.5 CaCl₂, 1.3 MgCl₂, and 3.5 KCl. Electrophysiological recordings were sampled at 20 kHz and filtered at 10 kHz using a Multiclamp 700B amplifier and pClamp 11.2 software (Axon Instruments). Spontaneous firing rate was measured using 2-min recordings in gap-free mode and analyzed in Clampfit 11.2.

Pharmacology

IRL-1620 (Tocris, #1196), ET-3 (Tocris, #1162), and 3-(2-carboxyp-iperazin-4-yl)propyl-1-phosphonic acid (CPP, Abcam, ab144495) were dissolved in water. A-192621 (Med Chem Express, HY-120295) and 2-cyano-3-(4-hydroxyphenyl)-2-propenoic acid (CHC; Tocris, #5029) were dissolved in dimethyl Sulfoxide (DMSO) before being added to recording solution for a final working concentration of 0.03% and 0.2% DMSO, respectively.

Fluorescence in Situ Hybridization

Fresh frozen brain tissue was sectioned at 15-µm thickness through the SCN and mounted on glass slides. Slides were stored at −80°C until ready for staining. Sections were stained using the manufacturer-designed probes for Ednrb (Advanced Cell Diagnostics Bio catalog #473801), Eno2 (catalog #517261), and Gfap (catalog #313211) according to the manufacturer’s recommended protocols. Briefly, slides were fixed in 10% neutral buffered formalin at 4°C for 15 min. Slides were then dehydrated through an ethanol series, then dried at RT for 5 min. A hydrophobic barrier was drawn around the section and the tissue incubated with protease IV for 30 min at RT. After washing in phosphate buffered saline (PBS) , slides were incubated with the appropriate probe mix for 2 h at 40°C in the HybEZ II Oven (ACD Bio). Sections were then processed with the RNAScope Fluorescent Multiplex Assay v2 reagents using the manufacturer’s recommendations.

Images were quantified using a custom semi-automated pipeline in Cell Profiler (v4.2.1). Briefly, DAPI was used to identify nuclei within the boundaries of the SCN, which were then used to identify Ednrb-positive nuclei. The region of interest (ROI) for Ednrb-positive nuclei was then expanded out by 10 pixels. Ednrb expression was measured as the percentage of an ROI area, which was covered by Ednrb signal.

Statistical Analysis

All data were analyzed using SPSS 29 (IBM). All data were tested for assumptions of normality and homogeneity of variances, and violations of these assumptions were handled with transformations or use of nonparametric tests as needed. Behavioral data were analyzed using independent-samples t-tests or 2-way, repeated-measures analysis of variance (ANOVA) with Holm-Bonferroni-corrected post hoc. Fluorescence in situ hybridization (FISH) data were square root-transformed to correct for a non-normal distribution and analyzed with a linear mixed model with a first-order auto-regressive covariance structure using “batch” as a covariate and “time” and “diet” as fixed factors and repeated measures across cells within each brain. Estimated marginal means (EMM) and 95% CI reported in results were back transformed for clarity of interpretation. Due to the non-normal distribution of SCN spike rate data, loose-patch results were analyzed with nonparametric Kruskal-Wallis tests (Figure 4) or the Scheirer-Ray-Hare extension of the Kruskal-Wallis tests, which allows for a 2 × 2 design. Sex was included as a factor in initial analysis of electrophysiological data and found to have no significant main effect or interactions; thus, data from male and females animals were combined for the final analysis. Contingency data were analyzed with 3-way loglinear analysis and χ² tests. Significance was ascribed at P < .05.

Results

To examine the impact of dietary salt intake on daily activity patterns, male C57 mice were placed in wheel-running cages and fed high (4% NaCl) or normal (0.49% NaCl) salt diet. Consistent with our previous work in rats,²⁵ HSD did not impact the overall activity levels in mice (Figure 1A and D). However, HSD changed the organization of activity across the 24-h cycle (Figure 1B). Specifically, HSD-fed animals had increased activity at ZT 19. Interestingly, this time of the night is typically referred to as the “siesta” period—the approximately 2-h period in the late night where mice will sleep.²² In fact, when we examined the timing of the mid-point of the “siesta” period (as defined in the “Materials and Methods” section), HSD-fed mice trended toward a later siesta (mean ± SEM: HSD, 21.83 ± 0.62; t-test, t₍₅₎ = 2.169, P = .082) but with significantly greater variability across animals; whereas in NSD-fed mice, the siesta mid-point occurred at ZT 20.5 for every mouse (Levene’s test, F_(1,10) = 7.191, P = .023; Figure 1C arrowheads). Moreover, the first nightly bout of activity was >2 h longer in HSD-fed mice on average (Figure 1F), while there was a trend for the second, post-rest, bout of activity to be shorter in HSD-fed mice (Figure 1G). This change to the late-night rest period is was also reflected in the trend for HSD-fed mice to spend more time of the night in an activity bout (Figure 1H), despite significantly longer alpha lengths (time between activity onset and activity offset) of NSD-fed mice (Figure 1E). Together, these results suggest, HSD-fed mice are less likely to take a siesta, and if a siesta is present, it is likely to be shorter in duration.

We next sought to determine whether changes in SCN output could account for the HSD-induced changes in behavior. To do this we examined spontaneous action potential firing from SCN neurons during the day or night after 2 weeks on an HSD or NSD. As expected, NSD-fed mice had a day/night difference in SCN spike rate, with higher activity during the day (median [interquartile range or IQR]: NSD-day, 2.59 [0.65-5.40]; NSD-night, 0.19 [0.00-1.49 Hz]; Figure 2). This day/night difference was blunted in neurons from mice fed an HSD (median [IQR]: HSD-day, 2.52 [0.64-5.55]; HSD-night, 1.28 [0.07-4.07 Hz]). Interestingly, this change was specifically driven at night, at which time the average spike rate of neurons from HSD-fed mice was more than double that of NSD neurons. Additionally, HSD decreased the proportion of silent cells compared to NSD across both times of day, with HSD cells 2.07 times more likely to be spiking at night than NSD cells (Figure 2C).

Previous work has demonstrated that HSD increases the expression of components of the ET system in various extrarenal tissues including endothelial cells.^26–28 However, less is known about the impact of the ET system in the central regulation of acclimation to an HSD despite evidence for endothelial cell expression of Edn3 in the SCN (Supplemental Figure S1; Wen et al.¹²), as well as expression of ET_B receptors in multiple hypothalamic regions, including the SCN.^9,12,29 To determine whether there is differential expression of Ednrb, the gene encoding ET_B receptors, about the day versus night or HSD versus NSD, we next used FISH to probe for Ednrb in SCN sections from NSD- and HSD-fed animals collected at ZT 6 (day) or ZT 18 (night). SCN cells exhibited a significant diurnal difference in Ednrb expression, with higher expression during the day (main effect of time, F₍₁₉₇₅₎ = 4.779, P = .029); however, this day/night difference was blunted in HSD-fed animals (Figure 3). Specifically, HSD significantly increased Ednrb expression by 1.26 fold compared to NSD at night (EMM [95% CI]: HSD-night, 24.9% [23.7-26.2]; NSD-night, 19.7% [18.4-21.1]) but did not have a significant effect on Ednrb expression during the day (EMM [95% CI]: HSD-day, 23.3% [22.2-24.3]; NSD-day, 24.0% [22.9-25.1]), consistent with the HSD-induced changes in SCN firing rate.

We next sought to confirm the presence of functional ET_B receptors in the SCN and determine whether their activation is sufficient to increase SCN neuronal activity. To do this, we treated acute brain slices from NSD-fed mice with either a selective ET_B agonist (IRL-1620, 10 n m), ET-3 (10 n m), or vehicle (water), and recorded from SCN neurons during the projected night. Both treatments significantly increased SCN firing compared to vehicle-treated controls (median [IQR]: Veh, 0.11 [0.00-2.18]; IRL-1620, 1.26 [0.00-3.11]; ET-3, 1.24 [0.07-3.50 Hz]; Figure 4A and B). Moreover, ET-3-treated cells were 2.93 times more likely to be spiking than vehicle-treated controls (Figure 4C). Together, these results demonstrate that bath application of ET-3 sufficiently mimics the effects of HSD feeding on nighttime SCN excitability.

Given that ET_B receptor activation was capable of increasing SCN firing at night, we next determined whether ET_B receptor activation is necessary for the HSD-induced increase in SCN neuronal activity at night. To do this, we fed mice NSD or HSD for 2 weeks and then recorded from SCN neurons at night in the presence of chronically bath-applied ET_B antagonist (A-1922621, 1 µm) or vehicle control (DMSO, 0.03%). As expected, in vehicle-treated slices, SCN cells from HSD-fed animals had a significantly higher spike rate compared to NSD controls (median [IQR]: NSD-veh, 0.08 [0.00-1.31]; HSD-veh, 2.21 [0.24-4.51 Hz]; Figure 5A and B). Treatment with the ET_B antagonist significantly reduced the spontaneous firing rate of cells from HSD-fed mice but had no effect on cells from NSD-fed animals (median [IQR]: HSD, 0.20 [0.00-2.56]; NSD, 0.00 [0.00-1.29 Hz]). Although the HSD + antagonist cells were still significantly more excitable than NSD + antagonist cells (P = .018), neuronal activity was restored to levels similar to the NSD + veh group (P = .521). Moreover, the proportion of silent cells were similar across all groups except the HSD + veh, with HSD + veh cells 3.44 times more likely to be spiking than HSD + antagonist cells (Figure 5C). Taken together, these results suggest that local ET_B signaling in the SCN contributes to HSD-induced changes in neuronal excitability.

In other brain areas, ET_B receptors are localized to both neurons and astrocytes¹³; however, a recent single-cell RNA-seq study in the SCN reported a substantial enrichment of Ednrb specifically in astrocytes.^12,29 Similarly, we found co-expression of Ednrb in cells expressing the astrocytic marker, Gfap, but not in cells expressing the neuronal marker, Eno2 (Supplemental Figure S2), confirming the astrocytic enrichment of Ednrb in the SCN. In the SFO, activation of glial ET_B receptors promotes the release of lactate, which can be taken up into neurons through monocarboxylate transporters (MCTs), and leads to increased excitability in local neurons.¹⁴ To test whether lactate was also involved in the SCN neuronal response to ET-3, we performed loose-patch recordings of SCN neurons from NSD-fed animals at night (projected ZT 14-18) in the presence of vehicle (DMSO, 0.2%), ET-3 (10 n m), the MCT inhibitor, CHC (200 µm), or ET-3 + CHC. Consistent with our above findings, ET-3 treatment significantly increased the spike rate of SCN neurons compared to vehicle-treated controls (median [IQR]: veh, 0.01 [0.00-0.80]; ET-3, 1.27 [0.00-3.56 Hz]; Figure 6A and B). Treatment with the MCT blocker, CHC, completely blocked this ET-3-induced excitability (median [IQR]: CHC, 0.02 [0.00-0.71]; ET-3 + CHC, 0.02 [0.00-0.92 Hz]), while CHC had no effect on SCN firing rate in vehicle-treated slices. Furthermore, ET-3-treated slices had a significantly higher percentage of spiking cells compared to all other treatment groups, with ET-3-treated cells 2.19 times more likely to be spiking than ET-3 + CHC–treated cells (Figure 6C).

In a nearby hypothalamic region, the supraoptic nucleus (SON), astrocytic ET_B receptor activation has been shown to augment neuronal activity through glutamatergic signaling.¹³ Moreover, recent work has demonstrated that glutamate serves as a mechanism by which astrocytes regulate SCN neuronal synchrony.³⁰ Therefore, we next tested whether glutamatergic signaling underlies the increase in neuronal activity in response to ET-3 in the SCN. To do this, we performed loose-patch recordings from SCN neurons from NSD-fed animals at night (projected ZT 14-18) in the presence of vehicle (water), ET-3 (10 n m), the NMDA blocker, CPP (10 µm), or ET-3 + CPP. Consistent with our above findings, ET-3 treatment significantly increased the spike rate of SCN neurons, compared to vehicle-treated controls (median [IQR]: veh, 0.02 [0.00-0.85]; ET-3, 1.22 [0.05-4.34 Hz]; Figure 7A and B). Remarkably, treatment with the NMDA antagonist, CPP, completely blocked this ET-3-induced excitability (median [IQR]: CPP, 0.08 [0.00-1.25]; ET-3 + CPP, 0.07 [0.00-1.08 Hz]). In the presence of vehicle, CPP had no effect on SCN firing rate. Similar to the effect on spike rate, ET-3-treated slices had a significantly lower percentage of silent cells, compared to all other treatment groups, with ET-3-treated cells 3.9 times more likely to be spiking than ET-3 + CPP–treated cells (Figure 7C). Together, these results suggest that glutamatergic signaling mediates the excitatory effects of ET-3 on SCN neurons at night.

Discussion

The goal of our current study was to determine the impact of high salt intake on output signaling of the central circadian pacemaker, as well as explore the ET system as a possible mechanism underlying these changes. Here, we present 3 major findings: (1) 2 weeks of HSD feeding is sufficient to change the organization of daily wheel running without impacting the overall activity levels, (2) chronic HSD enhances SCN neuronal excitability at night due to local ET signaling involving activation of ET_B receptors localized to SCN astrocytes, and (3) ET-3-driven neuronal activation depends on MCT function and NMDA receptor activation. To our knowledge, our results are the first to demonstrate a novel mechanism for the circadian effects of HSD mediated by the ET-3 peptide that appears to be derived from endothelial cells to activate ET_B receptors in the SCN including a novel role for SCN astrocytes.

Our initial finding that HSD feeding disrupts typical circadian behavior contributes to the growing body of literature highlighting the SCN as a mediator of fluid-electrolyte control. A recent study reported that increases in late-night neuronal activity in arginine vasopressin-expressing SCN neurons projecting to the OVLT were responsible for an increase in water intake during the end of the active phase.⁷ C57BL6 mice exhibit a distinct pattern of nocturnal behavior with 2 distinct bouts of activity separated by a period of sleep often referred to as the “siesta.”^22,23,31 In this study, we found that 2 weeks of HSD led to a reorganization of nighttime activity such that the initial activity bout was extended into the mid-to-late night “siesta” period. This resulted in a shortened second bout of activity, with the second bout of activity entirely lost in half of the HSD-fed animals. These changes in locomotor activity coincided with an increase in SCN neuronal activity in the early-mid night (ZT 14-18). Although future experiments are needed to determine the physiological changes driving this reorganization of nighttime activity, it is possible that this increase in SCN firing extends the active phase in order to lengthen the period of drinking (offsetting the high salt intake). While we do not measure water intake in this study, this theory is consistent with our previous work showing increased water intake in HSD-fed rats²⁵ and mice,³² primarily during the active (ie, dark) phase. Future experiments examining fluid intake at a higher resolution are needed to fully elucidate the timing of these changes. Interestingly, in humans, sodium intake has been reported to have an opposite effect on water consumption in the long term.³³ Given that SCN activity rhythms appear to be similar in both nocturnal and diurnal mammals, with higher activity in the light phase,³⁴ further research into SCN control of thirst could provide valuable insight into how these dynamics change across temporal niche.

Our finding that HSD-induced neuronal excitability is driven by local endothelin signaling is supported by multiple pieces of evidence. First, we demonstrate that endothelial cells within the SCN express Edn3, the gene encoding ET-3, consistent with a previous single-cell RNA-sequencing study.¹² Second, the increase in SCN spike rate persists for up to 6 h after external influences have been removed. Moreover, acute treatment with an ET_B antagonist reverses the effects of HSD ex vivo. Although this evidence supports a model with local SCN endothelial cell release of ET-3, it is important to note that circulating ETs in the blood or surrounding brain regions could also contribute to the behavioral effects of HSD. Thus, future research should investigate the signaling mechanisms of dietary sodium to SCN endothelial cells, the length of time that the effects of HSD persist after diet reversal, and the identification of other rhythmic processes in the SCN that are impacted by HSD. Prior single-cell RNA-sequencing study of the SCN did not reveal detectable expression of ET-1 (Edn1)¹² that is considered the primary endothelial-derived isoform. Consistent with these findings, we were unable to detect ET-1 expression in the SCN using in situ hybridization (data not shown).

Although this study demonstrated a role for ET signaling in the SCN response to HSD, more work is needed to determine the functional role of ET in central circadian regulation under normal, unchallenged, conditions. Recent work has demonstrated the existence of the portal vein system in which blood flows from the SCN to the nearby OVLT at a higher rate at night than during the day, driven by an as yet undetermined mechanism.³⁵ Interestingly, the rhythmic expression of Edn3 in SCN endothelial cells also peaks during the early night phase,¹² which raises the possibility for ET-3 contributing to the changes in flow rate across time of day. Given that HSD-fed animals had an increase in Ednrb expression at night, when Edn3 is reportedly higher, it is possible that an increase in ET_B receptors sensitizes the SCN to a normally occurring nightly increase in ET-3 release. However, a limitation of this study is the lack of measurement of ET-3 release in the SCN in response to HSD. Thus, we cannot rule out the possibility that a change in ET-3 release also contributes to the HSD-induced excitability in the SCN. In fact, previous work has demonstrated an increase in ET-3 release in the SFO following dehydration,¹⁴ so future research should examine whether a similar process occurs within the SCN in response to HSD.

Although ET_B receptors have been found to be present on both astrocytes and neurons in other brain regions,^9,13 our data suggest these receptors are restricted to astrocytes alone in the SCN. This is consistent with recent single-cell RNA sequencing data showing Ednrb as an astrocytic marker in the SCN.¹² Additionally, we found that ET-3-induced excitability was dependent on MCTs, which allow for the transport of lactate from astrocytes to neurons,³⁶ consistent with previous findings in the SFO.¹⁴ As regulators of the blood-brain barrier, astrocytes are perfectly positioned to serve as a mediator between circulating signals and neurons. In fact, even in a brain region with both astrocytic and neuronal expression of ET_B receptors, like the SON, the neuronal response to ET_B signaling appears to be indirect. Specifically, ET_B-induced changes in neuronal activity were dependent on astrocytic modulation of glutamate or nitric oxide for excitatory or inhibitory responses, respectively.¹³ The present results show that ET-3 excited SCN neurons at night, and thus, we focused on glutamate-mediated signaling and found that this excitatory effect was dependent upon NMDA receptor activation. However, future work should examine the possible mechanisms by which astrocytic ET_B activation may augment glutamatergic signaling, such as Ca²⁺-dependent exocytosis or modulation of excitatory amino acid transporters.³⁶ In fact, previous work has shown that neuronal lactate uptake can lead to potentiation of NMDA receptors through alteration of redox state.³⁷ Given that extracellular levels of glutamate are already elevated in the SCN at night,³⁰ lactate-driven changes in NMDA receptor sensitivity could provide a mechanism for ET-3-induced excitability without a need for additional glutamate release. Interestingly, astrocytic regulation of daily rhythms in extracellular glutamate has recently been shown to contribute to SCN network synchrony.³⁰ Therefore, future work should examine the impact of HSD consumption and/or endothelin signaling on intra-SCN coupling as well.

Author Contributions

Jodi R. Paul (Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing—original draft, Writing—review & editing), Megan K. Rhoads (Conceptualization, Data curation, Formal analysis, Methodology, Visualization, Writing—review & editing), Anna Elam (Data curation), David M. Pollock (Conceptualization, Formal analysis, Funding acquisition, Investigation, Resources, Software, Supervision, Visualization, Writing—original draft, Writing—review & editing), and Karen L. Gamble (Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Visualization, Writing—review & editing)

Funding

National Institute of Diabetes and Digestive and Kidney Diseases: R01 DK134562 (to D.M.P. and K.L.G.); American Heart Association Postdoctoral Fellowship: 827566 (to M.K.R.); and American Heart Association Career Development Award: 24CDA1046338 (to M.K.R.).

Conflict of Interest Statement

D.M.P. holds the position of Editor-in-Chief for FUNCTION and is blinded from reviewing or making decisions for the manuscript. K.L.G. holds the position of Editorial Board Member for FUNCTION and is blinded from reviewing or making decisions for the manuscript.

Data Availability

Source data are available at 10.6084/m9.figshare.28525553.

References