Activation of ionotropic and group I metabotropic glutamate receptors stimulates kisspeptin neuron activity in mice

Abstract

Different populations of hypothalamic kisspeptin (KISS1) neurons located in the rostral periventricular area of the third ventricle (RP3V) and arcuate nucleus (ARC) are thought to generate the sex‐specific patterns of gonadotropin secretion. These neuronal populations integrate gonadal sex steroid feedback with internal and external cues relayed via the actions of neurotransmitters and neuropeptides. The excitatory amino acid neurotransmitter glutamate, the main excitatory neurotransmitter in the brain, plays a role in regulating gonadotropin secretion, at least partially through engaging KISS1 signaling. The expression and function of individual glutamate receptor subtypes in KISS1 neurons, however, are not well characterized. Here, we used GCaMP‐based calcium imaging and patch‐clamp electrophysiology to assess the impact of activating individual ionotropic (iGluR) and group I metabotropic (mGluR) glutamate receptors on KISS1 neuron activity in the mouse RP3V and ARC. Our results indicate that activation of all iGluR subtypes and of group I mGluRs, likely mGluR1, consistently drives activity in the majority of KISS1 neurons within the RP3V and ARC of males and females. Our results also revealed, somewhat unexpectedly, sex‐ and region‐specific differences. Indeed, activating (S)‐α‐amino‐3‐hydroxy‐5‐methyl‐4‐isoxazolepropionic acid (AMPA) type iGluRs evoked larger responses in female ARC^KISS1 neurons than in their male counterparts whereas activating group I mGluRs induced larger responses in RP3V^KISS1 neurons than in ARC^KISS1 neurons in females. Together, our findings suggest that glutamatergic neurotransmission in KISS1 neurons, and its impact on the activity of these cells, might be sex‐ and region‐dependent in mice.

1. INTRODUCTION

Gonadotropin‐releasing hormone (GnRH) neurons drive luteinizing (LH) and follicle‐stimulating hormone secretion from the anterior pituitary gland, thereby controlling gonadal function. GnRH secretory patterns are regulated by gonadal steroid hormonal feedback loops that are mediated through neuronal circuits afferent to the GnRH neurons. ¹ Kisspeptin (KISS1) neurons of the arcuate nucleus (ARC) and rostral periventricular area of the third ventricle (RP3V) project to GnRH neurons, drive their electrical activity, promote LH secretion through the release of KISS1, and mediate gonadal hormone feedback in females. ² , ³ , ⁴ , ⁵ , ⁶ , ⁷ , ⁸ In addition to gonadal feedback, the GnRH neuronal network also integrates internal and external cues relayed through the release of various neurotransmitters and neuropeptides acting on their cognate receptors to regulate GnRH and KISS1 neuron activity. ¹ , ⁹ , ¹⁰ , ¹¹ , ¹²

The impact of the excitatory amino acid glutamate on the GnRH neuronal network and on LH secretion has been long recognized (reviewed in reference 13). Administration of glutamate, or of ionotropic glutamate receptor (iGluR) and group I metabotropic glutamate receptor (mGluR) agonists, increases GnRH and LH secretion in rodents through a central mechanism. ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ More importantly, activation of iGluRs by endogenous glutamate has been shown to play a key role in pulsatile ²¹ , ²² , ²³ , ²⁴ and surge ¹³ , ¹⁹ , ²⁵ , ²⁶ , ²⁷ LH secretion. These classic observations, however, do not provide any information on where in the GnRH neuronal network GluR activation plays its role in regulating LH secretion.

Exogenous glutamate as well as iGluR and group I mGluR agonists stimulate GnRH neuron activity. ²⁸ , ²⁹ , ³⁰ , ³¹ , ³² , ³³ Interestingly, intact KISS1 signaling is required for at least some of the effects of activating GluRs on LH secretion, ³⁴ , ³⁵ , ³⁶ suggesting actions at KISS1 neurons. KISS1 neurons express genes for all iGluR subtypes in the ARC and RP3V and for most mGluRs in the RP3V. ³⁷ , ³⁸ , ³⁹ KISS1 neurons also increase their activity in response to AMPA and N‐methyl‐D‐aspartic acid (NMDA) receptor activation. ⁴⁰ , ⁴¹ , ⁴² , ⁴³ Both KISS1 populations exhibit AMPA/kainate receptor‐mediated post‐synaptic currents (PSCs) ⁴⁴ , ⁴⁵ , ⁴⁶ and, importantly, blockade of fast glutamatergic synaptic transmission suppresses ARC^KISS1 neuron coordinated activity in vitro and in vivo, ⁴⁷ , ⁴⁸ , ⁴⁹ suggesting key roles for iGluR activation in pulsatile GnRH secretion. However, the impact of iGluRs and mGluRs on KISS1 neuron activity remains to be fully characterized.

Here, we used GCaMP‐based intracellular Ca²⁺ concentration ([Ca²⁺]_i) imaging and patch‐clamp electrophysiology in brain slices to determine the impact of iGluR and group I mGluR signaling on ARC^KISS1 and RP3V^KISS1 neuron activity. Our results reveal that activation of these receptors consistently stimulates these cells, with potential sex‐dependent and regional differences.

2. MATERIALS AND METHODS

2.1. Animals

For [Ca²⁺]_i imaging, we used mice expressing the genetically‐encoded Ca2⁺ indicator GCaMP6f ⁵⁰ in KISS1 neurons, obtained by crossing mice that express the DNA‐recombining enzyme Cre recombinase (Cre) in KISS1‐expressing cells (Kiss1‐Cre; RRID:IMSR_JAX:023426) ⁵¹ with mice that carry Cre‐dependent GCaMP6f at the ROSA26 locus (flox‐STOP‐GCaMP6f; RRID:IMSR_JAX:028865). ⁵² Male and female offspring (2–6 months old) heterozygous for the Kiss1‐Cre and flox‐STOP‐GCaMP6f alleles (Kiss1‐Cre::GCaMP6f) were used in experiments, as previously described. ⁵³ For electrophysiology, female mice expressing the humanized renilla green fluorescent protein (hrGFP) in KISS1 neurons (Kiss1‐hrGFP; RRID:IMSR_JAX:023425) ⁵⁴ were used. Unless otherwise noted, female mice were used in the diestrous stage, as determined by vaginal lavage (3.5–5.0 μL H₂O, taken at zeitgeber [ZT] 1–3) smeared on slides, stained with methylene blue and examined by light microscopy. ⁵⁵ Mice were group‐housed with littermates under controlled temperature (23 ± 2°C) and lighting conditions (12‐h light/dark cycles) with ad libitum access to food and water. Mice were assigned to experiments based on their sex and genotype. All experiments were approved by Kent State University's Institutional Animal Care and Use Committee.

2.2. Brain slice preparation

Brain slices were obtained as described previously. ⁵³ Briefly, mice were killed between ZT2 and ZT7 by decapitation under isoflurane anesthesia, and their brains were quickly extracted. Thereafter, 200–250 μm‐thick coronal brain slices including the RP3V, or the ARC were obtained using a vibrating blade microtome (HM650V, Microm International GmbH, or VT1200S, Leica Microsystems, RRID:SCR_020243) in an ice‐cold solution containing (in mM): 87 NaCl, 2.5 KCl, 25 NaHCO₃, 1.25 NaH₂PO₄, 0.5 CaCl₂, 6 MgCl₂, 25 glucose and 75 sucrose. Brain slices were left to incubate at 30–34°C for at least 1 h in artificial cerebrospinal fluid (aCSF) containing (in mM): 125 NaCl, 2.5 KCl, 26 NaHCO₃, 1.25 NaH₂PO₄, 2.5 CaCl₂, 1.2 MgCl₂ and 11 glucose. All solutions were equilibrated to pH 7.4 with a mixture of 95% O₂/5% CO₂.

2.3. [Ca²⁺]_i imaging and electrophysiology

Individual brain slices were placed under an upright epifluorescence microscope (Scientifica or Prior Scientific) and constantly perfused (1.0–1.5 mL/min) with warm (32–34°C) aCSF.

2.3.1. [Ca ²⁺]_i imaging

Variations in [Ca²⁺]_i in KISS1 neurons were estimated by measuring fluorescence changes in individual GCaMP6f‐expressing neurons in brain slices from Kiss1‐Cre::GCaMP6f mice. Slices were illuminated through a ×40 immersion objective, using a light‐emitting diode light source (pE300ultra LED; CoolLED, RRID:SCR_021972) filtered for blue light excitation (460–487 nm bandpass filter; Semrock). Epifluorescence (emission 500–546 nm bandpass; Semrock) was collected using an ORCA‐FLASH 4.0 LT+ CMOS camera (Hamamatsu, RRID:SCR_021971). LED and camera were controlled and synchronized with the μ‐manager 1.4 software (RRID:SCR_000415). ⁵⁶ After a 10‐ to 30‐min stabilization period in the recording chamber, a focal plane including several fluorescent cell bodies was chosen and acquisitions (100 ms light exposure at 2 Hz for 10–15 min) started. Low‐intensity LED illumination was used (~0.1–0.8 mW) to minimize GCaMP6f photobleaching.

2.3.2. Electrophysiology

KISS1 neuron action potential firing was recorded in brain slices from Kiss1‐hrGFP mice. Green fluorescent protein (GFP)‐expressing RP3V or ARC neurons were first visualized using brief LED illumination (excitation and emission as above), then approached with glass micropipettes under infrared differential interference contrast illumination. Recording micropipettes (tip resistance: 2–6 MΩ) were made with borosilicate glass (catalog #BF150‐110‐7.5, Sutter Instruments), pulled using a Model P‐1000 micropipette puller (Sutter Instruments, RRID:SCR_021042). Action potential firing was recorded in voltage‐clamp mode (no holding potential applied) in the minimally invasive cell‐attached configuration (12–30 MΩ initial seal resistance). Micropipettes were filled with aCSF and the recording configuration was achieved by applying the lowest amount of suction required to detect spontaneous, downward deflections in the current trace (spikes), which correspond to single action potentials. Electrical signals were recorded, low‐pass filtered at 2 kHz and digitized at 20 kHz using a double integrated patch amplifier controlled with the SutterPatch software (Sutter Instruments).

2.4. Drug applications

All drugs were dissolved to the appropriate stock concentration in de‐ionized water, aliquoted and stored at −20°C. Stock solutions were diluted to working concentrations in aCSF prior to performing experiments. For [Ca²⁺]_i imaging, all agonists were bath applied for 2 min after a ≥ three‐minute baseline period. For electrophysiology, agonists were applied locally via a glass micropipette connected to the valve of a picospritzer (NPI electronic GmbH, Germany, RRID:SCR_022657). The pressure micropipette was filled with agonist‐containing aCSF and was placed in the tissue 20 to 50 μm from the recorded cell. Pressure applications (2–14 psi) occurred for 5 s after a ≥10‐s baseline. Because pressure applications typically caused tissue displacement around recorded cells, control experiments in which aCSF was locally applied were carried out to confirm the specificity of the reported effects (see results section).

(S)‐α‐amino‐3‐hydroxy‐5‐methyl‐4‐isoxazolepropionic acid (AMPA, catalog #HB0052), kainic acid (catalog #HB0355) and (R,S)‐2‐amino‐2‐(2‐chloro‐5‐hydroxyphenyl)acetic acid (CHPG, catalog #HB0034) were purchased from HelloBio (RRID:SCR_021047). (S)‐3,5‐dihydroxyphenylglycine (DHPG, catalog#0805) and N‐methyl‐D‐aspartic acid (NMDA, catalog #0114) were purchased from Tocris Bio‐Techne (RRID:SCR_003689).

2.5. Analysis of [Ca²⁺]_i signals

GCaMP6f fluorescence image time‐series were processed in FIJI (RRID:SCR_002285). ⁵⁷ Regions of interest (ROIs) were drawn around individual, in‐focus fluorescent somata. Mean fluorescence intensity within each ROI was measured in each frame. Fluorescence intensity data were analyzed using scripts written in R, versions 4.2.1 and 4.3.1 (RRID:SCR_001905). For each ROI, time‐dependent bleaching was corrected by subtracting the linear downward trend of the mean fluorescence trace. Normalized fluorescence was then calculated as F _t/F × 100, where F is the baseline fluorescence intensity calculated as the mean fluorescence intensity over a 1‐min period preceding agonist applications and F _t is the fluorescence measured at any time point. ROIs in which normalized fluorescence traces reversibly increased from baseline around the time of agonist application were recorded as being excited. For statistical comparisons, agonist effects were calculated as the mean normalized fluorescence over 60 s at the peak of the effect minus that over 60 seconds of baseline.

2.6. Analysis of electrophysiological signals

Analysis of electrophysiological traces was undertaken using Clampfit 10.7 (Molecular Devices, USA, RRID:SCR_011323). Cell‐attached current traces were high‐pass filtered (Bessel 10 Hz cutoff) to remove slow variations in the baseline current. Spikes were then detected using the threshold crossing method and spike time stamps were organized into one‐second bins. To determine if agonists (or aCSF) pressure applications affected KISS1 neuron firing, the mean baseline firing rate was measured during the 5 s preceding applications. Recordings in which the mean firing rate changed by greater than 50% of the mean baseline firing rate during the five‐second application were recorded as displaying a response. Because the effect of DHPG lasted longer than that of iGluR agonists (see results section), the impact of DHPG on firing was assessed during the 10 s following application onset. For statistical comparisons, mean firing rates were calculated during 5 s of baseline and during the 5 s of the application (or 10 s for DHPG).

2.7. Statistical analyses

Statistical analyses were performed using Prism (GraphPad, RRID:SCR_002798). Data are reported in tables as mean ± standard error of the mean (SEM), and in figures as mean ± SEM or ± 95% confidence intervals. Normality of data distributions was estimated using the Shapiro–Wilk test. For electrophysiology, comparisons between two paired groups were made using the paired t‐test (effect of aCSF, AMPA, or kainate pressure application) or the Wilcoxon signed rank test (effect of NMDA pressure application), as appropriate. A Mann–Whitney test was carried out to compare DHPG‐induced changes in firing in the ARC and RP3V. For [Ca²⁺]_i imaging and electrophysiology, comparisons between multiple groups were made with the two‐way ANOVA test (AMPA, kainate, NMDA, DHPG, and CHPG for [Ca²⁺]_i signals; DHPG for electrophysiology) followed by Tukey's multiple comparison tests, as required. Details regarding tests used are given in tables. Sample sizes in statistical tests were the number of slices for [Ca²⁺]_i imaging, and the number of cells for electrophysiology. For the latter, each group includes cells from at least 3 different mice. Differences were considered statistically significant for p < .05.

3. RESULTS

We first used [Ca²⁺]_i imaging in brain slices from adult male and female Kiss1‐Cre::GCaMP6f mice to assess the effects of stimulating GluRs on RP3V^KISS1 and ARC^KISS1 neuron activity (Figure 1). Individual GCaMP6f‐expressing neurons in the RP3V and ARC of males and females displayed large and transient increases in GCaMP6f fluorescence (i.e., [Ca²⁺]_i) upon two‐minute bath applications of GluR agonists (Figure 1Ab,Bb).

FIGURE 1.

Activation of glutamate receptors (GluRs) increases intracellular Ca²⁺ concentration ([Ca²⁺]_i) in rostral periventricular area of the third ventricle (RP3V) and arcuate nucleus (ARC) kisspeptin (KISS1) neurons. (Aa and Ba) Representative RP3V and ARC slices obtained from female and male Kiss1‐Cre::GCaMP6f mice, respectively. White circles and numbers illustrate the positioning of individual regions of interest (ROIs) for analysis. 3V = third ventricle. (Ab and Bb) Example responses to bath applications of N‐methyl‐D‐aspartic acid (NMDA) (50 μM; Ab) and (S)‐3,5‐dihydroxyphenylglycine (DHPG) (50 μM; Bb) for the cells within the ROIs delineated in Aa and Ba. Blue shaded regions indicate the timing of agonist applications.

3.1. iGluR agonists stimulate KISS1 neuron activity

AMPA (10 μM), kainate (5 μM) and NMDA (50 μM) all caused large increases in [Ca²⁺]_i (ranging from ~2.5 to ~47.5% change in normalized GCaMP6f fluorescence on average, depending on agonist, sex, or brain region) in the vast majority of KISS1 neurons (~78%–100% of cells) in the RP3V and ARC of males and females (Figure S1; Table 1). Two cells were inhibited by kainate in the male ARC. Because this was seen in a minority of cases (<5% of cells), it was not explored further. Interestingly, two‐way ANOVA analyses revealed significant main effects of sex (AMPA and kainate) and brain region (NMDA) on responses to these agonists, with a significant interaction between sex and brain region for AMPA responses (Table 1). Post‐tests revealed that the magnitude of AMPA responses showed a substantial sex difference (males < females) in the ARC and a regional difference (ARC < RP3V) in males (Figure 2Aa,Ab; Figure S1; Table 1). No meaningful differences between individual groups were found for kainate or NMDA in post‐tests (Figure 2B,C; Figure S1).

TABLE 1.

Summary of effects of GluR agonists on KISS1 neuron [Ca²⁺]_i.

	Region	Normalized change in fluorescence (%; mean ± SEM)	Two‐way ANOVA results	Proportion of cells excited (%; mean ± SEM)	Two‐way ANOVA results	Number of cells, slices and mice
AMPA	ARC	♂ 14.06 ± 2.87 ♀ 47.40 ± 5.06 a	♂ versus ♀ F (1,42) = 11.730, p = .001; ARC versus RP3V F (1,42) = 0.233, p = .632; interaction F (1,42) = 18.350, p < .001	♂ 100.0 ± 0.0 ♀ 100.0 ± 0.0	No variation	♂ 179,13,10 ♀ 129,10,7
	RP3V	♂ 34.67 ± 5.38 b ♀ 30.96 ± 3.75		♂ 100.0 ± 0.0 ♀ 100.0 ± 0.0		♂ 113,12, 9 ♀ 138,11,8
Kainate	ARC	♂ 3.95 ± 1.73 ♀ 7.86 ± 2.07	♂ versus ♀ F (1,20) = 7.276, p = .014; ARC versus RP3V F (1,20) = 0.217, p = .646; interaction F (1,20) = 1.067, p = .314	♂ 77.8 ± 16.5 ♀ 94.2 ± 3.8	♂ versus ♀ F (1,20) = 0.106, p = .749; ARC versus RP3V F (1,20) = 0.022, p = .883; interaction F (1,20) = 1.290, p = .270	♂ 55 6,5 ♀ 92,6,3
	RP3V	♂ 2.61 ± 0.81 ♀ 11.39 ± 3.77		♂ 88.9 ± 7.0 ♀ 79.8 ± 13.1		♂ 37,6, 4 ♀ 72,6,4
NMDA	ARC	♂ 9.98 ± 2.33 ♀ 13.39 ± 3.14	♂ versus ♀ F (1,26) = 1.633, p = .213; ARC versus RP3V F (1,26) = 5.969, p = .022; interaction F (1,26) = 0.146, p = .705	♂ 98.9 ± 1.1 ♀ 100.0 ± 0.0	♂ versus ♀ F (1,26) = 0.417, p = .524; ARC versus RP3V F (1,26) = 0.417, p = .524; interaction F (1,26) = 0.417, p = .524	♂ 159,11,9 ♀ 108,7,5
	RP3V	♂ 17.83 ± 4.70 ♀ 24.15 ± 5.73		♂ 100.0 ± 0.0 ♀ 100.0 ± 0.0		♂ 55,6,5 ♀ 82,6,5
DHPG	ARC	♂ 6.31 ± 1.90 ♀ 6.08 ± 1.06 c	♂ versus ♀ F (1,27) = 6.263, p = .019; ARC versus RP3V F (1,27) = 5.587, p = .023; interaction F (1,27) = 6.971, p = .014	♂ 98.7 ± 0.9 ♀ 91.3 ± 4.4	♂ versus ♀ F (1,27) = 1.536, p = .226; ARC versus RP3V F (1,27) = 2.751, p = .109; interaction F (1,27) = 1.536, p = .226	♂ 139,9,8 ♀ 133,10,7
	RP3V	♂ 5.94 ± 1.41 d ♀ 14.63 ± 2.21		♂ 100.0 ± 0.0 ♀ 100.0 ± 0.0		♂ 47,6,5 ♀ 64,6,4
CHPG	ARC	♂ 1.10 ± 1.08 ♀ −0.20 ± 0.20	♂ versus ♀ F (1,20) = 0.923, p = .348; ARC versus RP3V F (1,20) = 0.144, p = .709; interaction F (1,20) = 1.656, p = .213	♂ 35.1 ± 14.1 ♀ 4.7 ± 2.2	♂ versus ♀ F (1,20) = 1.131, p = .300; ARC versus RP3V F (1,20) = 0.157, p = .696; interaction F (1,20) = 3.724, p = .068	♂ 66,6,4 ♀ 69,6,4
	RP3V	♂ 0.14 ± 0.17 ♀ 0.33 ± 0.30		♂ 11.5 ± 5.6 ♀ 20.3 ± 13.3		♂ 41,6,4 ♀ 84, 6,4

Abbreviations: AMPA, (S)‐α‐amino‐3‐hydroxy‐5‐methyl‐4‐isoxazolepropionic acid; ARC, arcuate nucleus; CHPG, (R,S)‐2‐amino‐2‐(2‐chloro‐5‐hydroxyphenyl)acetic acid; DHPG, (S)‐3,5‐dihydroxyphenylglycine; [Ca²⁺]_i, intracellular Ca²⁺ concentration; iGluR, ionotropic glutamate receptors; KISS1, kisspeptin; NMDA, N‐methyl‐D‐aspartic acid; RP3V, rostral periventricular area of the third ventricle.

^a q = 7.679, df = 42, p < .001 ♂ ARC versus ♀ ARC;

^b q = 4.989, df = 42, p = .006 ♂ ARC versus ♂ RP3V;

^c q = 5.113, df = 27, p = .006 ♀ ARC versus ♀ RP3V;

^d q = 4.647, df = 27, p = .014 ♂ RP3V versus ♀ RP3V; Tukey's multiple comparison tests.

FIGURE 2.

Activation of ionotropic glutamate receptors (iGluRs) stimulates arcuate nucleus (ARC) and rostral periventricular area of the third ventricle (RP3V) kisspeptin (KISS1) neurons—sex and region differences. (Aa–Ca) Example recordings of individual Kiss1‐Cre::GCaMP6f brain slice responses to (S)‐α‐amino‐3‐hydroxy‐5‐methyl‐4‐isoxazolepropionic acid (AMPA) (10 μM; Aa), kainate (5 μM; Ba), and N‐methyl‐D‐aspartic acid (NMDA) (50 μM; Ca). Thin colored traces are individual cells (AMPA: 15 cells for ♂ ARC and 14 for ♀ ARC; kainate: 12 cells; NMDA: 14 cells). Thick black traces are averages for each slice. Note the sex difference in responses to AMPA in the ARC. Blue shaded regions indicate the timing of iGluR agonist applications. (Ab–Cb) Summary graphs of agonist effects on KISS1 neuron intracellular Ca²⁺ concentration ([Ca²⁺]_i) in the ARC and RP3V of male and female mice. Each data point represents the mean change in normalized GCaMP6f fluorescence across all cells in a single slice. **p = .006 and ***p < .001 Tukey's multiple comparisons tests after two‐way ANOVA. See Table 1 for statistical details.

3.2. Stimulation of group I mGluRs increases KISS1 neuron activity

Activation of group I mGluRs with DHPG (50 μM) also reliably increased [Ca²⁺]_i in male and female ARC^KISS1 and RP3V^KISS1 neurons (ranging from ~5.9 to ~14.6% mean change in normalized GCaMP6f fluorescence in >90% of cells; Figure 3Aa; Figure S1; Table 1). We also note that DHPG inhibited two and four cells in the male and female ARC, respectively, representing <3% of cells. Two‐way ANOVA analysis revealed significant main effects of sex and brain region, and a significant interaction between the two, for responses to DHPG. Post‐tests showed that the magnitude of the DHPG effect was sex‐dependent (male < female) in the RP3V and region‐dependent (ARC < RP3V) in females (Figure 3Aa,Ab; Supplementary Figure S1; Table 1). Because the effect of DHPG may be mediated via activation of mGluR1 and/or mGluR5, we then tested the effect of CHPG—a selective mGluR5 agonist. As seen in Figure 3B, CHPG (100 μM) only had marginal excitatory effects, except in the male ARC (~1.1% mean change in normalized GCaMP6f fluorescence in ~35% of cells; Figure 3B and Table 1), suggesting that the excitatory effect of DHPG might be mainly due to activation of mGluR1. One cell in the male ARC and 4 cells in the female RP3V (<5% of cells for both) were inhibited by CHPG in the male ARC and the female RP3V.

FIGURE 3.

Activation of group I metabotropic glutamate receptors (mGluRs) stimulate arcuate nucleus (ARC) and rostral periventricular area of the third ventricle (RP3V) kisspeptin (KISS1) neurons—sex and region differences. (Aa and Ba) Example recordings of individual Kiss1‐Cre::GCaMP6f brain slice responses to (S)‐3,5‐dihydroxyphenylglycine (DHPG) (50 μM; Aa) in slices of ♀ ARC (13 cells; colored traces) and RP3V (14 cells), and of (R,S)‐2‐amino‐2‐(2‐chloro‐5‐hydroxyphenyl)acetic acid (CHPG) (100 μM; Ba) in a slice of ♂ RP3V (6 cells). Thick black traces are averages for each slice. Note the regional difference in magnitude of DHPG effects and the marginal effect of CHPG. (Ab and Bb) Summary graphs of DHPG (Ab) and CHPG (Bb) effects. *p = .014 and **p = .006 Tukey's multiple comparison tests after two‐way ANOVA. See Table 1 for statistical details. Blue shaded regions indicate the timing of mGluR agonist applications.

In the course of conducting these experiments, we observed that male and female ARC and female RP3V slices displayed short bouts (~5–20 s) of spontaneous, coordinated increases in [Ca²⁺]_i in up to 20 KISS1 neurons (Figure S2) that are reminiscent of those recently described in the ARC, ⁴⁸ , ⁴⁹ suggesting that activity coordination in brain slices might be a common property of KISS1 neurons in the ARC and RP3V. These coordinated events took a variable time to develop, occurring within a few to >60 min of starting a recording. We observed these events in slices treated with AMPA, NMDA, and DHPG but they likely did not directly result from stimulation of GluRs as they occurred at similar rates in untreated slices (not shown).

3.3. GluR activation stimulates action potential firing in KISS1 neurons

We then used loose patch cell‐attached electrophysiology in slices from Kiss1‐hrGFP female mice to validate some of our [Ca²⁺]_i imaging results. Focusing on female RP3V^KISS1 neurons, we examined if GluR agonists evoked changes in KISS1 neuron action potential firing. In control experiments, five‐second local pressure applications of aCSF transiently, but significantly, suppressed RP3V^KISS1 neuron action potential firing by ~30% on average (Table 2 and Figure 4 insets), likely reflecting the effects of mechanical disturbances associated with local pressure applications. As illustrated in Figures 4 and 5A, however, pressure applications of either AMPA (10 μM), kainate (5 μM), NMDA (50 μM) or DHPG (50 μM) significantly increased firing >threefold in the majority (>85%) of RP3V^KISS1 neurons (Tables 2 and 3). The effect of DHPG lasted longer than that of iGluR agonists (Figure 5A inset), likely reflecting the metabotropic nature of group I mGluR signaling. Together, these observations confirm that agonist‐induced increases in [Ca²⁺]_i are associated with increases in action potential firing in KISS1 neurons. We further validated our [Ca²⁺]_i imaging observations by comparing the effect of DHPG on RP3V^KISS1 and ARC^KISS1 neuron firing in females. Two‐way ANOVA analysis of RP3V^KISS1 and ARC^KISS1 neuron firing during baseline and in response to DHPG revealed significant main effects of DHPG and brain region (Table 3). Although DHPG did increase firing in individual female ARC^KISS1 neurons (Figure 5B; Table 3), DHPG‐induced increases in KISS1 firing were statistically significant in the RP3V but not in the ARC, while KISS1 firing in the presence of DHPG was higher in the RP3V than in the ARC (Figure 5C; Table 3). Moreover, DHPG‐induced increases in firing were significantly larger in the RP3V than in the ARC (RP3V^KISS1 neurons: 4.23 ± 0.70 Hz, n = 8; ARC^KISS1 neurons: 0.97 ± 0.44 Hz, n = 7; p = 0.007, U statistic = 5.5; Mann–Whitney test; not illustrated). These observations confirm that the effect of DHPG on female KISS1 neuron activity is region‐specific.

TABLE 2.

Summary of the effect of iGluR agonists on KISS1 neuron action potential firing.

	Region	Baseline firing rate (Hz; mean ± SEM)	Puff firing rate (Hz; mean ± SEM)	Number of cells that changed firing rate (proportion, %)	Results of statistical comparisons	Number of cells (mice)
aCSF control	♀ RP3V	5.49 ± 1.14	3.97 ± 1.14 a	3 inhibited (42.9)	t = 3.341, df = 6, p = .016; paired t‐test	7 (3)
AMPA	♀ RP3V	2.40 ± 0.70	9.34 ± 2.09 b	6 excited (85.7)	t = 3.848, df = 6, p = .009; paired t‐test	7 (5 c )
Kainate	♀ RP3V	3.43 ± 0.75	12.29 ± 3.13 a	7 excited (100)	t = 3.417, df = 6, p = .014; paired t‐test	7 (5)
NMDA	♀ RP3V	1.94 ± 0.64	6.69 ± 2.09 a	6 excited (85.7)	W = 28.000, p = .016; Wilcoxon matched‐pairs test	7 (4 d )

Abbreviation: aCSF, artificial cerebrospinal fluid; AMPA, (S)‐α‐amino‐3‐hydroxy‐5‐methyl‐4‐isoxazolepropionic acid; iGluR, ionotropic glutamate receptors; KISS1, kisspeptin; NMDA, N‐methyl‐D‐aspartic acid; RP3V, rostral periventricular area of the third ventricle.

^a p < .05.

^b p < .01.

^c Including one proestrous mouse.

^d Including two proestrous mice.

FIGURE 4.

Activation of ionotropic glutamate receptors (iGluRs) stimulates rostral periventricular area of the third ventricle (RP3V)^KISS1 neuron electrical activity. Example traces illustrating that local (S)‐α‐amino‐3‐hydroxy‐5‐methyl‐4‐isoxazolepropionic acid (AMPA) (10 μM; A), Kainate (5 μM; B) and N‐methyl‐D‐aspartic acid (NMDA) (50 μM; C) applications increased action potential firing in individual RP3V^KISS1 cells. Insets show the time course of normalized firing in response to local applications of iGluR agonists (mean ± SEM; n = 7 cells each); the response to control artificial cerebrospinal fluid (aCSF) applications (n = 7) is also shown. Blue shaded regions illustrate the timing of local agonist applications.

FIGURE 5.

Activation of group I metabotropic glutamate receptors (mGluRs) stimulates kisspeptin (KISS1) neuron firing. (A) Example trace of the effect of (S)‐3,5‐dihydroxyphenylglycine (DHPG) (50 μM) on a female rostral periventricular area of the third ventricle (RP3V)^KISS1 neuron action potential firing. Inset shows the normalized firing time course in response to DHPG or to artificial cerebrospinal fluid (aCSF) (means ± SEM; n = 8 and 7 cells, respectively). (B) Traces of the DHPG‐evoked increase in firing in two different female arcuate nucleus (ARC)^KISS1 neurons. Inset shows the time course of normalized firing in response to DHPG (mean ± SEM; n = 7). Blue shaded regions illustrate the timing of local DHPG applications. (C) Summary graph of KISS1 neuron mean firing rate in response to DHPG in the ARC and RP3V of female mice. **p = .003, ^## p = .009, Tukey's multiple comparison tests after two‐way ANOVA.

TABLE 3.

Summary of the effect of DHPG on KISS1 neuron action potential firing.

	Region	Baseline firing rate (Hz; mean ± SEM)	Puff firing rate (Hz; mean ± SEM)	Number of cells that changed firing rate (proportion, %)	Two‐way ANOVA results	Number of cells (mice)
DHPG	♀ RP3V	1.78 ± 0.50	6.00 ± 0.43 a	7 excited (87.5)	Baseline versus DHPG, F (1,26) = 10.710, p = .003 RP3V versus ARC, F (1,26) = 8.304, p = .008 Interaction, F (1,26) = 4.200, p = .051	8 (4)
	♀ ARC	1.11 ± 0.86	2.09 ± 1.27 b	6 excited (85.7)		7 (4)

Abbreviations: ARC, arcuate nucleus; DHPG, (S)‐3,5‐dihydroxyphenylglycine; KISS1, kisspeptin; RP3V, rostral periventricular area of the third ventricle.

^a q = 5.509, df = 26, p = 0.003, RP3V baseline versus DHPG;

^b >q = 4.931, df = 26, p = 0.009, RP3V DHPG versus ARC DHPG; Tukey's multiple comparison tests.

4. DISCUSSION

We report here that selective iGluR and group I mGluR agonists consistently increase KISS1 neuron [Ca²⁺]_i in the RP3V and ARC of male and female mice. In the RP3V, we show that activating iGluRs and group I mGluRs also increases KISS1 neuron action potential firing. In addition, we observed sex‐ and region‐specific differences in KISS1 neuron [Ca²⁺]_i responses to GluR agonists, notably with AMPA in the ARC having larger effects in females than in males and DHPG in females having a larger effect in the RP3V than in the ARC. We validated this latter observation by providing evidence that the effect of DHPG on KISS1 neuron firing is region‐specific in females.

Recent RNAseq data indicate that KISS1 neurons express genes encoding all iGluR and most mGluR subtypes. ³⁷ , ³⁸ , ³⁹ Previous observations revealed that activating AMPA and NMDA receptors increases activity in KISS1 neurons. ⁴⁰ , ⁴² , ⁵⁸ Here, we expand upon these previous findings by providing new insights into KISS1 neuron GluR make‐up, with evidence that these cells may express AMPA, NMDA, kainate and group I mGluR receptors and that their activation increases KISS1 neuron [Ca²⁺]_i and action potential firing. In addition, our [Ca²⁺]_i imaging data indicates that the effect of activating group I mGluRs is primarily mediated by mGluR1, except in the male ARC where mGluR5 may have a small contribution. Seeing that the effects of some GluR agonists on LH secretion require intact KISS1 signaling, our findings together provide a potential mechanism through which these agonists stimulate LH secretion in vivo. ¹⁶ , ¹⁷ , ¹⁹ , ³⁴ , ³⁵ , ³⁶ , ⁵⁹

Our observations that local agonist applications evoked KISS1 neuron firing suggest that the effects of GluR agonists we report here are likely direct, although additional indirect effects are possible. Ruling out these possible indirect effects would require recordings carried out under conditions isolating KISS1 neurons from synaptic inputs. Agonist‐induced increases in [Ca²⁺]_i may result from recruitment of multiple sources of Ca²⁺ ions. As NMDA receptors and subtypes of AMPA and kainate receptors are Ca²⁺‐permeable, ⁶⁰ , ⁶¹ , ⁶² Ca²⁺ influx through these receptors may contribute to the signals we observed. In addition, because the agonists we tested evoked KISS1 neuron action potential firing, Ca²⁺ influx through voltage‐gated Ca²⁺ channels (VGCC) may also contribute to these signals. Lastly, it is also possible that Ca²⁺ release from internal stores plays a role, particularly in response to group I mGluR activation.

Several reports have documented that KISS1 neurons in the RP3V and ARC exhibit spontaneous fast excitatory PSCs (sEPSCs). ⁴⁴ , ⁴⁵ , ⁴⁶ The respective role of AMPA, kainate and NMDA receptors in mediating these events remains unclear, although the contribution of NMDA receptors is likely limited in conditions under which sEPSCs are typically recorded (i.e., near the resting membrane potential, at which NMDA receptors conduct little current). Based on our findings and on the demonstrated roles of synaptic kainate and NMDA receptors elsewhere in the brain [reviewed in ⁶⁰ , ⁶² ], it is possible that these receptors contribute to glutamatergic synaptic transmission along with AMPA receptors in KISS1 neurons, at least under certain conditions. Regarding group I mGluRs, whether these receptors can be activated by endogenous glutamate in KISS1 neurons is, to our knowledge, unknown. It is likely that their activation by synaptic glutamate release requires sustained presynaptic activity as it does in other brain areas. ⁶³ , ⁶⁴

Our [Ca²⁺]_i imaging experiments revealed sex differences in the effects of some GluR agonists on KISS1 neurons. In the ARC, AMPA caused larger increases in KISS1 neuron activity in females than in males. This difference might result from male and female ARC^KISS1 neurons expressing different levels of membrane AMPA receptors. Additionally, it is possible that female ARC^KISS1 neurons express more of the Ca²⁺‐permeable, GLUA2‐lacking AMPA receptors ⁶¹ than male ARC^KISS1 neurons, thereby contributing to larger [Ca²⁺]_i responses to AMPA. The observed difference may also result from sex differences in VGCC expression and/or in [Ca²⁺]_i handling mechanisms (i.e., Ca²⁺ storage and/or release). This latter possibility seems somewhat less likely as sex differences were not observed in the ARC with all agonists. Further investigations will be needed to determine the origin and, importantly, functional significance of the observed sex difference in responses to AMPA. Interestingly, activation of AMPA receptors by endogenous glutamate was recently suggested to play a role in coordinating activity within the ARC^KISS1 neuron network. ⁴⁸ , ⁴⁹ Whether or not the sex difference we report here has an impact on this pattern of ARC^KISS1 neurons remains to be seen, although we note that AMPA receptor‐mediated ARC^KISS1 neuron activity coordination is seen in females and in males. ⁴⁸ , ⁴⁹ We also observed sex differences in DHPG effects in the RP3V, with greater responses in females than in males. As discussed previously, ¹¹ , ⁵³ the poor correspondence between Cre‐dependent fluorescent reporters and KISS1 immunoreactivity in the adult male RP3V of KISS1‐Cre driver lines ⁶⁵ , ⁶⁶ increases uncertainty around the identity of GCaMP6f‐expressing neurons in the male RP3V. The fact that we cannot rule out that subpopulations of the GCaMP6f‐expressing neurons we recorded in the male RP3V are not Kiss1‐expressing cells somewhat limits the interpretative value of the observed sex difference. It is, however, tempting to draw a parallel with the female‐specific role of RP3V^KISS1 neurons in the rodent preovulatory LH surge. Recordings made from fully identified KISS1 neurons in the male RP3V will be required to unequivocally answer this question.

Specific agonists also displayed regional differences in their impact on KISS1 neuron activity. This is the case, for example, of AMPA in males, which had larger effects in the RP3V than in the ARC. Here again, the caveats associated with recording KISS1 neuron [Ca²⁺]_i in the male RP3V prevent any conclusions from being made. DHPG, however, displayed interesting region‐specific effects. DHPG‐induced increases in KISS1 neuron [Ca²⁺]_i and action potential firing were substantially larger in the female RP3V than in the female ARC. As is the case for sex differences in AMPA responses in the ARC, multiple factors may explain these observations, including regional differences in mGluR1 expression levels and/or in mGluR1 coupling with ion channels and intracellular Ca²⁺ stores. As female ARC^KISS1 neurons have lower resting membrane potentials and lower firing rates than their RP3V counterparts [reviewed in ¹¹ ], regional differences in KISS1 neuron excitability may also play a role. This latter possibility, however, seems unlikely as regional differences were not as pronounced for the effects of NMDA and were not seen for the effects of kainate on KISS1 neuron activity. The role of activation of group I mGluRs by endogenous glutamate in regulating KISS1 neuron activity patterns and, indeed, whether the stronger effect of DHPG on KISS1 neuron activity seen in the RP3V vs the ARC impacts the regulation of LH secretory patterns remains to be seen.

In summary, our findings reveal that GluR activation promotes KISS1 neuron activity, increasing [Ca²⁺]_i and action potential firing. This work also uncovered potential sex and regional differences in signaling by certain GluR subtypes in KISS1 neurons, which could play a role in KISS1 neuron‐dependent regulation of LH secretory patterns.

AUTHOR CONTRIBUTIONS

Robin J. Bearss: Data curation; investigation; formal analysis; writing – review and editing; visualization. Isabella A. Oliver: Data curation; investigation; formal analysis. Peighton N. Neuman: Data curation; investigation; formal analysis. Wahab I. Abdulmajeed: Data curation; investigation; formal analysis. Jennifer M. Ackerman: Investigation. Richard Piet: Conceptualization; data curation; investigation; funding acquisition; formal analysis; visualization; writing – review and editing; writing – original draft; project administration; supervision.

FUNDING INFORMATION

This work was supported by NICHD grant R01HD109337 to Richard Piet.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

PEER REVIEW

The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer-review/10.1111/jne.13456.

Supporting information

Figure S1. Effects of glutamate receptor agonists on KISS1 [Ca²⁺]_i in the ARC and RP3V of male and female mice. Average traces (black lines) ± 95% confidence interval (gray lines) of all cells in each group are displayed. The blue shaded regions illustrate the application of agonists. Agonist concentrations were 10, 5, 50, 50 and 100 μM for AMPA, kainate, NMDA, DHPG, and CHPG, respectively. Responses to DHPG and CHPG are displayed on the same scale to highlight differences in their effects. Spikes in some of the average traces are due to coordinated [Ca²⁺]_i increases occurring spontaneously in some recordings.

Figure S2. Coordinated activity in KISS1 neurons. Example traces of three different KISS1 neurons displaying spontaneous coordinated increases in [Ca²⁺]_i in a brain slice including the ARC obtained from a male mouse (A) and in a brain slice including the RP3V obtained from a female mouse (B). Insets show single coordinated events on a shorter time scale. Slices were treated with DHPG 45 min prior to these recordings. Traces were processed with the FluoroSNNAP software ⁶⁷ to remove slow variations in the fluorescence traces and analyze coordinated events.

Bearss RJ, Oliver IA, Neuman PN, Abdulmajeed WI, Ackerman JM, Piet R. Activation of ionotropic and group I metabotropic glutamate receptors stimulates kisspeptin neuron activity in mice. J Neuroendocrinol. 2025;37(1):e13456. doi: 10.1111/jne.13456

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.