Neuromedin B Excites Central Lateral Amygdala Neurons and Reduces Cardiovascular Output and Fear-Potentiated Startle

Abstract

Neuromedin B (NMB) and gastrin-releasing peptide (GRP) are the two mammalian analogs in the bombesin peptide family that exert a variety of actions including emotional processing, appetitive behaviors, cognition, and tumor growth. The bombesin-like peptides interact with three receptors: the NMB-preferring bombesin 1 (BB1) receptors, the GRP-preferring bombesin 2 (BB2) receptors and the orphan bombesin 3 (BB3) receptors. Whereas injection of bombesin into the central amygdala reduces satiety and modulates blood pressure, the underlying cellular and molecular mechanisms have not been determined. As administration of bombesin induces the expression of Fos in the lateral nucleus of the central amygdala (CeL) which expresses BB1 receptors, we probed the effects of NMB on CeL neurons using in vitro and in vivo approaches. We showed that activation of the BB1 receptors increased action potential firing frequency recorded from CeL neurons via inhibition of the inwardly rectifying K⁺ (Kir) channels. Activities of phospholipase Cβ and protein kinase C were required, whereas intracellular Ca²⁺ release was unnecessary for BB1 receptor-elicited potentiation of neuronal excitability. Application of NMB directly into the CeA reduced blood pressure and heart rate and significantly reduced fear-potentiated startle. We may provide a cellular and molecular mechanism whereby bombesin-like peptides modulate anxiety and fear responses in the amygdala.

1. Introduction

Neuromedin B (NMB) and gastrin-releasing peptide (GRP) are the mammalian analogues of the bombesin and bombesin-like family of peptides originally characterized in amphibians and first isolated from porcine gastric tissue and spinal cord (Minamino, Kangawa, & Matsuo, 1983). The biological actions of bombesin-like peptides are mediated by at least three receptors: the NMB-preferring bombesin 1 (BB1) receptors, the GRP-preferring bombesin 2 (BB2) receptors (González et al., 2009; Ohki-Hamazaki, 2000) and the orphan bombesin 3 (BB3) receptors whose natural ligand is unknown (N. Gonzalez, Moreno, & Jensen, 2015; M. Li et al., 2019). All three bombesin receptors are coupled to G proteins resulting in the activation of phospholipase Cβ (PLCβ) which hydrolyzes phosphatidylinositol 4,5-bisphosphate (PIP₂) generating inositol 1,4,5-trisphosphate (IP₃) to stimulate intracellular Ca²⁺ release and diacylglycerol (DAG) to activate protein kinase C (PKC) (Nieves Gonzalez, Moody, Igarashi, Ito, & Jensen, 2008; N. Gonzalez et al., 2015; Shapira, Way, Lipinsky, Oron, & Battey, 1994), although the bombesin receptors also stimulate tyrosine phosphorylation of a number of signaling proteins (Nieves Gonzalez et al., 2008; Ramos-Álvarez et al., 2015). BB1 receptor activation regulates pituitary-thyroid function, fear and anxiety responses, satiety and tumor growth; BB2 receptors play important roles in modulating pruritus, lung development, small intestinal mucosal defense and CNS processes such as learning and memory; BB3 receptors are involved in controlling obesity and glucose intolerance, lung response to injury, tumor growth and gastrointestinal tract motility (Nieves Gonzalez et al., 2008; Ramos-Álvarez et al., 2015). However, the cellular and molecular mechanisms whereby the bombesin-like peptides modulate these physiological functions and pathological disorders have not been fully determined.

The amygdala is a critical structure involved in regulating emotional processing (Dejean et al., 2015; Janak & Tye, 2015; Ressler, 2010; Tye et al., 2011), pain (Neugebauer, 2015; Neugebauer, Li, Bird, & Han, 2004; Veinante, Yalcin, & Barrot, 2013), substance abuse disorders (Gilpin, Herman, & Roberto, 2015; Silberman, Shi, Brunso-Bechtold, & Weiner, 2008; Warlow, Robinson, & Berridge, 2017) and appetitive behaviors (Petrovich, 2011, 2013; Smith & Lawrence, 2018; Zanchi et al., 2017). The amygdala is comprised of the lateral amygdala (LA), the basolateral amygdala (BLA), and the central amygdala (CeA). The GABAergic medial intercalated cells (ITCs) (Ehrlich et al., 2009) separate the BLA from the CeA (Nitecka & Ben-Ari, 1987). Whilst the glutamatergic pyramidal neurons are the principal neurons in the LA and BLA, GABAergic neurons are the primary neuronal type in the CeA which consists of 3 subnuclei named as capsular (CeC), lateral (CeL), and medial (CeM) nucleus of CeA (Joseph E. LeDoux, 2000). The LA receives and integrates multisensory information from the thalamus (J. E. LeDoux, Cicchetti, Xagoraris, & Romanski, 1990; Tully, Li, Tsvetkov, & Bolshakov, 2007), the cortex (McDonald, 1998) and the brainstem (Johansen, Cain, Ostroff, & LeDoux, 2011) which also projects to the CeA (Bernard, Huang, & Besson, 1992; Neugebauer, Galhardo, Maione, & Mackey, 2009). The principal neurons in the LA make glutamatergic synapses onto glutamatergic neurons in the BLA and GABAergic neurons in the medial ITCs (Ehrlich et al., 2009) as well as the CeL. The BLA then sends glutamatergic projections to the CeL, the CeM (Krettek & Price, 1978; A. Pitkänen et al., 1995; Savander, Go, LeDoux, & Pitkänen, 1995) and the ITCs which further innervate CeA neurons (Royer, Martina, & Pare, 1999). The CeL and CeM also receive GABAergic afferents from other structures (Le Gal LaSalle, Paxinos, & Ben-Ari, 1978) and contain local GABAergic interneurons to inhibit each other via axon collaterals (H.-C. Pape & D. Pare, 2010) and GABAergic projection neurons to relay information out of the amygdala (McDonald & Augustine, 1993; Paré & Smith, 1993). The CeL projects to the CeM, with no reciprocal projection from the CeM to the CeL (A Pitkänen, 2000). The CeM is the major output nucleus of the amygdala and projects to the structures involved in emotional (Hopkins & Holstege, 1978; H. C. Pape & D. Pare, 2010; A Pitkänen, 2000) and autonomic control (Galeno & Brody, 1983; Iwata, Chida, & LeDoux, 1987; Mogenson & Calaresu, 1973; Viviani et al., 2011), although the CeL sends GABAergic projections to behavioral and physiologic effector regions as well (Penzo, Robert, & Li, 2014).

The CeA expresses not only bombesin-like peptides (Moody, Getz, O'Donohue, & Rosenstein, 1988) and NMB mRNA (Wada, Way, Lebacq-Verheyden, & Battey, 1990), but also high densities of binding sites for bombesin-like peptides (Moody et al., 1988) and NMB (M. C. Lee, Jensen, Coy, & Moody, 1990). Consistent with the distribution of bombesin-like peptides and their receptors in the CeA, injection of bombesin into the CeA induces an increase in mean arterial pressure (Brown & Gray, 1988) and reduces food intake (Kyrkouli, Stanley, & Leibowitz, 1987; Vigh, Lenard, Fekete, & Hernadi, 1999). Intraperitoneal or cerebroventricular administration of bombesin induces Fos-like immunoreactivity, a marker of neuronal activation, in the CeL (B. H. Li & Rowland, 1996). However, the effects of bombesin and bombesin-like peptides on neuronal excitability in the CeA and the underlying cellular and molecular mechanisms have not been determined. Here, we studied the effects of NMB on the excitability of CeL neurons based on the results showing that high densities of NMB binding sites are distributed in the CeA (M. C. Lee et al., 1990) and administration of bombesin induces Fos expression in the CeL suggesting that activation of bombesin receptors increases neuronal activity in the CeL. Our results indicate that activation of BB1 receptors by NMB excites CeL neurons by PLCβ and PKC-mediated depression of inwardly rectifying K⁺ (Kir) channels. Moreover, microinjection of NMB into the CeA reduced blood pressure and heart rate in unanesthetized male and female rats and reduced fear-potentiated startle responses. Our results may provide a cellular and molecular mechanism to explain the roles of bombesin and bombesin-like peptides in the amygdala.

2. Materials and Methods

2.1. Preparation of amygdala slices

Coronal brain slices (300 μm) were prepared from virgin male and female Sprague-Dawley (SD) rats (22-38 days old), as described previously (C. A. Boyle, Hu, Quaintance, & Lei, 2021; Binqi Hu, Boyle, & Lei, 2020). The number of males and females for each experiment was kept as equal as possible. Rats were deeply anesthetized with isoflurane and subsequently decapitated for brain dissection. The cerebellum was trimmed, and the caudal pole of the brain was glued to the plate of a vibratome (Leica VT1200S) and then bathed in an ice-cold solution that contained (in mM) 130 N-methyl-D-glucamine (NMDG)-Cl, 24 NaHCO₃, 3.5 KCl, 1.25 NaH₂PO₄, 0.5 CaCl₂, 8.0 MgCl₂ and 10 glucose, saturated with 95% O₂ and 5% CO₂ (pH 7.4, adjusted with HCl). Cuts began at the rostral pole and slices were collected from both hemispheres when the structure of the amygdala was apparent. Slices were then incubated in the above solution except NMDG-Cl was replaced with NaCl at 35°C for 1 h for recovery and kept at room temperature (~22°C) until use. All procedures and experiments presented in this study were approved by the Institutional Animal Use and Care Committee of the University of North Dakota and performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

2.2. Recordings of action potentials, resting membrane potentials and holding currents from amygdala neurons

Whole-cell patch-clamp recordings using a Multiclamp 700B amplifier (Molecular Devices, Sunnyvale, CA) in current- or voltage-clamp mode were conducted from the CeL neurons (Figure 1, a) visually identified with infrared video microscopy (Olympus BX51WI) and differential interference contrast optics. The extracellular solution contained (in mM) 130 NaCl, 24 NaHCO₃, 3.5 KCl, 1.25 NaH₂PO₄, 2.5 CaCl₂, 1.5 MgCl₂ and 10 glucose, saturated with 95% O₂ and 5% CO₂ (pH 7.4). Recording pipettes were filled with the internal solution containing (in mM) 120 K⁺-gluconate, 10 KCl, 5 NaCl, 2 MgCl₂, 10 HEPES, 0.5 EGTA, 2 ATPNa₂, 0.4 GTPNa, and 5 phosphocreatine (pH 7.3). Data were filtered at 2 kHz, digitized at 10 kHz, acquired and analyzed subsequently using pClamp 10.4 software (Molecular Devices, Sunnyvale, CA). For action potential (AP) recordings, the aforementioned external solution was supplemented with kynurenic acid (1 mM) to block glutamatergic transmission and picrotoxin (100 μM) to block GABAergic transmission. Frequency of APs was measured by Clampfit 10.7 with “Event Detection” and “Threshold Search”. The data were then output to Excel and binned per minute with a custom formula in Excel. The resting membrane potentials (RMPs) or holding currents (HCs) at −60 mV were recorded in the extracellular solution supplemented with tetrodotoxin (TTX, 0.5 μM), kynurenic acid (1 mM), and picrotoxin (100 μM). Pharmacological inhibitors were applied to the cells either extracellularly or intracellularly via the recording pipettes. For extracellular application, slices were pretreated for at least 2 h to ensure permeation of reagents into the cells in the slices and the extracellular solution continuously contained the same concentration of the reagents, unless stated otherwise. For intracellular application, we waited for >15 min after the formation of whole-cell configuration to ensure the diffusion of the inhibitors into the cells. To prevent agonist-induced receptor desensitization, each slice was limited to a single application of NMB.

Figure 1. NMB increases AP firing frequency in CeL neurons.

(a), microscopic photograph showing the location of CeL where electrophysiological recordings were conducted. LA: lateral nucleus; BLA: basolateral nucleus; CeC: capsular central amygdala; CeL: lateral central amygdala; CeM: medial central amygdala. Scale bar: 500 μm. (b), bath application of NMB (0.3 μM) increased AP firing frequency in regular spiking (RS) neurons in the CeL region. Upper panel: voltage responses elicited by injection of currents from −30 pA to 20 pA at an increment of 10 pA in a duration of 400 ms. Lower panel: current traces recorded from the same neuron prior to, during and after the application of NMB. The firing frequency of APs in RS neurons was increased by a 5-minute bath application of NMB. (c), bath application of NMB (0.3 μM) increased AP firing frequency in low-threshold bursting (LTB) neurons of the CeL region. The firing frequency of APs in LTB neurons was increased by a 5-minute bath application of NMB. (d), bath application of NMB (0.3 μM) increased AP firing frequency in late firing (LF) neurons in the CeL region. The firing frequency of APs in LF neurons was increased by a 5-minute bath application of NMB. (e), summarized time course of NMB-induced potentiation of AP firing frequency recorded from 33 CeL neurons.

2.3. Recording of Kir currents from CeL neurons

The aforementioned intracellular K⁺-gluconate solution was used to record Kir currents from CeL neurons. We replaced the CaCl₂ in the extracellular solution with the same concentration of MgCl₂ and included TTX (0.5 μM) to reduce the contaminations of voltage-gated Ca²⁺ and Na⁺ channels, respectively. CeL neurons were held at −60 mV and stepped from −140 to −40 mV for 400 ms at a voltage increment of 10 mV every 10 s. Steady-state current values were measured within 5 ms prior to the end of the step voltage protocol.

2.4. Stereotaxic surgery, microinjection, and histology

Surgical procedures and cannula placement were performed according to our previous publications (Cody A. Boyle, Hu, Quaintance, Mastrud, & Lei, 2022; Deng et al., 2009; Xiao et al., 2014; Xiao et al., 2012) under aseptic conditions. Male (368 ± 10.4 g, n = 17) and female (225 ± 5.8 g, n = 20) SD rats were deeply anaesthetized using vaporized isoflurane and secured in a stereotaxic frame (Stoelting Co., IL). Guide cannula (23 GA, 8.5 mm length; P1 Technologies Inc., Roanoke, VA) were bilaterally implanted targeting the CeA with coordinates obtained from the rat brain atlas (from bregma: −2.1-2.3 mm, mediolateral: ± 4.1-4.2 mm, dorsoventral: −7.6 mm (Paxinos & Watson, 2007). The end of the guide cannula was positioned 0.5 - 1 mm above the CeA. A small incision was made from the base of the skull extending forward toward the front of the skull exposing the suture lines. Cannulas were secured with dental cement bonded to 3 stainless steel screws (4.8 mm, P1 Technologies Inc., Roanoke, VA) inserted in burr holes drilled into the skull. To prevent obstruction of the guide cannula, dummy cannulas were screwed into each guide cannula until microinjections were performed. Following surgery, animals were monitored for adverse post-surgical complications and allowed a 7–10 d recovery period. During this recovery period, animals received daily handling to simulate the microinjection process. Saline, NMB, TQ, and BIM23042 (all in 1 μl volume per side) were bilaterally injected into the CeA through an internal cannula (30 GA, 8.5 mm; P1 Technologies Inc., Roanoke, VA). Two 5 mL Hamilton syringes fixed to an automated pump (Harvard Apparatus, MA) were used to administer drug injections at a rate of 0.2 μl min ⁻¹. Following experiments, rats were anesthetized and 80 μm serial sections were collected on a vibrating microtome (VT1200, Leica Biosystems Inc., IL) to confirm cannula placement. Animals with incorrect cannula placement were excluded from analysis.

2.5. Acoustic startle and Fear-potentiated startle responses

Acoustic startle response (ASR) and fear-potentiated startle (FPS) experiments were conducted using two identical SR-LAB startle chambers with cylindrical animal holders (San Diego Instruments, CA) as previously described (Cody A. Boyle et al., 2022). White noise bursts (90 dB, 95 dB, and 105 dB WNBs) were elicited from a high-frequency loudspeaker mounted 24 cm above each animal holder within each cabinet. A visual conditioned stimulus was produced by a single light bulb in the ceiling of each cabinet. Foot shocks were used as a conditioned stimulus and were delivered from a stainless-steel floor grind in each animal holder. Experimental protocols were designed and implemented via the SR-LAB software. The experimental paradigm occurred over a 4-day period as shown below (Figure 8, b). On Day 1, animals were moved in their home cage to the behavioral testing room for a 1 h handling session followed by placement inside the animal holder within each startle chamber for 30 minutes of habituation. On Day 2, animals underwent a 5-minute habituation period and were presented with 21 WNBs (7- 90 dB; 7- 95 dB; 7- 105 dB; 50 ms duration) at a 30 s intertrial interval (ITI) for startle responses in the presence of a background white noise (70 dB). A subset of male and female rats was microinjected with saline or NMB (0.3 nmol) 15 min prior to ASR testing. On Day 3, to test the involvement of BB1 receptor activation on FPS, test compounds were microinjected 10 minutes before fear conditioning, unless otherwise stated. Like previous days 1 and 2, animals were allowed a 5-minute habituation. Animals were presented with a series of ten 3.7 s light cues (neutral stimuli) that co-terminated with a 0.5 s foot shock (0.5 mA, pseudorandom ITI 30-180 s). There was no background noise present during the fear conditioning trials. On Day 4, fear acquisition testing was performed using FPS. Following a 5-minute habituation period within the startle chambers, the animals were exposed to 30 WNB trials in the presence of a continuous 70 dB white noise background. The startle amplitude was specified as the maximum peak voltage recorded within the first 200 ms of the response to the WNB. The first 10 WNBs (5- 90 dB; 5- 95 dB) were used to determine basal startle amplitude in the presence of background noise without the light cue. The following 20 WNBs were split into groups of 10 WNBs in the presence of white noise background: 10 WNBs (105 dB) were paired with the light cue (cued), 10 WNBs (105 dB) without (non-cued) in a pseudorandom order. FPS responses were reported as raw startle responses as previously reported (Ayers, Missig, Schulkin, & Rosen, 2011; Missig, Ayers, Schulkin, & Rosen, 2010). Responses to the 0.5 mA foot shock recorded during fear conditioning were the maximum peak voltage.

Figure 8. Schematic diagram showing the cardiovascular recording, ASR and FPS procedures and cannula placement.

(a), For recording cardiovascular parameters, animals received daily handling to habituate them to the experimental procedure for at least 7 days. Baseline measurements were recorded for 2 days. Cardiovascular parameters were recorded during and for at least 15 min after the injection (top). (b), In the acoustic startle response (ASR) experiments, rats were given intra-CeA saline (1 μL in each side) or NMB (0.3 nmol in 1 μL) 10 mins prior to the test session (middle). In fear-potentiated startle (FPS) experiments (bottom), rats were given intra-CeA saline or NMB 10 mins prior to the fear conditioning session. FPS measured as cued or non-cued fear responses were tested 24 h later. (c), cannula tip placement in a subset of experiments displayed onto an atlas figure adapted from Paxinos and Watson (Paxinos & Watson, 2007); White arrow denotes the cannula tip placement.

2.6. Recordings of mean arterial pressure and heart rate

Measures of cardiovascular activity were recorded from age-matched male and female SD rats using the VPR noninvasive blood pressure monitoring system (CODA-6, Kent Scientific, Torrington, CT) connected to a Powerlab system (ADInstruments Pty Ltd., Bella Vista, New South Wales, Australia) (Figure 8, a). Measurements obtained with the VPR noninvasive blood pressure monitoring system provide an accurate measure comparable to more invasive techniques (Feng et al., 2008). The numbers of male and female rats were kept as equal as possible. Measurements were taken during the light cycle and SD rats were habituated for at least 7 days via light handling in a clean towel (Lipták, Kaprinay, & Gáspárová, 2017). To improve recording signal, animals were placed on a warming pad prior to data collection and care was taken to not overheat nor over-restrict the movement of the animals during blood pressure monitoring. Baseline cardiovascular parameters (systolic, diastolic, mean arterial pressure, and heart rate) were recorded for at least a 10-minute habituation period during each session. A series of measurements were recorded during and immediately following drug injection and data were reported as the net change in mean arterial pressure (MAP) and heart rate (HR) from baseline as a result of saline or drug injection, as previously reported (Hem, Phie, Chilton, & Kinobe, 2019; Lipták et al., 2017; Luo et al., 2017).

2.7. Data analysis

Data are presented as the means ± S.E.M. For the in vitro data, “n” refers to the number of cells recorded, whereas during the in vivo experiments, “n” refers to the number of animals tested. Wilcoxon matched-pairs signed-rank test (abbreviated as Wilcoxon test in the text), Mann-Whitney test, one-way analysis of variance (ANOVA) followed by Dunnett's multiple comparisons test or two-way repeated-measures ANOVA followed by Šídák’s multiple comparison tests were used as appropriate for statistical analysis; P values were reported throughout the text and significance was set as P < 0.05. ASR, FPS, and baseline cardiovascular responses were analyzed using a mixed-effects two-way ANOVA, as appropriate and significant findings were analyzed post hoc using the Šídák’s multiple comparisons test. Shock reactivity was analyzed using the nonparametric Mann-Whitney test. Net changes in cardiovascular values were compared using a one-way ANOVA and post-hoc differences were assessed using Dunnett’s or Tukey’s multiple comparisons test. The IC₅₀ value was calculated from raw data measurements using non-linear regression. Statistical analysis was conducted using GraphPad Prism Version 9 or Origin 7.

2.8. Chemicals

The following chemicals were purchased from R&D Systems: NMB, kynurenic acid, picrotoxin, U73122, U73343, heparin, thapsigargin, BAPTA, chelerythrine, bisindolylmaleimide II (Bis II), phorbol 12-myristate 13-acetate (PMA), TTX, ML 133, tertiapin-Q (TQ), tertiapin-LQ, BIM23042, and PD168368. Stock solutions of drugs were prepared and aliquoted for storage at −20°C until use. For chemicals dissolved in dimethyl sulfoxide (DMSO), the final concentration of DMSO was less than 0.1%.

3. Results

3.1. NMB increases neuronal excitability in the CeL

We chose NMB as an example to test the effects of bombesin-like peptides on neuronal excitability in the CeL (Figure 1, a) based on the results showing that the CeA expresses both NMB mRNA (Wada et al., 1990) and NMB binding sites (M. C. Lee et al., 1990) and administration of bombesin induces Fos expression in the CeL (B. H. Li & Rowland, 1996), suggesting that activation of bombesin receptors facilitates neuronal activity. CeL neurons can be classified electrophysiologically into three types: regular spiking (RS, ~54%), low-threshold bursting (LTB, ~34%), and late firing (LF, ~12%) (Amano, Amir, Goswami, & Paré, 2012). We thus identified the types of the recorded neurons first after formation of the whole cell recording configuration by injection of a series of negative and positive currents. As neurons in the CeL did not show spontaneous firing, we injected a persistent positive current to induce discrete AP firing. We examined 33 neurons of which 23 were RS, 7 were LTB, and 3 were LF. Bath application of the selective BB1 receptor agonist, NMB at 0.3 μM, a near-saturating concentration (Viviani et al., 2011), augmented the firing frequency of APs recorded from each neuronal type (Figure 1, b-d). We therefore pooled the data recorded from each neuronal type. NMB increased the AP firing frequency to 274 ± 38% of control (Control: 0.51 ± 0.08 Hz, NMB: 1.11 ± 0.15 Hz, n = 33, P < 0.0001, Wilcoxon test, Figure 1, e), suggesting that activation of BB1 receptors excited CeL neurons. NMB-elicited augmentation of AP firing frequency was irreversible after a wash for 30 minutes (Control: 0.51 ± 0.08 Hz, wash 30 min: 1.38 ± 0.34 Hz, n = 33, P = 0.0003, Wilcoxon test, Figure 1, e). The NMB-mediated increases in AP firing frequency were sex-independent as NMB exerted similar extent of increases (P = 0.15, Two-sample t-test) in males (248 ± 53% of control, n = 17) and female (434 ± 117% of controls, n = 16) virgin rats. We thus used both sexes and kept the numbers of male and female rats as equal as possible for the remaining experiments.

3.2. NMB-induced excitation of CeL neurons requires the functions of G proteins and PLCβ

We then sought to determine the intracellular signaling mechanisms involved in NMB-mediated excitation of CeL neurons. BB1 receptors are coupled to Gα_q proteins resulting in the activation of PLCβ which hydrolyzes PIP₂ to generate IP₃ to increase intracellular Ca²⁺ release and DAG to activate PKC (Shapira et al., 1994). Thus, we examined the roles of this signaling pathway in NMB-mediated facilitation of neuronal excitability in the CeL. Intracellular perfusion of the selective G protein inactivator GDP-β-S (0.5 mM) prevented NMB-induced augmentation of AP firing frequency (Control : 0.74 ± 0.31 Hz, NMB: 0.65 ± 0.28 Hz, 99 ± 19% of control, n = 10, P = 0.75, Wilcoxon test; F _(1,41) = 9.523, P = 0.0036 vs. NMB alone, two-way ordinary ANOVA, Figure 2, a), indicating that the function of G proteins is required for NMB-mediated enhancement of AP firing.

Figure 2. G proteins, PLCβ and PKC are required for NMB-mediated potentiation of AP firing frequency.

(a), intracellular dialysis of the G protein inactivator, GDP-β-S (0.5 mM), blocked NMB-induced facilitation of AP firing frequency (n = 10). (b), pretreatment of slices with the PLC inhibitor, U73122 (5 μM), blocked NMB-induced potentiation of AP firing frequency (n = 17), whereas application of NMB still significantly enhanced AP firing frequency in slices pretreated with the inactive analog, U73343 (5 μM, n = 18). (c), intracellular dialysis of the IP₃ receptor blocker, heparin (2 mg/ml, n = 16), or the inhibitor of the smooth-endoplasmic reticular Ca²⁺-ATPase, thapsigargin (10 μM, n = 21) did not significantly affect NMB-induced augmentation of AP firing frequency. (d), intracellular application of ryanodine (100 μM) to inhibit ryanodine receptors did not significantly influence NMB-elicited augmentation of AP firing frequency (n = 18), whereas intracellular perfusion of the Ca²⁺ chelator BAPTA (10 mM) blocked NMB-induced enhancement of AP firing frequency (n = 18). (e), replacement of extracellular Ca²⁺ with the same concentration of Mg²⁺ did not affect NMB-induced increases in AP firing frequency (n = 18). (f), pretreatment of slices with and continuous bath application of the selective PKC inhibitor chelerythrine (Chele, 10 μM) blocked NMB-induced augmentation of AP firing frequency (n = 22). (g), pretreatment of slices with and continuous bath application of the selective PKC inhibitor Bis II (1 μM) blocked NMB-induced augmentation of AP firing frequency (n = 10). (h), bath application of the PKC activator PMA (1 μM) significantly increased the AP firing frequency (n = 14), whereas pretreatment of slices with and continuous bath application of Bis II (1 μM, n = 15) or chelerythrine (10 μM, n = 10) blocked PMA-induced potentiation of AP firing frequency.

We assessed the roles of PLCβ in NMB-mediated excitation of CeL neurons. Slices were pretreated with the selective PLC inhibitor, U73122 (5 μM), for >2 h. Separate slices were treated with the inactive analogue U73343 (5 μM) in the same fashion as a control. Under these conditions, application of NMB did not significantly augment AP firing frequency in slices pretreated with U73122 (Control: 1.36 ± 0.25 Hz, NMB 1.28 ± 0.31 Hz, 99 ± 11% of control, n = 17 , P = 0.88, Wilcoxon test, Figure 2, b), but still significantly enhanced the AP firing frequency in slices pretreated with U73343 (Control: 0.51 ± 0.10 Hz, NMB 0.98 ± 0.19 Hz, 221 ± 35% of control , n = 18, P = 0.0007, Wilcoxon test; F _(1,33) = 9.25, P = 0.005 vs. U73122, two-way ordinary ANOVA, Figure 2, b). These results demonstrate that PLCβ is required for NMB-mediated increases in neuronal excitability in the CeL.

3.3. Intracellular Ca²⁺ release is unnecessary, but PKC is required for NMB-elicited excitation of CeL neurons

We investigated the involvement of Ca²⁺ release from intracellular stores in NMB-mediated facilitation of neuronal excitability. Intracellular application of the IP₃ receptor blocker heparin (2 mg/ml) through the recording pipettes failed to significantly affect NMB-mediated enhancement of AP firing frequency (Control: 0.56 ± 0.10 Hz, NMB: 1.79 ± 0.64 Hz, 299 ± 49% of control, n = 16, P = 0.0006, Wilcoxon test; F _(1,47) = 0.047, P = 0.83 vs. NMB alone, two-way ordinary ANOVA, Figure 2, c), indicating that IP₃ receptors are not required for NMB-induced augmentation of neuronal excitability. Similarly, intracellular application of the sarcoplasmic-endoplasmic reticulum Ca²⁺-ATPase inhibitor thapsigargin (10 μM) via the recording pipettes did not significantly alter NMB-mediated augmentation of AP firing frequency (Control: 0.52 ± 0.11 Hz, NMB: 0.82 ± 0.15 Hz, 174 ± 20% of control, n = 21, P < 0.0001, Wilcoxon test; F _(1,52) = 0.289, P = 0.59 vs. NMB alone, two-way ordinary ANOVA, Figure 2, c). We further examined the potential involvement of Ca²⁺ released from the ryanodine-sensitive store by intracellular perfusion of ryanodine (100 μM) to inhibit ryanodine receptors. Intracellular application of ryanodine did not significantly affect NMB-mediated increases in neuronal excitability in CeL neurons (Control: 1.97 ± 0.59 Hz, NMB: 3.98 ± 1.14 Hz, 244 ± 64% of control, n = 18, P = 0.002, Wilcoxon test; F _(1,50) = 0.09, P = 0.77 vs. NMB alone, two-way ordinary ANOVA, Figure 2, d). These results demonstrate that intracellular Ca²⁺ release is not involved in the NMB-elicited increases in AP firing. Interestingly, intracellular dialysis of the Ca²⁺ chelator, BAPTA (10 mM), significantly attenuated NMB-mediated increases in AP firing frequency (Control: 0.90 ± 0.19 Hz, NMB: 0.79 ± 0.14 Hz, 110 ± 14% of control, n = 18, P = 0.48, Wilcoxon test; F _(1,49) = 15.69, P = 0.0002 vs. NMB alone, two-way ordinary ANOVA, Figure 2, d). As PKC was involved in NMB-elicited excitation of CeL neurons (see below) and some PKC isoforms are Ca²⁺-dependent, one explanation for the BAPTA result is that BAPTA lowered the basal intracellular Ca²⁺ level, which may be required for the functions of Ca²⁺-dependent signaling molecules such as PKC. Lastly, we investigated the role of extracellular Ca²⁺ in NMB-elicited augmentation of neuronal excitability by replacing the extracellular Ca²⁺ with an equal concentration of Mg²⁺. Under these conditions, application of NMB similarly enhanced AP firing frequency (Control: 0.58 ± 0.10 Hz, NMB: 1.05 ± 0.16 Hz, 213 ± 28% of control, n = 18, P = 0.003, Wilcoxon test; F _(1,49) = 0.196, P = 0.66 vs. NMB alone, two-way ordinary ANOVA, Figure 2, e), indicating that the NMB-mediated increase in neuronal excitability is not dependent on extracellular Ca²⁺.

We further probed the role of PKC in NMB-mediated excitation of CeL neurons. Slices were pretreated with the selective PKC inhibitor, chelerythrine (10 μM), for >2 h, and the bath was continuously perfused with the same concentration of chelerythrine to ensure persistent inhibition of PKC. In this situation, application of NMB did not significantly increase the AP firing frequency (Control: 0.75 ± 0.18 Hz, NMB: 0.71 ± 0.17 Hz, 102 ± 15% of control, n = 22, P = 0.71, Wilcoxon test; F _(1,53) = 16.57, P = 0.0002 vs. NMB alone, two-way ordinary ANOVA, Figure 2, f). Moreover, pretreatment of slices with and continuous bath application of another selective PKC inhibitor, bisindolylmaleimide II (Bis II, 1 μM), blocked NMB-mediated increases in AP firing frequency (Control: 1.04 ± 0.32 Hz, NMB: 0.91 ± 0.26 Hz, 101 ± 15% of control, n = 10, P = 0.77, Wilcoxon test; F _(1,41) = 9.52, P = 0.0036 vs. NMB alone, two-way ordinary ANOVA, Figure 2, g). The involvement of PKC was further supported by bath application of the PKC activator, phorbol 12-myristate 13-acetate (PMA, 1 μM), resulting in a significant increase of AP firing frequency in CeL neurons (Control: 0.68 ± 0.21 Hz, PMA: 1.74 ± 0.43 Hz, 281 ± 70% of control, n = 14, P = 0.003, Wilcoxon test, Figure 2, h). In slices pretreated and continuously superfused with Bis II (1 μM), application of PMA failed to significantly increase AP firing frequency (Control: 0.42 ± 0.11 Hz, PMA: 0.33 ± 0.10 Hz, 75 ± 9% of control, n = 15, P = 0.06, Wilcoxon test; F _(1,27) = 19.11, P = 0.0002 vs PMA alone, Figure 2, h). Moreover, application of chelerythrine (10 μM) in the same fashion blocked PMA-induced increases in AP firing frequency (Control: 0.60 ± 0.31 Hz, PMA: 0.65 ± 0.29 Hz, 101 ± 15% of control, n = 10, P = 0.63, Wilcoxon test; F _(1,22) = 9.81, P = 0.005 vs. PMA alone, two-way ordinary ANOVA, Figure 2, h). These data together indicate that PKC is required for NMB-mediated excitation of CeL neurons.

3.4. NMB depolarizes CeL neurons and increases the input resistance and membrane time constants of CeL neurons

We then included TTX (0.5 μM) in the extracellular solution to block Na⁺-dependent AP firing to determine the effects of NMB on passive membrane properties of CeL neurons. Bath application of NMB induced significant depolarization of CeL neurons (Control: −63.1 ± 1.8 mV, NMB: −59.6 ± 1.9 mV, net depolarization: 3.5 ± 0.8 mV, n = 16, P = 0.0004, paired t-test, Figure 3, a1-a2). To confirm that NMB-elicited excitation of CeL neurons was indeed mediated by activation of BB1 receptors, we pretreated slices with and continuously bath-applied the selective BB1 antagonist, BIM23042 (0.3 μM) (Blais, Sethi, & Tabarean, 2016; Flynn, 1997). In the presence of BIM23042, NMB failed to depolarize CeL neurons significantly (BIM23042: −63.5 ± 1.2 mV, BIM23042 + NMB: −63.3 ± 1.3 mV, net depolarization: 0.20 ± 0.16 mV, n = 10, P = 0.24, paired t-test, Figure 3, a3). Furthermore, application of PD168368 (10 μM), another selective BB1 receptor antagonist (Moody, Jensen, Garcia, & Leyton, 2000), in the same fashion blocked NMB-induced depolarization (PD168368: −64.3 ± 1.9 mV, PD168368 + NMB: −64.1 ± 2.0 mV, net depolarization: 0.25 ± 0.14 mV, n = 10, P = 0.12, Wilcoxon test, Figure 3, a3). In voltage-clamp mode, NMB elicited an inward current at −60 mV (−19.6 ± 5.2 pA, n = 17, P = 0.002, paired t-test, Figure 3, b1-b2). Taken together, these results demonstrate that activation of BB1 receptors increases neuronal excitability through membrane depolarization.

Figure 3. NMB induces membrane depolarization and increases the input resistances and membrane time constants.

(a1-a3), bath application of NMB depolarized CeL neurons via activation of BB1 receptors. (a1), RMP recorded from a CeL neuron before, during, and after the application of NMB. (a2), summary data for NMB-induced depolarization (n = 16). Green circles represent the values from individual cells and the red circles represent their average. (a3), pretreatment of slices with and continuous bath application of the selective BB1 receptor antagonist BIM23042 (BIM, 0.3 μM, n = 10) or PD168368 (PD, 10 μM, n = 10) blocked NMB-induced depolarization. (b1-b2), bath application of NMB induced an inward current from CeL neurons in voltage-clamp. (b1), holding current recorded at −60 mV from a CeL neuron before, during, and after the application of NMB. (b2), summary of net holding currents induced by NMB (n = 17). Green circles represent the values from individual cells and the bar graph represent their average. (c1-c3), NMB increased the input resistance (R_in). (c1), voltage responses evoked by injection of negative currents from −100 to −20 pA at an interval of 20 pA before (left) and during (right) the application of NMB. (c2), I-V relationship averaged from 14 cells. R_in was obtained by linear fitting of the I-V relationship. (c3), summary graph for the R_in before and after the application of NMB (n = 14). (d1-d3), NMB increased membrane time constants. (d1), voltage response evoked by −100 pA current injection before and after the application of NMB. (d2), expansion of the voltage transient shown in the box in (d1) to demonstrate NMB-induced enlargement of membrane time constant. (d3), summary graph for membrane time constants before and after the application of NMB. (e1-e4), activation of BB1 receptors excited CeL neurons through an inhibition of inwardly rectifying K⁺ channels. (e1), the voltage-step protocol used. Cells were held at −60 mV and stepped from −140 mV to −40 mV for 400 ms in 10 mV voltage intervals every 10 s. (e2), representative currents elicited by the voltage-step protocol before and after bath application of NMB and the net currents acquired by subtraction. (e3), I-V curves averaged from 13 cells before and after the application of NMB. **** P < 0.0001, two-way repeated measures ANOVA followed by Šídák’s multiple comparison test. (e4), I-V curve of the net currents acquired by subtracting the currents in control condition from those after the application of NMB (n = 13).

We further determined the effect of BB1 receptor activation on the input resistance of CeL neurons by injecting a series of negative currents from −20 to −100 pA with a 20 pA step every 6 s before and after the application of NMB. We then fit the current-voltage relationship to a linear function for each cell to obtain the input resistance (R_in) which was the slope of the linear fitting (Figure 3, c1-c3). Bath application of NMB significantly increased R_in (Control: 260 ± 35 MΩ, NMB: 290 ± 36 MΩ, n = 14, P = 0.007, Wilcoxon test, Figure 3, c3), indicating that NMB increased R_in. We acquired the membrane time constants by fitting a single exponential function to the voltage transient generated by negative current injection (−100 pA, 100 ms from the end of the baseline). Bath application of NMB significantly increased the membrane time constants (Control: 38.3 ± 5.7 ms, NMB: 51.8 ± 7.8 ms, n = 14, P = 0.023, Wilcoxon test, Figure 3, d1-d3). These results together suggest that NMB excites CeL neurons by suppressing a membrane conductance.

3.5. Activation of BB1 receptors inhibits an inwardly rectifying K⁺ channel in CeL neurons

We next determined the ionic mechanisms by which BB1 receptor activation depolarizes CeL neurons. As shown in Figure 2, e, the effect of NMB on CeL neurons was independent of extracellular Ca²⁺. We therefore replaced extracellular Ca²⁺ with the same concentration of Mg²⁺ to avoid contamination of voltage-gated Ca²⁺ channels. TTX (0.5 μM) was included in the Ca²⁺-free extracellular solution to block voltage-gated Na⁺ channels. Cells were held at −60 mV and stepped from −140 mV to −40 mV for 400 ms at a voltage interval of 10 mV every 10 s (Figure 3, e1) before and after bath application of NMB when the maximal effect was observed. Steady-state currents were measured within 5 ms before the end of the step voltage protocols. Under these circumstances, the currents recorded before and after the application of NMB showed inward rectification (n = 13, Figure 3, e2-e4), suggesting that Kir channels are expressed in CeL neurons. Subtraction of the current-voltage (I-V) relationship in the control condition from that after the application of NMB displayed an inwardly rectified I-V curve (n = 13, Figure 3, e4). These results support that activation of the BB1 receptors excites CeL neurons by inhibiting Kir channels.

Micromolar concentrations (100 – 300 μM) of extracellular Ba²⁺ have been shown to block Kir channels by at least 80% (Binqi Hu et al., 2020; B. Hu, Cilz, & Lei, 2017; H. Li, Hu, Zhang, Boyle, & Lei, 2019). Inclusion of 300 μM Ba²⁺ in the extracellular solution inhibited a current showing inward rectification (n = 12, F _(1,11) = 39.65, P < 0.0001, two-way repeated measures ANOVA followed by Šídák’s multiple comparison test, Figure 4, a1-a3) suggesting that this concentration of Ba²⁺ inhibited Kir channels. In the presence of Ba²⁺, application of NMB failed to elicit more currents significantly (n = 12, F _(1,11) = 0.28, P = 0.61, two-way repeated measures ANOVA followed by Šídák’s multiple comparison test, Figure 4, a4-a6), demonstrating that NMB-induced excitation of CeL neurons is mediated by depression of Ba²⁺-sensitive Kir channels.

Figure 4. NMB-mediated inhibition of Kir currents is Ba²⁺-sensitive but insensitive to the Kir2 subfamily blocker ML 133.

(a1-a6), bath application of Ba²⁺ (300 μM) alone inhibited Kir channel currents and blocked NMB-elicited inhibition of Kir channel currents. (a1), currents evoked by the voltage-step protocol before (left) and during (middle) the application of Ba²⁺ (300 μM) and the net currents acquired by subtraction (right). (a2), I-V curves of the currents elicited by the voltage-step protocol before and during the application of Ba²⁺. **** P < 0.0001, ** P < 0.01. (a3), Net currents acquired by subtraction of the currents recorded in the control condition from those recorded from the same cell in the presence of Ba²⁺ (n = 12). Note that the Ba²⁺-sensitive currents showed inward rectification. (a4), currents recorded from a CeL neuron in response to the voltage-step protocol in the presence of Ba²⁺ alone (left) and Ba²⁺ plus NMB (middle). The net currents acquired by subtraction were shown in the right panel. (a5), I-V curves of the currents elicited by the voltage-step protocol in the presence of Ba²⁺ alone and Ba²⁺ plus NMB (n = 12). (a6), Net currents acquired by subtracting the currents in the presence of Ba²⁺ alone from those recorded from the same cell in the presence of Ba²⁺ plus NMB. NMB-elicited net currents in the control condition (green circles) were co-plotted as a comparison. (b1-b6), bath application of ML 133 (30 μM), a blocker of Kir2 subfamily channels, by itself inhibited Kir currents, but failed to block NMB-mediated inhibition of Kir currents. (b1), currents elicited by the voltage-step protocol before (left) and during (middle) the application of ML 133 (30 μM). The net currents inhibited by ML 133 were shown on the right. (b2), I-V curves of the currents elicited by the voltage-step protocol before and during the application of ML 133 (n = 12). (b3), net currents acquired by subtraction of the currents in the control condition from those recorded from the same cell in the presence of ML 133. Note that the ML133-sensitive currents displayed inward rectification. (b4), current traces recorded from a CeL neuron in response to the voltage-step protocol in the presence of ML 133 (left) and ML 133 plus NMB (middle) and the net current acquired via subtraction (right). (b5), I-V curves of the currents generated by the voltage-step protocol in the presence of ML 133 alone and ML 133 plus NMB (n = 12). (b6), net currents acquired by subtraction of currents recorded from CeL neurons in the presence of ML 133 from those in the presence of ML 133 plus NMB. Net currents elicited by NMB in control condition without application of ML133 were co-plotted for comparison. **** P < 0.0001, *** P < 0.001, ** P < 0.01, * P < 0.05, two-way repeated measures ANOVA followed by Šídák’s multiple comparison test.

3.6. GIRK channels are involved in BB1 receptor-mediated excitation of CeL neurons

We aimed at identifying the subtype of Kir channels involved in NMB-mediated facilitation of neuronal excitability. Kir channels are divided into four functional groups: 1) the constitutively active Kir2 subfamily channels including Kir2.1, Kir2.2, Kir2.3, and Kir2.4; 2) the G protein-gated GIRK (Kir3 subfamily) channels consisting of Kir3.1 (GIRK1), Kir3.2 (GIRK2), Kir.3.3 (GIRK3), and Kir3.4 (GIRK4); 3) the ATP-sensitive K⁺ (K_ATP) channels comprising Kir6.1 and Kir6.2; 4) K⁺ transport channels involving Kir1.1, Kir4.1, Kir.4.2, and Kir7.1 (Hibino et al., 2010). We tested the role of the Kir2 subfamily channels in NMB-elicited depression of Kir channels by utilizing the selective Kir2 channel blocker, ML 133 (30 μM) (Ford & Baccei, 2016; X. Huang, Lee, Lu, Sanders, & Koh, 2018; Kim et al., 2015; Sonkusare, Dalsgaard, Bonev, & Nelson, 2016; Wang et al., 2011). Bath application of ML 133 alone induced a significant inhibition of Kir currents from −120 to −140 mV (n = 12, F _(1,11) = 7.29, P = 0.021, two-way repeated measures ANOVA followed by Šídák’s multiple comparison test, Figure 4, b1-b3). In the presence of ML 133, application of NMB still significantly depressed Kir currents at potentials −100 to −140 mV (n = 12, F _(1,11) = 12.40, P = 0.005, two-way repeated measures ANOVA followed by Šídák’s multiple comparison test, Figure 4, b4-b6). I-V responses elicited in the presence of ML 133 and NMB were significantly different from those obtained in CeL neurons treated with NMB alone (F _(1,253) = 4.44, P = 0.036, two-way ordinary ANOVA followed by Šídák’s multiple comparison test, Figure 4, b6), but did not differ significantly at any recorded voltages according to post-hoc analysis. Together, these results suggest that the Kir2 subfamily channels are not involved in the NMB-mediated excitation of CeL neurons.

We further probed the role of GIRK channels in NMB-mediated inhibition of Kir channels. GIRK1 channels are strongly expressed in the BLA and cortical nuclei with low expression in the CeA (DePaoli, Bell, & Stoffel, 1994). Bath application of ML 297 (10 μM), an activator of GIRK1-containing channels (Kaufmann et al., 2013) failed to significantly affect Kir currents (n = 9, F _(1,8) = 0.67, P = 0.44, two-way repeated measures ANOVA followed by Šídák’s multiple comparison test, Figure 5, a1-a3), suggesting that there were no functional GIRK1 channels in the CeL neurons and NMB-elicited excitation of CeL neurons was unlikely to be mediated by depressing GIRK1 channels.

Figure 5. GIRK channels are involved in NMB-elicited depression of Kir channel currents.

(a1-a3), application of ML 297, the GIRK1-containing channel activator, failed to alter Kir currents. (a1), currents recorded from a CeL neuron in response to the voltage-step protocol before (left) and during (middle) the application of ML 297 (10 μM) and the net current acquired by subtraction (right). (a2), I-V curves of the currents elicited by the voltage-step protocol before and during the application of ML 297 (n = 9). (a3), net currents acquired by subtracting the currents in the control condition from those recorded from the same cells in the presence of ML 297. Application of ML 297 did not significantly affect Kir currents at each voltage. (b1-b6), bath application of tertiapin-Q (TQ) by itself inhibited Kir currents and significantly depressed NMB-mediated inhibition of Kir currents. (b1), currents recorded from a CeL neuron in response to the voltage-step protocol before (left) and during (middle) the bath application of TQ (250 nM) and the net current generated by subtraction (right). (b2), I-V curves of the currents elicited by the voltage-step protocol before and during the application of TQ (n = 10). (b3), net currents generated by subtracting the currents in the control condition from those recorded from the same neurons in the presence of TQ. TQ-sensitive currents displayed inward rectification. (b4), currents recorded from a CeL neuron in response to the voltage-step protocol in the presence of TQ alone (left) and TQ plus NMB (middle) and the net current acquired by subtraction (right). (b5), I-V curves of the currents elicited by the voltage-step protocol in the presence of TQ or TQ + NMB (n = 10). (b6), net currents acquired by subtracting the currents in the presence of TQ alone from those recorded from the same cells in the presence of TQ plus NMB. NMB-elicited net currents in the control condition (green circles) were co-plotted as a comparison. Note that TQ significantly reduced NMB-elicited depression of Kir currents at −130 mV and −140 mV. ** P < 0.01, ordinary two-way ANOVA. (c1-c6), bath application of tertiapin-LQ (T-LQ) alone did not alter Kir currents and failed to significantly change NMB-mediated inhibition of Kir currents. (c1), currents recorded from a CeL neuron in response to the voltage-step protocol before (left) and during (middle) the application of T-LQ (100 nM) and the net current acquired by subtraction (right). (c2), I-V curves of the currents elicited by the voltage-step protocol before and during the application of T-LQ (n = 7). (c3), net currents acquired by subtracting the currents in the control condition from those recorded from the same cells in the presence of T-LQ. T-LQ failed to alter Kir currents at each voltage. (c4), currents recorded from a CeL neuron in response to the voltage-step protocol in the presence of T-LQ alone (left) and T-LQ plus NMB (middle) and the net current acquired by subtraction (right). (c5), I-V curves of the currents elicited by the voltage-step protocol in the presence of T-LQ and T-LQ + NMB (n = 7). (c6), net currents acquired by subtracting the currents in the presence of T-LQ alone from those recorded from the same cells in the presence of T-LQ plus NMB. NMB-sensitive net currents in control conditions were co-plotted for comparison. Application of T-LQ failed to significantly alter the NMB-induced depression of Kir currents. **** P < 0.0001, *** P < 0.001, ** P < 0.01, * P < 0.05, two-way repeated measures ANOVA followed by Šídák’s multiple comparison test.

We then used tertiapin-Q (TQ), a blocker for GIRK (K_i = 13.3 nM) and Kir1.1 (K_i = 1.3 nM) channels (Felix et al., 2006; Jin, Klem, Lewis, & Lu, 1999; Jin & Lu, 1999). Bath application of TQ (250 nM) by itself depressed a current displaying inward rectification (n = 10, F_(1,9) = 23.16, P = 0.001, two-way repeated measures ANOVA followed by Šídák’s multiple comparison test, Figure 5, b1-b3). There were significant differences in the I-V curve before and during the application of TQ at a voltage range from −100 to −140 mV (n = 10, Figure 5, b1-b3), suggesting that CeL neurons express functional GIRK channels. Whereas application of NMB in the presence of TQ still significantly depressed Kir currents (F_(1,9) = 24.95, P = 0.0007, two-way repeated measures ANOVA followed by Šídák’s multiple comparison test, Figure 5, b4-b6), the NMB-elicited depression of Kir currents in the presence of TQ was significantly smaller at −130 mV and −140 mV compared with the effect of NMB in the control condition (Felix et al., 2006; Jin et al., 1999; Jin & Lu, 1999) (Figure 5, b6). These results together indicate that NMB-induced excitation of CeL neurons occurred at least partially through the depression of GIRK channels.

Because TQ inhibits both GIRK and Kir1.1 channels (Felix et al., 2006; Jin et al., 1999; Jin & Lu, 1999), we probed the potential involvement of Kir1.1 channels in NMB-induced inhibition of Kir currents. Bath application of the selective Kir1.1 channel blocker, tertiapin-LQ (100 nM), which lacks effects on GIRK channels (Ramu, Xu, & Lu, 2008), failed to alter significantly the I-V relationship (n = 7, F_(1,6) = 1.58, P = 0.26, two-way repeated measures ANOVA, Figure 5, c1-c3), suggesting that CeL neurons do not express functional Kir1.1 channels. Application of NMB in the continuous presence of tertiapin-LQ still inhibited a comparable extent of Kir currents (n = 7, F _(1,6) = 31.59, P = 0.001, two-way repeated measures ANOVA followed by Šídák’s multiple comparison test, Figure 5, c4-c6). The NMB-induced depression of Kir currents in the presence of tertiapin-LQ was not significantly different from that in the control condition at all voltages tested (F _(1,18) = 2.44, P = 0.14, two-way ordinary ANOVA, Figure 5, c6). These data suggest a lack of involvement of Kir1.1 channels in NMB-elicited depression of Kir currents. Together, these data suggest that activation of the BB1 receptor excites CeL neurons through the inhibition of GIRK channels.

3.7. G proteins, PLCβ and PKC are required for BB1 receptor-mediated depression of Kir channels

As our results indicated that activation of BB1 receptors excited CeL neurons by depressing the GIRK type of the Kir channels, we further tested the roles of G proteins, PLCβ and PKC in NMB-mediated inhibition of Kir channel currents. Inclusion of GDP-β-S (0.5 mM) in the intracellular recording solution blocked NMB-induced depression of Kir currents (n = 14, F _(1,13) = 1.54, P = 0.24, two-way repeated measures ANOVA followed by Šídák’s multiple comparison test, Figure 6, a1-a3). Additionally, pretreatment of slices with the PLCβ inhibitor U73122 (5 μM) significantly reduced NMB-induced depression of Kir currents (n = 15, F_(1,14) = 0.14, P = 0.71, two-way repeated measures ANOVA, Figure 6, b1-b3), compared with the effect of NMB in slices pretreated with the inactive analog, U73343 (5 μM, n = 5, F _(1,4) = 22.22, P = 0.009, two-way repeated measures ANOVA followed by Šídák’s multiple comparison test, Figure 6, c1-c3). These results demonstrate that the activity of PLCβ is required for NMB-mediated depression of Kir channels in the CeL.

Figure 6. G proteins and PLCβ are involved in BB1 receptor-mediated depression of Kir channels.

(a1-a3), intracellular dialysis of GDP-β-S (0.5 mM) via the recording pipettes blocked NMB-elicited inhibition of Kir currents. (a1), currents recorded from a CeL neuron in response to the voltage-step protocol before (left) and during (middle) the application of NMB in the presence of GDP-β-S and the net current generated by subtraction (right). (a2), I-V curves of the currents elicited by the voltage-step protocol before and during the application of NMB (n = 14). (a3), net currents generated by subtracting the currents in control conditions from those recorded from the same neurons during the bath application of NMB with GDP-β-S in the recording pipettes. Note that inclusion of GDP-β-S in the recording pipettes blocked NMB-elicited depression of Kir currents. (b1-b3), pretreatment of slices with the selective PLC inhibitor, U73122 (5 μM), significantly depressed NMB-elicited Kir currents. (b1), currents recorded from a CeL neuron in response to the voltage-step protocol before (left) and during (middle) the application of NMB in a slice pretreated with U73122 and the net currents acquired by subtraction (right). (b2), I-V curves of the currents generated by the voltage-step protocol before and during the bath application of NMB (n = 15). (b3), net currents acquired by subtracting the currents in control conditions from those during the bath application of NMB from the same neurons in slices pretreated with U73122. Note that pretreatment of slices with U73122 significantly attenuated NMB-mediated depression of Kir currents. (c1-c3), bath application of NMB in slices pretreated with the inactive analog U73343 (5 μM), still significantly depressed Kir currents. (c1), currents recorded from a CeL neuron in response to the voltage-step protocol before (left) and during (middle) the application of NMB in a slice pretreated with U73343 and the net current acquired by subtraction (right). (c2), I-V curves of the currents recorded from the same neurons before and during the application of NMB (n = 5) in slices pretreated with U73343. (c3), net currents acquired by subtracting the currents in control conditions from those during the application of NMB from the same neurons in slices pretreated with U73343. Note that bath application of NMB still significantly depressed Kir currents. **** P < 0.0001, *** P < 0.001, ** P < 0.01, * P < 0.05, two-way repeated measures ANOVA followed by Šídák’s multiple comparison test.

We then explored the role of PKC in NMB-induced depression of Kir currents. Slices were pretreated with Bis II (1 μM), a selective PKC inhibitor, and the same concentration of Bis II was continuously bath-applied in the extracellular solution. Bath application of NMB in this condition failed to significantly alter Kir channel currents at all voltages recorded (n = 12, F _(1,11) = 0.03, P = 0.86 two-way repeated measures ANOVA, Figure 7, a1-a3). Compared with NMB-mediated depression in the control condition, slices treated with Bis II elicited significantly smaller currents (F _(1,253) = 94.63, P < 0.0001 vs. NMB-alone, ordinary two-way ANOVA, Figure 7, a3), supporting a functional requirement of PKC in NMB-mediated depression of Kir channels. Similarly, bath application of a PKC activator, PMA (1 μM), significantly inhibited Kir currents (n = 13, F _(1,12) = 12.97, P = 0.004, two-way repeated measures ANOVA followed by Šídák’s multiple comparison test, Figure 7, b1-b3). The PMA-mediated depression was significantly reduced in slices pretreated with Bis II (1 μM, n = 12, F _(1,11) = 1.80, P = 0.21, two-way repeated measures ANOVA, Figure 7, c1-c3). These results further support the involvement of PKC in NMB-mediated depression of Kir channels. Lastly, we determined the effects of intracellular Ca²⁺ levels on NMB-elicited inhibition of Kir channels. Intracellular perfusion of BAPTA (10 mM) via the recording pipettes significantly attenuated NMB-elicited depression of Kir currents (F _(1,286) = 25.88, P < 0.0001 vs. NMB alone, ordinary two-way ANOVA, Figure 7, d1-d3), supporting the involvement of the Ca²⁺-dependent PKC isoform in NMB-mediated inhibition of Kir channels.

Figure 7. PKC is involved in NMB-induced depression of Kir channels.

(a1-a3), pretreatment of slices with and continuous bath application of the selective PKC inhibitor, Bis II (1 μM), blocked NMB-mediated inhibition of Kir currents at each voltage. (a1), currents recorded from a CeL neuron elicited by the voltage-step protocol before (left) and during (middle) the application of NMB in a slice treated with Bis II and the net currents acquired by subtraction (right). (a2), I-V curves of the currents generated by the voltage-step protocol before and during the application of NMB (n = 12) recorded from the same neurons in slices treated with Bis II. Note that application of Bis II blocked NMB-induced depression of Kir currents at each voltage. (a3), net currents acquired by subtracting the currents in control conditions from those during the application of NMB recorded from the same neurons in slices treated with Bis II. NMB-induced net currents in control conditions without Bis II were co-plotted for comparison (green circles). (b1-b3), application of the PKC activator, PMA (1 μM), inhibited Kir currents. (b1), current traces recorded from a CeL neuron elicited by the voltage-step protocol before (left) and during (middle) the application of PMA and the net current acquired by subtraction (right). (b2), I-V curves generated by the voltage-step protocol before and during the application of PMA (n = 13). Note that PMA significantly attenuated Kir currents. (b3), net currents produced by subtracting the currents in control conditions from those recorded from the same cells during the application of PMA. (c1-c3), pretreatment of slices with and continuous bath application of Bis II blocked PMA-mediated depression of Kir currents. (c1), currents recorded from a CeL neuron in response to the voltage-step protocol before (left) and during (middle) the bath application of PMA in a CeL neuron treated with Bis II and the net current acquired by subtraction (right). (c2), I-V curves of the currents generated by the voltage-step protocol before and during the bath application of PMA (n = 12) in slices treated with Bis II. (c3), net currents acquired by subtracting the currents in control conditions from those during the application of PMA recorded from the same neurons in slices treated with Bis II. (d1-d3), intracellular application of the Ca²⁺ chelator BAPTA (10 mM) via the recording pipettes significantly reduced NMB-mediated depression of Kir currents. (d1), currents recorded from a CeL neuron in response to the voltage-step protocol before (left) and during (middle) the application of NMB with the intracellular solution containing BAPTA and the net current acquired by subtraction (right). (d2), I-V curves of the currents generated by the voltage-step protocol before and during the application of NMB (n = 15) with the intracellular solution containing BAPTA. (d3), net currents acquired by subtracting the currents in control conditions from those during the bath application of NMB from the same neurons with the intracellular solution containing BAPTA. NMB-elicited net currents in control conditions without BAPTA were co-plotted for comparison (green circles). **** P < 0.0001, *** P < 0.001, ** P < 0.01, * P < 0.05, two-way repeated measures ANOVA followed by Šídák’s multiple comparison test.

3.8. BB1 receptor-mediated excitation of CeL neurons promotes inhibitory cardiovascular responses

The distinct CeA projections to the brainstem involved in the alterations of cardiac output in response to emotionally relevant stimuli are not completely understood (Galeno & Brody, 1983; Iwata et al., 1987; Mogenson & Calaresu, 1973; Viviani et al., 2011). As bombesin receptors are implicated in cardiovascular responses, we tested the roles of BB1 receptor-mediated excitation of CeA neurons in modulation of cardiac function. Guide cannulae were bilaterally implanted into the CeA and NMB and/or other compounds were microinjected into the CeA to probe the effects of BB1 receptor activation on mean arterial pressures (MAP; mmHg) and heart rate (HR; bpm) (Figure 8, a & c). There were no significant differences in baseline cardiovascular responses between groups (F _{(2.4, 24.3)} = 2.76, P = 0.07, mixed-effect two-way ANOVA). A two-way ANOVA revealed saline injection did not significantly affect any cardiovascular parameters (Treatment: F _(1,10) = 0.007, P = 0.93, two-way ANOVA). Microinjection of NMB produced a significant dose-dependent reduction in MAP (F _(3,39) = 6.21, P = 0.0015, one-way ANOVA; saline vs. 0.3 nmol NMB: P = 0.035; saline vs. 1 nmol NMB: P = 0.0005 via Dunnett’s multiple comparisons test, Figure 9, a1, Table 1) and HR (F _(3,39) = 7.55, P = 0.0004, one-way ANOVA; saline vs. 0.1 nmol: P = 0.006; saline vs. 0.3 nmol NMB: P = 0.0001 Dunnett’s multiple comparisons test, Figure 9, a2, Table 1). The effects of NMB were not sex-dependent, as comparison of MAP (F _(1,35) = 0.198, P = 0.66) or HR (F _(1,35) = 0.961, P = 0.33) from male and female rats showed no significant differences (Two-way ordinary ANOVA). The IC₅₀ value was calculated to be 0.362 nmol obtained by fitting the MAP data at different doses of NMB. We therefore used 0.3 nmol NMB for the remaining in vivo experiments.

Figure 9. BB1 receptor activation in the CeA reduces cardiovascular output and affects ASR and FPS.

(a1-a4) Microinjection of NMB into the CeA lowered MAP and reduced HR through inhibition of GIRK channels. (a1), NMB dose-dependently reduced MAP, and (a2) HR. * P < 0.05, ** P <0.01, *** P < 0.001 vs saline (one-way ANOVA followed by Dunnett’s test). The circles and triangles represent data obtained from male and female rats, respectively. (a3), microinjection of NMB (0.3 nmol) into the CeA reduced MAP, whereas prior administration of the BB1 receptor antagonist, BIM 23042 (0.3 nmol) or GIRK channel blocker, TQ (250 pmol) blocked NMB-mediated depression of MAP. (a4), microinjection of NMB (0.3 nmol) into the CeA produced bradycardia, whereas prior microinjection of the BB1 receptor antagonist, BIM 23042 (0.3 nmol) or GIRK channel blocker, TQ (250 pmol) blocked NMB-induced bradycardia. *** P < 0.01, **** P < 0.0001 (one-way ANOVA followed by Dunnett’s test), n. s, no significance (Tukey's test). (b1-b3), microinjection of NMB increased ASR while reducing FPS. (b1), microinjection of NMB (0.3 nmol) into the CeA enhanced ASR to the 105 dB WNB without altering ASR to the 90 dB or 95 dB WNBs. ** P < 0.01 (Two-way ANOVA followed by Šídák’s multiple comparisons test). (b2), application of NMB into the CeA prior to fear conditioning significantly reduced cued responses without effects on non-cued responses during the FPS testing session. ** P < 0.01 (Two-way ANOVA followed by Šídák’s multiple comparisons test). (b3), behavioral responses to foot shock on Day 3 did not differ significantly in response to NMB microinjection. The circles and triangles represent data obtained from male and female rats, respectively.

Table 1:

Cardiovascular parameters

	Baseline Mean (± SEM)	NMB Mean (± SEM)	P-value
Systolic (mmHg)	160.2 (4.8)	141.9 (3.7)	0.03
Diastolic (mmHg)	123.6 (5.0)	101.1 (4.4)	0.007
MAP (mmHg)	136.5 (5.2)	110.1 (4.2)	0.001
HR (bpm)	433.3 (13.8)	371.7 (8.4)	< 0.001

We then tested the involvement of BB1 receptors in NMB-mediated depression of MAP and HR by microinjection of the selective BB1 receptor antagonist BIM23042 (0.3 nmol)(Flynn, 1997). Application BIM23042 by itself had no significant effect on MAP (P > 0.99 vs. saline, Figure 9, a3) or HR (P = 0.85 vs. saline, Figure 9, a4). However, prior microinjection of BIM23042 prevented NMB-induced depression of MAP (P = 0.99, one-way ANOVA followed by Tukey's multiple comparisons test, Figure 9, a3) and HR (P = 0.99, one-way ANOVA followed by Tukey's multiple comparisons test, Figure 9, a4), corroborating the involvement of BB1 receptors.

Our in vitro results indicate that activation of BB1 receptors excites CeL neurons by depressing GIRK channels. We further probed the roles of GIRK channels in NMB-mediated suppression of MAP and HR. TQ is a selective GIRK channel inhibitor that effectively inhibits GIRK channels in picomolar range (Mazarati et al., 2006; Morgan, Tran, Wescom, & Bobeck, 2020). Bilateral microinjection of TQ (250 pmol) by itself had no significant effects on MAP (P = 0.97 vs. saline, Figure 9, a3) and HR (P = 0.77 vs. saline, Figure 9, a4), but blocked NMB-elicited depression of MAP (P = 0.96, one-way ANOVA followed by Tukey's multiple comparisons test, Figure 9, a3) and HR (P = 0.80, one-way ANOVA followed by Tukey's multiple comparisons test, Figure 9, a4), indicating that GIRK channels are required for NMB-mediated depression of MAP and HR.

3.9. BB1 receptor activation augments ASR but attenuates FPS

The microcircuits of the CeA are involved in fear processing (Ciocchi et al., 2010; Haubensak et al., 2010; Viviani et al., 2011) and send projections to brainstem nuclei involved in fear expression and cardiovascular responses (Viviani et al., 2011). Whereas NMB and its receptor have been implicated in stress responses (Nieves Gonzalez et al., 2008; Merali, Graitson, Mackay, & Kent, 2013; Merali, Kent, & Anisman, 2002; Roesler, Kent, Schröder, Schwartsmann, & Merali, 2012), the effect of NMB on fear expression remain less understood (Bédard, Mountney, Kent, Anisman, & Merali, 2007; Zul Merali, Tania Bédard, et al., 2006b; Zul Merali, Pamela Kent, et al., 2006; Ohki-Hamazaki et al., 1999). FPS is a translatable paradigm used to measure the acquisition and expression of conditioned fear responses in both rodents and humans (M. Davis, 1989; Michael Davis, 2001; M. Davis, Falls, Campeau, & Kim, 1993). Taking advantage of a simple reflex (ASR), FPS measures conditioned fear by an increase in the amplitude of ASR in the presence of a previously paired cue with a foot shock. We probed the effects of BB1 receptor activation on both ASR and FPS by microinjecting NMB or saline into the CeA (Figure 8, b & c).

Microinjection of NMB (0.3 nmol) into the CeA significantly augmented ASR to the 105 dB WNB compared to the rats injected with saline (0.9 % NaCl) (NMB: 1013.77 ± 214.88; Saline: 599.19 ± 92.10 arbitrary units, n = 11, P = 0.0204, Šídák’s multiple comparisons test, Figure 9, b1). These results were independent of the sex of the rat (F _(1,54) = 3.39, P = 0.07). Furthermore, microinjection of NMB into the CeA (0.3 nmol) significantly reduced FPS responses to cued stimuli (P = 0.004, Šídák’s multiple comparisons test, Figure 9, b2), with no significant effects on baseline (Preshock) (P = 0.97, Šídák’s multiple comparisons test, Figure 9, b2) or non-cued responses (P = 0.91, Šídák’s multiple comparisons test, Figure 9, b2). Microinjection of NMB into the CeA prior to fear conditioning had no significant effect on shock reactivity (P = 0.57, Mann-Whitney test, Figure 9, b3). Moreover, the shock reactivity amplitudes measured between male and female rats injected with saline (P = 0.22) or NMB (P = 0.98) were not significantly different, suggesting that the alterations in FPS were independent of aversive stimuli or sex of the animal.

4. Discussion

Whilst injection of bombesin into the CeA increases mean arterial pressure (Brown & Gray, 1988) and decreases food intake (Kyrkouli et al., 1987; Vigh et al., 1999), the cellular and molecular mechanisms whereby bombesin and bombesin-like peptides modulate these physiological functions have not been determined. As NMB peptide mRNA (Wada et al., 1990) and NMB binding sites (M. C. Lee et al., 1990) are expressed in the CeA and peripheral and central administration of bombesin increases Fos-like immunoreactivity, a marker of neuronal activation, in the CeL (B. H. Li & Rowland, 1996), we tested the hypothesis that activation of BB1 receptors facilitates neuronal excitability in the CeL. Our results demonstrate that BB1 receptor activation excited CeL neurons via inhibition of GIRK type Kir channels. BB1 receptor-mediated excitation of CeL neurons and depression of Kir channel currents required the functions of G proteins, PLCβ and PKC, but were independent of intracellular Ca²⁺ release. We further showed that BB1 receptor activation reduced MAP and HR, depending on GIRK channel activity. In ASR and FPS testing, microinjection of NMB into the CeA increased ASR, but reduced FPS to cued stimuli. Our results may provide a cellular and molecular mechanism by which bombesin and bombesin-like peptides regulate physiological functions in vivo.

Neuropeptides of the bombesin family increase neuronal excitability mainly by two ionic mechanisms: depression of Kir channels and activation of cation channels. For example, GRP excites spinal cord neurons (Pagani et al., 2019), paraventricular thalamic neurons (Hermes, Kolaj, Coderre, & Renaud, 2013) and interneurons in the entorhinal cortex (Zhang et al., 2014) by depressing Kir channels; GRP excites paraventricular thalamic neurons by activating TRPV1 channels (Hermes et al., 2013); both NMB and GRP excite neuropeptide Y-containing neurons in the arcuate nucleus by activation of nonselective cation channels and sodium/calcium exchangers (A. N. van den Pol et al., 2009) and hippocampal interneurons by activating a cation channel (K. Lee et al., 1999). Consistent with this scenario, our results support a role of GIRK type Kir channels in BB1 receptor-elicited excitation of CeL neurons. Inhibition of a membrane conductance should increase input resistance and membrane time constants. Consistent with this anticipation, application of NMB augmented the input resistance and the membrane time constants of the CeL neurons. Moreover, the I-V curve of the NMB-sensitive current exhibited inward rectification with a reversal potential near the K⁺ reversal potential. Micromolar concentrations of Ba²⁺ selectively block Kir channels. Our results showed that application of 300 μM Ba²⁺ inhibited Kir channels by itself and blocked NMB-elicited depression of Kir currents, further supporting a role of Kir channels in NMB-mediated excitation of CeL neurons. Kir channels are classified into 4 functional subfamilies: Kir2 (Kir2.1, Kir2.2, Kir2.3, Kir2.4), Kir3 (GIRK channels, Kir3.1, Kir3.2, Kir3.3, Kir3.4), Kir6 (K_ATP channels, Kir6.1, Kir6.2), and K⁺ transport channels (Kir1.1, Kir4.1, Kir7.1) (Hibino et al., 2010). Our intracellular solution contained 2 mM ATP which would exert inhibition on K_ATP channels. Application of the selective Kir2 subfamily blocker ML 133 at a saturating concentration (30 μM) (Wu et al., 2010) attenuated Kir currents, suggesting that the CeL neurons express functional Kir2 subfamily channels. Consistent with our electrophysiological results, the amygdala expresses mRNAs for both Kir2.1 and Kir2.2 (Karschin, Dissmann, Stuhmer, & Karschin, 1996). However, NMB-induced depression of Kir currents was not significantly altered by inclusion of ML 133 in the extracellular solution, suggesting that the Kir2 subfamily channels were not involved in NMB-elicited excitation of CeL neurons. TQ inhibits both GIRK and Kir1.1 channels (Felix et al., 2006; Jin et al., 1999; Jin & Lu, 1999), whereas tertiapin-LQ blocks Kir1.1 channels without effects on GIRK channels (Ramu et al., 2008). Application of TQ by itself elicited significant inhibition of Kir currents, whereas application of tertiapin-LQ failed to alter Kir currents significantly. These results suggest that the CeL neurons express tonically functional GIRK channels without expression of Kir1.1 channels, consistent with the notion that GIRK channels are constitutively active (Chen & Johnston, 2005; J. C. Gonzalez, Epps, Markwardt, Wadiche, & Overstreet-Wadiche, 2018; Lüscher, Jan, Stoffel, Malenka, & Nicoll, 1997). Furthermore, NMB-induced inhibition of Kir currents was significantly reduced in the presence of TQ, but application of tertiapin-LQ failed to alter NMB-induced inhibition of Kir currents significantly. These results together suggest that activation of BB1 receptors excite CeL neurons by suppressing GIRK channels.

Our result that application of TQ alone significantly inhibited Kir channel currents satisfies the prerequisite that GIRK channels must be tonically active in the CeL neurons. Another precondition to explain our results is that the CeL neurons should physically express GIRK channels. The GIRK channels include GIRK1, GIRK2, GIRK3 and GIRK4, and these subunits are widely expressed in the brain, existing predominantly as heterotetramers of GIRK1, GIRK2 and/or GIRK3, or as homotetramers of the GIRK2 subunit (Hibino et al., 2010; Luscher & Slesinger, 2010). GIRK4 is not expressed in the amygdala (Karschin et al., 1996; Murer et al., 1997) but restricted to some neuronal populations such as Purkinje cells and neurons of the globus pallidus and the ventral pallidum, suggesting that GIRK4 channels are unlikely to be the molecular target of BB1 receptor activation in the CeL. Furthermore, our results did not support the involvement of GIRK1 channels in BB1 receptor-mediated excitation of CeL neurons, because application of ML 297, an activator selective for GIRK1-containing channels, did not affect Kir currents significantly in the CeL neurons, suggesting that there is no functional GIRK1 in the CeL neurons. Consistent with our electrophysiological data, the expression of GIRK1 mRNA has been detected more prominently in the BLA and cortical nuclei with less expression in the CeA of the amygdala (DePaoli et al., 1994). Thus, the most possible subunit compositions involved in BB1 receptor-mediated excitation of CeL neurons are GIRK2 and GIRK3, which could exist as GIRK2/GIRK3 heterotetramers or GIRK2 homotetramers (Hibino et al., 2010; Luscher & Slesinger, 2010). In line with this speculation, GIRK2 (Karschin et al., 1996; Murer et al., 1997) and GIRK3 (Karschin et al., 1996) are expressed in the amygdala, and activation of both BB1 and BB2 receptors results in depression of GIRK channels including GIRK1, GIRK2 and GIRK4 expressed in xenopus oocytes (Stevens, Shah, Pinnock, & Lee, 1999). BB1 receptors are coupled to Gα_q/11 resulting in activation of PLCβ which hydrolyzes PIP₂ generating IP₃ to release intracellular Ca²⁺ and DAG to activate PKC (Nieves Gonzalez et al., 2008; N. Gonzalez et al., 2015; Shapira et al., 1994). Our results suggested that intracellular Ca²⁺ release was not required, whereas the activities of PLCβ and PKC were necessary, for NMB-elicited excitation of CeL neurons and inhibition of Kir channels. Consistent with our results, PLC is required for NMB and GRP-mediated excitation of hippocampal interneurons although the involved ion channels are cation channels, not Kir channels (K. Lee et al., 1999). In accordance with our results, activation of PKC phosphorylates and inhibits GIRK channels (Adney, Ha, Meng, Kawano, & Logothetis, 2015; Mao et al., 2004; Stevens et al., 1999) and PLC and PKC are involved in BB1 or BB2 receptor-mediated depression of GIRK channels co-expressed in xenopus oocytes (Stevens et al., 1999).

PKC is a serine-threonine kinase family of isozymes with different activation requirements of intracellular Ca²⁺ and DAG (K.-P. Huang, 1989). Classical/conventional PKC (cPKC) include PKCα, PKCβ, and PKCγ and are dependent on Ca²⁺ and DAG; Novel PKC (nPKC) include PKCδ, PKCε, PKCη, and PKCθ that are Ca²⁺ dependent but DAG independent; and atypical PKC (aPKC) that include PKCζ and PKCλ which are independent of both Ca²⁺ and DAG. We demonstrated that application of the DAG analogue, PMA, significantly increased AP firing frequency but depressed Kir currents, suggesting the involvement of cPKC and nPKC. Chelation of intracellular Ca²⁺ with BAPTA via the recording pipettes blocked NMB-elicited augmentation of AP firing and lessened NMB-induced suppression of Kir currents, suggesting the involvement of cPKC. Agreeably, activation of cPKC isoforms phosphorylate and inhibit GIRK channels (Niemeyer, Rinne, & Kienitz, 2019). Our results indicate that activation of BB1 receptors enhances the excitability of CeL neurons via PLCβ and PKC-mediated depression of GIRK type Kir channels. Consistent with our results, modulations of Kir channels (Pravetoni & Wickman, 2008; Victoria et al., 2016; Wydeven et al., 2014), PLCβ (McOmish, Burrows, Howard, & Hannan, 2008; Xiao et al., 2012) and PKC (Bowers, Collins, Tritto, & Wehner, 2000; Hodge et al., 2002; Lesscher et al., 2008; Liu, Feng, & Wang, 2014; Weeber et al., 2000) affect anxiety and fear responses.

Both bombesin-like peptides and the CeA are involved in modulating fear/anxiety and appetitive behaviors (Petrovich, 2011, 2013; Smith & Lawrence, 2018; Zanchi et al., 2017), although bombesin-like peptides and their activity in the amygdala in the modulation of these physiological functions have not been fully determined. Injection of bombesin into the CeA reduces food intake (Kyrkouli et al., 1987; Vigh et al., 1999) and food ingestion increases the release of bombesin-like peptides from the CeA (Zul Merali, Judy McIntosh, Pamela Kent, David Michaud, & Hymie Anisman, 1998). Consistent with our results, the function of PKC is required for amphetamine-mediated anorectic action (Hsieh, Yang, Chiou, & Kuo, 2005). NMB and GRP activate the hypothalamic-pituitary-adrenal axis resulting in increased release of ACTH and corticosterone (Z. Merali et al., 2006) which are closely involved in emotional regulation. BB1 receptor knockout mice displayed decreased emotionality (Yamada et al., 2002) and administration of a selective BB1 receptor agonist as well as antagonist paradoxically exerts anxiolytic effects (Bédard et al., 2007; Zul Merali, Tania Bédard, et al., 2006a) and attenuates the FPS response (Bédard et al., 2007). Furthermore, the bombesin-like peptides are released from the CeA in response to stressor exposure (Z. Merali, J. McIntosh, P. Kent, D. Michaud, & H. Anisman, 1998). Collectively, these results support that activation of BB1 receptors elicited by NMB modulate fear and anxiety responses. In our hands, microinjection of NMB significantly reduced FPS responses while increasing ASR with increasing intensity. We demonstrate a functional requirement of PKC for NMB-induced excitation of the CeA. In line with this, cPKC and nPKC isozymes are involved in G_q protein-mediated increased acoustic startle responses (Toth, Gresack, Hauger, Halberstadt, & Risbrough, 2013).

There are sex differences in the presentation of affective disorders (Rubinow & Schmidt, 2019) as well as sex differences in amygdala activation in response to emotionally relevant stimuli (Cahill, Uncapher, Kilpatrick, Alkire, & Turner, 2004). Our data indicate that NMB reduces FPS while enhancing ASR in both male and female rats. Due in part to limiting numbers of male and female rats and that bombesin-like peptide immunoreactivity in the limbic system does not vary with sex or estrous cycling (Micevych, Matt, & Go, 1988), the data from both sexes were pooled for the investigations of NMB on FPS and ASR. Because we did not determine the phase of the estrous cycle of the female rats in our study, one caveat of this study is that we cannot rule out the influences of the estrous cycle of the female rats on the effects of NMB. Whereas the estrous cycle of females affects anxiety-like behaviors and fear conditioning (Johnston & File, 1991), FPS phenomena are dependent on CeA activity, and are not subjected to sex hormone modulation in both male and female rats (M. Davis, 1989; Michael Davis, 2001; D. Toufexis, Davis, Hammond, & Davis, 2005; D. J. Toufexis, Davis, Hammond, & Davis, 2004).

Injection of bombesin into the CeA induces an increase in MAP (Brown & Gray, 1988), consistent with reports of an increased response of the autonomic nervous system in fear and anxiety conditions. We demonstrated microinjection of NMB into the CeA dose-dependently reduced MAP and HR via activation of BB1 receptors and depression of GIRK channels. As bombesin non-selectively activates BB1 receptors and BB2 receptors (González et al., 2009; Ohki-Hamazaki, 2000), the resultant increase in MAP by intra-CeA application of bombesin (Brown & Gray, 1988) may lend support to a hypothesis of distinct processes of each receptor on cardiovascular output from the CeA. Indeed, intraventricular application of GRP in trout produced significant increases in HR and MAP (Le Mével, Lancien, Mimassi, Kermorgant, & Conlon, 2015). Another possibility is that BB1 receptor activation disinhibits CeA GABAergic interneurons producing increases in the baroreceptor responses and decreases in MAP and HR. Activation of CeA inhibitory projections to the dorsal vagal complex result in a decrease of the baroreceptor reflex and subsequent increases in MAP and HR (Saha, 2005). As the CeA is populated by GABAergic interneurons that reciprocally inhibit each other to produce coordinated defensive responses (Haubensak et al., 2010; Moscarello & Penzo, 2022; Yu, Garcia da Silva, Albeanu, & Li, 2016), NMB-expressing microcircuits of the CeA may inhibit excitatory cardiovascular outputs from this nucleus. In addition, bombesin-like peptides have been demonstrated to interact with neuropeptide Y (Bungo et al., 2000; Mulholland & Simeone, 1993), a peptide documented to suppress central cardiac output (Walker, Grouzmann, Burnier, & Waeber, 1991; Yang, Yang, Tang, & Wang, 1992). NMB activates NPY-expressing neurons in the hypothalamic arcuate nucleus involved in energy homeostatic processing due to co-expression of BB1 receptors (Anthony N. van den Pol et al., 2009). As hypothalamic nuclei coordinate cardiovascular responses (Elsaafien et al., 2021) and these nuclei receive inputs from the CeA (Gray, Carney, & Magnuson, 1989; Weera, Shackett, Kramer, Middleton, & Gilpin, 2021), NMB microinjection into the CeA may alter cardiovascular output via NMB-expressing projections to the hypothalamus. Further characterization of NMB-specific projections of the CeA is warranted.

Data Availability Statement:

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