Genetic Activation of Locus Coeruleus Noradrenergic Neurons Modulates Cerebellar MF-GrC Synaptic Plasticity via Presynaptic α2-AR/PKA Signaling in Mice

TL;DR

This study shows that activating brain cells in mice that release a chemical called norepinephrine can change how connections in the cerebellum strengthen or weaken, using a specific signaling pathway.

Contribution

The study reveals a novel presynaptic α2A-AR/PKA signaling mechanism by which LC noradrenergic neurons modulate cerebellar synaptic plasticity.

Findings

Optogenetic activation of LC noradrenergic neurons impairs MF-GrC LTP via α2-AR signaling.

Facial stimulation-induced glutamate sensor fluorescence LTP is abolished by chemogenetic LC activation.

Optogenetically induced LTD is blocked by PKA inhibition but not by PKC inhibition.

Abstract

The aim of this study was to investigate how optogenetic or chemogenetic activation of locus coeruleus (LC) noradrenergic neurons modulates cerebellar mossy fiber-granule cell (MF-GrC) long-term synaptic plasticity in DBH-Cre mice. The results showed that facial stimulation induced MF-GrC long-term potentiation (LTP) under control conditions and that this LTP was impaired by optogenetic activation of LC noradrenergic neurons via α2-ARs. Meanwhile, facial stimulation induced glutamate sensor fluorescence LTP in the granular layer, which was abolished by chemogenetic activation of LC noradrenergic neurons. In the absence of N-methyl-D-aspartate receptor activity, optogenetic activation of LC noradrenergic neurons triggered facial stimulation-induced MF-GrC long-term depression (LTD) via α2A-ARs. Optogenetically induced MF-GrC LTD by LC noradrenergic neuron activation was abolished by protein kinase A (PKA) inhibition but not by protein kinase C inhibition. Immunofluorescence results revealed abundant α2A-AR expression in the granular layer, with particularly high levels in glomeruli. These results indicate that optogenetic activation of LC noradrenergic neurons impairs MF-GrC LTP by triggering presynaptic LTD via the α2A-AR/PKA signaling pathway. Collectively, these findings provide novel evidence that LC-derived noradrenergic afferents modulate synaptic plasticity within the cerebellar cortical circuitry in intact animals.

Locus coeruleus (LC) noradrenergic neurons project their axons to the cerebellar cortex and modulate cerebellar circuit function via distinct adrenergic receptor (AR) subtypes. The present study investigated the mechanism by which optogenetic activation of LC noradrenergic neurons modulates facial stimulation-evoked long-term synaptic plasticity at cerebellar mossy fiber-granule cell (MF-GrC) synapses in urethane-anesthetized DBH-Cre mice. Blockade of GABAA receptors, 20 Hz facial stimulation induced MF-GrC long-term potentiation (LTP) in the control group, and this LTP was impaired by optogenetic activation of LC noradrenergic neurons via α2-ARs. Meanwhile, facial stimulation induced LTP of glutamate sensor fluorescence in the granular layer, which was abolished by chemogenetic activation of LC noradrenergic neurons. Following NMDA receptor blockade, optogenetic activation of LC noradrenergic neurons triggered facial stimulation-induced MF-GrC long-term depression (LTD) via α2A-ARs. Optogenetically activated LC noradrenergic neuron-induced MF-GrC LTD was abolished by protein kinase A (PKA) inhibition but not by protein kinase C inhibition. Immunofluorescence results revealed abundant α2A-AR expression in the granular layer, with particularly high levels in glomeruli, and no colocalization with the glutamate sensor. These results indicate that optogenetic activation of LC noradrenergic neurons impairs facial stimulation-induced MF-GrC LTP by triggering presynaptic LTD via the α2A-AR/PKA signaling cascade.

Figures

Full text

1. Introduction

In the cerebellum, mossy fibers (MFs) are the major excitatory glutamatergic inputs to the cerebellar cortex. Boutons of MFs are large, contain multiple active zones, and make synapses with tens of granular cells (GrCs). Each MF bouton and its GrC dendritic targets are ensheathed by glia within specialized structures known as glomeruli. Sensory information transmitted via MFs induces glutamate release, thereby eliciting excitatory postsynaptic currents (EPSCs) in GrCs [1], as well as long-term synaptic plasticity at MF-GrC synapses [2,3]. Long-term MF-GrC synaptic plasticity was first demonstrated by Roggeri and colleagues [2,3], who found that 4 Hz facial stimulation induced MF-GrC long-term depression (LTD) dependent on γ-aminobutyric acid type A (GABA_A_) receptor activity, whereas the same stimulation elicited MF-GrC long-term potentiation (LTP) in the presence of a GABA_A_ receptor antagonist in vivo in rats. D’Errico and colleagues reported that distinct glutamate receptor subtypes converge onto a final common mechanism that translates the frequency and duration of MF discharges into the regulation of the LTP-LTD balance, a process that may play an important role in shaping spatiotemporal signal transformations at the cerebellar input stage [4]. Our previous results showed that 20 Hz facial stimulation evoked a N-methyl-D-aspartate (NMDA) receptor-nitric oxide (NO) cascade-dependent LTP at MF-GrC synapses in the absence of GABA_A_ receptor activity in vivo in mice [3]. Importantly, MF-GrC LTP/LTD has been proposed to play crucial roles in the cerebellum’s adaptation to excitatory inputs from native MFs and is also considered a leading cellular substrate for motor learning and memory in the brain [3,4,5,6,7,8,9].

Noradrenaline (NA) is a neuromodulator essential for cognition, alertness, and attention [10,11,12]. Following the discovery of cerebellar monoaminergic innervation by locus coeruleus (LC) neurons, the cerebellum has emerged as a structure where “point-to-point” and “global” neural circuits converge; monoamines may exert widespread effects on neurons via both direct synaptic transmission and paracrine signaling [13]. Anatomical studies have shown that noradrenergic fibers originating from the LC project to the cerebellar cortex via multilayered fiber tracts, with innervation of the molecular, granular, and Purkinje cell layers mediated by distinct subtypes of adrenergic receptors (ARs) [14]. AR subtypes are classified into α (α1 and α2) and β subtypes, all of which are widely distributed throughout the cerebellar cortex. In situ hybridization results have shown that the cerebellar cortex expresses α-AR mRNAs, with high levels of α2A-AR and α2B-AR mRNAs detected in cerebellar GrCs [15]. Immunohistochemical studies indicate that α2A-AR is abundantly expressed in the cerebellar cortical granule layer (GL) in rats and monkeys [16,17,18]. Moreover, behavioral studies have shown that deletion of the α2A-AR gene in mice leads to impaired motor coordination and increased anxiety-like behaviors, suggesting that α2-ARs play important roles in the GL of the cerebellar cortex [19].

Carey and Regehr [20] demonstrated that NA controlled the short- and long-term associative plasticity of parallel fiber-Purkinje cell synapses by activation of α2-ARs in vitro in mice. A growing body of evidence shows that specific activation of postsynaptic α2A-ARs on dendritic spines within prefrontal cortical neuronal circuits inhibits K^+^ channel activity via the cyclic adenosine monophosphate-Protein Kinase A (cAMP-PKA) signaling cascade, thus strengthening network connectivity, enhancing prefrontal cortical neuronal firing, and improving prefrontal cortical cognitive functions [21]. Furthermore, NA (released synaptically or applied exogenously) reduces the excitability of anterior piriform cortex pyramidal neurons via α2 receptors [22]. Under in vivo conditions, NA suppresses facial stimulation-evoked synaptic transmission at MF-GrC synapses via activation of α2-ARs [23] but facilitates facial stimulation-induced LTD at molecular layer interneuron-Purkinje cell synapses via the α2A-AR/PKA signaling cascade in the mouse cerebellar cortex [24]. Together, ARs are critical for synaptic transmission and plasticity within cerebellar neuronal circuits, suggesting that LC noradrenergic neurons modulate synaptic function of the cerebellar neuronal circuitry via distinct AR subtypes in intact animals. However, the mechanisms by which LC noradrenergic neurons modulate facial stimulation-evoked long-term plasticity at MF-GrC synapses remain unknown. Therefore, we herein investigated the mechanisms underlying the modulation of facial stimulation-evoked long-term synaptic plasticity at cerebellar MF-GrC synapses by optogenetic or chemogenetic activation of LC noradrenergic neurons in DBH-Cre mice.

2. Materials and Methods

2.1. Animals Preparation

2.1.1. Animals

DBH-Cre mice (stock #T005671) were purchased from GemPharmatech Co., Ltd. (Nanjing, China). Primers DBH-F: GGGCAG-TCTGG-TACTTCCAAGCT and DBH-R: ACTGTGTCTTTGCACG AACGTG (408 bp); Actin-F: CAGCAAAACCTGGCTGTGGATC and Actin-R: ATGAGCCACC ATGTGGGTGTC (412 bp) were used for genotype identification [22] (Figure 1A,B). A total of 120 DBH-Cre mice (60 male and 60 female) aged 8–12 weeks were used in all experiments.

All experimental procedures were approved by the Animal Care and Use Committee of Yanbian University and conducted in accordance with the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals (approval number: SYXK (Ji) 2025-0005, Approval Date: 20 August 2025). Mice were group-housed on a 12 h light (6:00–18:00) schedule, with free access to food and water, under constant temperature (24 ± 1 °C) and humidity (50 ± 5%). For electrophysiological study, the adult DBH-Cre mice (n = 92; both sexes equally) were randomly divided into control group and drug administration groups.

2.1.2. Stereotaxic Surgery and Viral Nano Injection

Mice were anesthetized in an induction chamber with 4% isoflurane (R500, RWD, Shenzhen, China) and positioned in a stereotaxic frame under sterile conditions (68801, RWD, Shenzhen, China). Anesthesia was maintained at 1–2% isoflurane throughout the surgical procedures. For in vivo fiber photometry experiments, adult DBH-Cre mice received unilateral injections of 300 nl of a glutamate sensor virus (AAV2/9-hSyn-SF-iGluSnFR.A184S-WPRE, Obio, Shanghai, China) into three sites within the cerebellar Crus II [anteroposterior (AP): −7.0 mm; mediolateral (ML): +3.02, +3.12, and +3.22 mm; dorsoventral (DV): −2.25 mm]. In experiments involving optogenetic or chemogenetic activation of LC noradrenergic neurons, mice were additionally injected bilaterally with 300 nl of AAV2/9-EF1α-DIO-hChR2(H134R)-mCherry-WPRE-hGH-pA and AAV-EF1a-DIO-mCherry-WPRE-hGH-polyA (BrainVTA, Wuhan, China) or AAV2/9-hSyn-DIO-hM3D(Gq)-mCherry-WPRE-pA and AAV2/9-hSyn-DIO-mCherry-WPRE-pA (Taitool, Shanghai, China) per side into the LC (AP: −5.45 mm; ML: ±1.25 mm; DV: −3.8 mm). Viral injections were performed using a Nanoject III microinjector (Drummond Scientific, Broomall, PA, USA) equipped with glass micropipettes (tip diameter ~20 μm) fabricated using a micropipette puller (P97, Sutter Instrument, Novato, CA, USA). Viruses were infused at a rate of 1 nl/s, and the pipette was left in place for an additional 10 min after injection to minimize backflow. One week before the optogenetic experiments, mice were implanted with bilateral optical fibers (R-FOC-BL200c-39NA, RWD, Shenzhen, China) targeting the LC, which were secured using dental adhesive (C&B Metabond, Parkell Inc., Brentwood, NY, USA). Mice were allowed to recover for at least 4 weeks following viral vector injection to ensure sufficient viral expression prior to optogenetic and chemogenetic experiments. After completion of in vivo electrophysiology experiments, mice were perfused to verify viral expression and projection mapping, as well as to confirm the placement of optical implants (Figure 1A–D).

2.1.3. Optogenetic/Chemogenetic Stimulation

The LC was additionally activated by optogenetic stimulation or chemogenetic manipulation using Clozapine-N-oxide (CNO). Prior to optical stimulation, optical fibers were connected to the implanted fibers using ferrule sleeves (LC-1.25, RWD, Shenzhen, China). Blue light (470 nm) generated by an optogenetic light source (Aurora400, Newdoon Inc., Hangzhou, China) was delivered simultaneously to the bilateral LC through bilateral optic fibers (OFJ-F-36-0.22-1.25-2C, Newdoon Inc., Hangzhou, China). Since 1–5 Hz photostimulation can sufficiently activate LC neurons and induce endogenous NE release, we applied 5 Hz optical stimulation to the LC (470 nm, 60 pulses) via an implanted optical fiber [25], paired with 20 Hz facial stimulation [3], to investigate the effect of LC noradrenergic neuronal activity on 20 Hz facial stimulation-induced MF-GrC LTP in the mouse cerebellar cortex. For chemogenetic activation of LC neurons, CNO was dissolved in artificial cerebrospinal fluid (ACSF: 125 mM NaCl, 3 mM KCl, 1 mM MgSO_4_, 2 mM CaCl_2_, 1 mM NaH_2_PO_4_, 25 mM NaHCO_3_, and 10 mM D-glucose) and applied to the cerebellar surface at a flow rate of 0.5 mL/min.

2.2. In Vivo Electrophysiological Recordings and Facial Stimulation

After constructing a watertight chamber, a craniotomy (1–1.5 mm in diameter) was performed to expose the cerebellar surface corresponding to Crus II, and the dura mater was carefully removed. A peristaltic pump (Gilson Minipulse 3; Villiers-le-Bel, Val-d’Oise, France) was used to perfuse the cerebellar surface with oxygenated ACSF at a flow rate of 0.5 mL/min. Rectal temperature was continuously monitored using a thermostatic device and maintained at 37.0 ± 0.2 °C.

The cerebellar Crus II region was visualized using a Nikon Eclipse FN1 microscope (Nikon Corp., Chiyoda-ku, Tokyo, Japan). Recording electrodes (3–5 MΩ) were fabricated using a PC-100 puller (Narishige, Shinjuku-ku, Tokyo, Japan) from thick wall borosilicate glass, which was filled with ACSF and with resistances of 3–5 MΩ. Local field potential recordings from GL were performed at depths of 300–400 μm under the pia mater membrane and identified according to our previous studies [3,23] in the absence of GABAergic inhibitory inputs. Local field potential recordings were performed with an Axopatch 700B amplifier (Molecular Devices, Foster City, CA, USA) under current clamp conditions (I = 0) and acquired through a Digidata 1550 series analog-to-digital interface through a personal computer using Clampex 10.4 software. Tactile stimulation of the ipsilateral whisker pad was delivered via air puffs (10 ms, 60 psi) through a 12-gauge stainless steel tube connected to a pressurized injection system (Picospritzer^®^ III; Parker Hannifin, Fairfield, NJ, USA). Air puff stimulation was computer-controlled and synchronized with electrophysiological recordings using a Master-8 controller (A.M.P.I., Jerusalem, Israel) and Clampex 10.4 software (Molecular Devices, Sunnyvale, CA, USA). All drugs were dissolved in ACSF and applied to the cerebellar surface. For induction of long-term MF-GC synaptic plasticity, 20 Hz air puff stimulation (10 ms, 60 psi, 240 pulses) was delivered 10 min after facial stimulation-evoked response traces became stable [3]. In Figure 7, paired air puff stimuli (10 ms, 50–60 psi, 50 ms interval) were used to evaluate whether activation of LC noradrenergic neurons triggered MF-GrC LTD accompanied by a change in the paired-pulse ratio (PPR). The PPR was calculated as the N2 amplitude divided by the N1 amplitude, which is commonly used to assess whether events occur at presynaptic or postsynaptic sites [3,26,27].

2.3. Glutamate Sensor Fluorescence Signal Acquisition

Four weeks later, AAV2/9-iGluSnFR.A184S and AAV2/9-DIO-hM3D(Gq)-mCherry-injected DBH-Cre mice (n = 24) were anesthetized with urethane (1.3 g/kg, intraperitoneal injection), tracheotomized, and fixed in a customized stereotaxic frame. After constructing a watertight chamber, the cerebellar surface corresponding to Crus II was exposed, and the dura mater was carefully removed. The fiber optics (R-FOC-BL200c-39NA, RWD, Shenzhen, China) were implanted into Crus II of the cerebellar cortex. The fibers were implanted vertically to a depth of 300–400 μm to target the GL glutamate probe-expressing region of Crus II. The cerebellar surface was perfused with oxygenated ACSF (containing 20 μM SR95531) at a flow rate of 0.5 mL/min and protected from light.

Fiber photometry recordings were performed using a commercial system (R821, RWD, Shenzhen, China), as previously described [28]. Briefly, excitation light at 470 nm and 410 nm was first directed into a fluorescence cube and then coupled into the optical fibers. The laser output at the fiber tip was measured with a power meter and adjusted to 50 μW before each experiment to ensure consistent excitation intensity. The 410 nm signal served as an isosbestic control to correct for motion artifacts. Emitted fluorescence signals from the glutamate sensor and control channel were collected at a sampling rate of 125 Hz. The signals were filtered through a 525 ± 25 nm bandpass filter, converted to electrical signals by a high-sensitivity silicon photodetector (integrated in the R821 system), and transmitted to the real-time data acquisition unit. Air puff stimulation (60 psi, 100 ms, 250 ms, and 500 ms) was applied to the ipsilateral whisker pad to evoke dynamic glutamate release in the cerebellar cortex. For induction of long-term MF-GrC synaptic plasticity, 20 Hz air puff stimulation (10 ms, 60 psi, 240 pulses) was delivered 10 min after the evoked MF-GrC synaptic responses had stabilized [3].

Baseline fluorescence (F_0_) was defined as the mean signal during a pre-stimulus window (−5 to 0 s relative to the onset of air puff stimulation). Normalized fluorescence changes were then expressed as ΔF/F (%). For each trial, fluorescence traces were time-locked to the onset of air puff stimulation (0 s). Signals were averaged across 20 repeated trials for each animal and then pooled across animals to calculate group means. Statistical analyses were performed on trial-averaged data unless otherwise specified.

2.4. Brain Slice Preparation, Whole-Cell Patch Clamp Recording and Optogenetic Stimulation of LC Neurons

Brain slices were prepared as described previously [29]. Briefly, the mice (n = 4) were anesthetized with isoflurane and decapitated quickly. The brain was immediately removed and placed into ice-cold oxygenated ACSF containing the following (in mM): 200 sucrose, 3 KCl, 2 CaCl_2_, 2 MgCl_2_, 26 NaHCO_3_, 1.25 NaH_2_PO_4_, and 10 D-glucose. The solution was bubbled with 95% O_2_-5% CO_2_ (pH 7.40). Coronal transverse pontine sections (300 μM) containing the LC area were prepared using a vibrating slicer (Leica VT1200S; Leica Biosystems Nussloch GmbH, Nussloch, Germany). The slices were transferred to normal ACSF in which sucrose was substituted with 124 mM NaCl and incubated at room temperature (24–25 °C) for 1 h before being used for recordings.

Whole-cell patch clamp recordings from LC neurons were visualized using a 40 × water immersion lens using a Nikon microscope (Eclipse FN1, Nikon Corp., Tokyo, Japan). The recording electrodes were pulled from borosilicate glass capillaries (World Precision Instruments, Sarasota, FL, USA) by a micropipette puller (model PC-10, NARISHIGE, Tokyo, Japan). The pipette resistance was 3–5 MΩ. The components of recording electrode solution contained (in mM): 130 K gluconate, 10 KCl, 10 HEPES, 2 Mg-ATP, 0.3 Na-GTP, and 0.4 EGTA (pH 7.3 with KOH, osmolarity adjusted to 300 mOsm). The ACSF solution was applied to the bath, containing the following (in mM): 124 NaCl, 3 KCl, 1.3 NaH_2_PO_4_, 2 MgCl_2_, 10 D-glucose, 26 NaHCO_3_, and 2 CaCl_2_ (pH 7.40 bubbled with 95% O_2_-5% CO_2_). Membrane potentials were monitored with an Axopatch 700B amplifier (Molecular Devices, Foster City, CA, USA), filtered at 2 kHz, and acquired through a Digidata 1440 series analog-to-digital interface on a personal computer using Clampex 10.4 software (Molecular Devices, Foster City, CA, USA). To confirm whether optogenetic stimulation could evoke the activity of LC noradrenergic neurons in the mouse cerebellar cortex, optical stimulation of the LC neurons (470 nm, 5 Hz, 10 pulses) [25] was delivered to acute brain slices via an optical fiber (Figure 1E).

2.5. Histological Examination

To determine AAV-DIO-hChR2(H134R)-mCherry and AAV-DIO-hM3D(Gq)-mCherry expression and assess colocalization with DBH, 6 mice were anesthetized with an intraperitoneal injection of 2,2,2-tribromoethanol (250 mg/kg) three weeks after AAV injection and perfused with ice-cold PBS followed by 4% paraformaldehyde. Brains were removed; post-fixed overnight; and dehydrated in 10%, 20%, and 30% sucrose at 4 °C. Coronal LC slices (40 μm) were cut using a freezing microtome (CM1900; Leica, Germany). Slices were first incubated with 0.3% Triton X-100 for 15 min at room temperature and then with 10% goat serum for 2 h. Next, slices were incubated with rabbit anti-DBH (1:2000; ABclonal, Wuhan, China) overnight at 4 °C, followed by donkey anti-rabbit DyLight 488 (1:500; Invitrogen, Carlsbad, CA, USA) for 2 h at room temperature. After several washes with PBS, sections were mounted with DAPI. Fluorescence images were acquired using a confocal laser scanning microscope (Nikon Eclipse C1, Tokyo, Japan).

To visualize viral expression and confirm optical fiber placement, mice were anesthetized with an intraperitoneal injection of 2,2,2-tribromoethanol after electrophysiological recordings and perfused with ice-cold PBS followed by 4% paraformaldehyde. Brains were removed and post-fixed overnight under the same fixation conditions. Coronal slices (40 μm) were cut using a freezing microtome, washed three times with PBS (5 min each), and mounted. Fluorescence images were acquired using a confocal laser scanning microscope.

To determine iGluSnFR.A184S probe expression and assess colocalization with α2A-AR, 3 mice were anesthetized with 2,2,2-tribromoethanol three weeks after AAV-iGluSnFR.A184S injection and perfused with ice-cold PBS followed by 4% paraformaldehyde. Brains were removed and post-fixed overnight. Sagittal cerebellar slices (40 μm) were cut using a freezing microtome. Slices were first incubated with 0.3% Triton X-100 for 15 min at room temperature and then with 10% goat serum for 2 h. Next, slices were incubated with rabbit anti-α2A-AR (1:1000; Sigma-Aldrich, Shanghai, China) overnight at 4 °C, followed by goat anti-rabbit Cyanine5 (Invitrogen, 1:500) for 2 h at room temperature. After several washes with PBS, sections were mounted with DAPI. Fluorescence images were acquired using a confocal laser scanning microscope (A1R, Nikon Corp., Tokyo, Japan).

2.6. Chemicals

Reagents included urethane; yohimbine (YHB), an α2-AR antagonist; and chelerythrine (protein kinase C inhibitor), all purchased from Sigma-Aldrich (Shanghai, China). KT5720 (selective protein kinase A inhibitor), D-2-amino-5-phosphonovaleric acid (D-APV; selective NMDA receptor antagonist), SR95531 hydrobromide (GABA_a_ receptor antagonist), CNO (an activator of hM3Dq DREADDs), and BRL44408 (BRL; an α2A-AR antagonist) were purchased from Tocris Cookson (Bristol, UK). 2,2,2-Tribromoethanol was purchased from GLPBIO (Montclair, CA, USA). For GABA_a_ receptor blockade, the GABA_a_ receptor antagonist SR95531 (20 μM) was added to ACSF in all experiments. For experiments involving PKA and protein kinase C (PKC) inhibition, KT5720 and chelerythrine were applied at least 30 min before recording and maintained throughout the experiment (Figures 8 and 9). For blockade of NMDA receptor-dependent components, the selective NMDA receptor antagonist D-APV (50 μM) was added to ACSF in selected experiments (Figures 6–9).

2.7. Statistical Analysis

All experiments were performed with post hoc power analysis using G*Power 3.1, and a statistical power greater than 0.8 (α = 0.05) was used to determine the sample size (n) for all comparisons. Electrophysiological data were analyzed with Clampfit 10.7 software (Molecular Devices, Foster City, CA, USA). The 10 min preceding 20 Hz stimulation delivery were used to calculate average baseline values. Post-stimulation values were calculated 40–50 min after the stimulation train. For each recording, the normalized amplitudes and area under the curve (AUC) of response waves and iGluSnFR fluorescence before (Pre) and after (Post) 20 Hz stimulation were divided by the mean of all baseline measurements in each group and then multiplied by 100. All data are expressed as mean ± S.E.M. The normality of experimental data was examined using the Kolmogorov-Smirnov test, and parametric or non-parametric statistics were applied accordingly. Paired t-tests and two-way repeated-measures ANOVA with Tukey’s post hoc test were used to determine statistical significance between experimental groups (SPSS 17.0; Chicago, IL, USA). p-values less than 0.05 were considered statistically significant.

3. Results

3.1. Expression of Optogenetic and Chemogenetic Viruses in LC Noradrenergic Neurons

To investigate the effects of optogenetic or chemogenetic activation of LC noradrenergic neurons on facial stimulation-induced cerebellar MF-GrC LTP in mice, we first verified the expression of optogenetic and chemogenetic viral constructs in LC noradrenergic neurons (Figure 1). For selective activation of LC noradrenergic neurons, we used a Cre-dependent viral strategy to drive specific expression of humanized Channelrhodopsin-2 [hChR2(H134R)] and the Gq-coupled human M3 muscarinic receptor DREADD [hM3D(Gq)] in LC noradrenergic neurons of DBH-Cre mice (Figure 1). Stereotaxic injections of AAV-DIO-hChR2(H134R)-mCherry and AAV-DIO-hM3D(Gq)-mCherry targeting the LC were performed in DBH-Cre mice (see Methods), which restricted expression of hChR2 (red; Figure 1C) and hM3D (red; Figure 1D) specifically to dopamine β-hydroxylase-positive neurons in the LC (i.e., noradrenergic neurons). Notably, hChR2 and hM3D were found in the granular layer (GL) of the cerebellar cortex, where they exhibited a scattered punctate distribution (Figure 1D). To confirm that optogenetic activation increases excitability of LC noradrenergic neurons, we performed ex vivo whole-cell patch clamp recordings to measure spontaneous firing frequency. Optical stimulation significantly increased the firing rate to 4.8 ± 0.4 Hz compared to baseline (3.9 ± 0.4 Hz; F(1,9) = 40.3, p < 0.001; n = 11 cells; Figure 1E). These results indicate that LC noradrenergic neurons express functional hChR2 and project axons into the cerebellar GL in DBH-Cre mice.

3.2. Optogenetic Activation of LC Noradrenergic Neurons Suppresses Facial Stimulation-Induced Cerebellar MF-GrC LTP In Vivo

Consistent with the previous studies [3,30], 20 Hz air puff facial stimulation (10 ms, 240 pulses) induced LTP of MF-GrC synaptic transmission when GABA_A_ receptors were blocked by SR95531 (20 μM). As shown in Figure 2, delivery of 20 Hz facial stimulation induced a significant increase in the amplitude of N under control conditions that persisted for over 50 min (Figure 2A,B). During 40–50 min after 20 Hz facial stimulation, the normalized amplitude (post) was 123.0 ± 4.4% of that before induction stimulation (pre: 100.0 ± 1.6%; F_(1,14)_ = 31.3, p < 0.001; n = 8 recordings; Figure 2C), and the normalized AUC (post) was 124.0 ± 4.1% of that before the 20 Hz stimulation (pre: 100.0 ± 2.3%; F_(1,14)_ = 26.2, p < 0.001 vs. pre; n = 8 recordings; Figure 2D). These results indicate that 20 Hz facial stimulation induces MF-GrC LTP in the mouse cerebellar cortical Crus II in vivo.

To investigate whether the activity of LC noradrenergic neurons modulates facial stimulation-induced MF-GrC LTP in the mouse cerebellar cortex, 20 Hz facial stimulation was paired with optical stimulation of the LC (470 nm, 5 Hz, 60 pulses) delivered via an implanted optical fiber. Notably, paired with optical stimulation of the LC, 20 Hz facial stimulation failed to induce MF-GrC LTP in the mouse cerebellar cortex (Figure 2A,B). The normalized amplitude of N (Post) during 40–50 min after 20 Hz facial stimulation was 100.6 ± 3.6% of baseline (Pre:100.0 ± 2.0%; F_(1,14)_ = 0.02, p = 0.88; n = 8 recordings; Figure 2C), which was significantly lower than that in control conditions (123.0 ± 4.4%, F_(1,14)_ = 15.3, p = 0.002; Figure 2C), and the normalized AUC of N (Post) was 98.5 ± 4.2% of baseline (Pre:100.0 ± 2.3%; F_(1,14)_ = 0.10, p = 0.75; n = 8 recordings; Figure 2D), which was significantly lower than that in control conditions (124.0 ± 4.1%; F_(1,14)_ = 18.8, p < 0.001; Figure 2D). In addition, the mean baseline N amplitude was 0.65 ± 0.05 mV in the control group, which was similar to that in the optogenetic condition (0.63 ± 0.06 mV; F_(1,14)_ = 0.04, p = 0.73; n = 8 recordings per group). After completion of in vivo electrophysiology experiments, mice were perfused to verify viral expression and projection mapping, as well as to confirm the placement of optical implants (Figure 2E,F). The results indicate that optical stimulation of the LC suppresses facial stimulation-induced MF-GrC LTP in the mouse cerebellum in vivo.

Behavioral studies have shown that deletion of the mouse α2A-AR gene leads to impaired motor coordination and increased anxiety-like behaviors, suggesting that α2-ARs play a critical role in modulation of MF-GrC synaptic activity in vivo in mice [19]. We then applied a selective α2-AR antagonist, YBH (50 μM), onto the cerebellar surface to examine whether optical stimulation of the LC suppressed facial stimulation-induced MF-GrC LTP in the mouse cerebellum via α2-ARs. When paired with optical stimulation of the LC, 20 Hz facial stimulation failed to induce MF-GrC LTP in the control group but induced MF-GrC LTP in the presence of YBH (Figure 3A,B). In the presence of YBH, the normalized amplitude of N (Post) during 40–50 min after 20 Hz facial stimulation was 123.4 ± 4.4% of baseline (Pre:100.0 ± 2.6%; F_(1,14)_ = 21.85, p < 0.001; n = 8 recordings; Figure 3C), which was significantly higher than that in control conditions (98.3 ± 3.4%, F_(1,14)_ = 21.22, p < 0.001; Figure 3C), and the normalized AUC of N (Post) during 40–50 min after 20 Hz facial stimulation was 124.0 ± 5.0% of baseline (Pre: 100.0 ± 2.9%; F_(1,14)_ = 13.39, p = 0.003; n = 8 recordings; Figure 3D), which was significantly higher than that in control conditions (99.8 ± 3.9%; F_(1,14)_ = 23.13, p < 0.001; Figure 3D). These results indicate that, when α2-AR in the cerebellar cortex is blocked, optogenetic activation of LC noradrenergic neurons fails to modulate facial stimulation-induced MF-GrC LTP in the cerebellar cortex, suggesting that optogenetic activation of LC noradrenergic neurons impairs MF-GrC LTP through activation of α2-ARs in the cerebellar cortex in vivo in mice.

3.3. Chemogenetic Activation of LC Noradrenergic Neurons Impairs Facial Stimulation-Induced LTP of Glutamate Fluorescence (glu-LTP) in the Mouse Cerebellar Cortex

Fs are the major glutamatergic excitatory inputs to the cerebellar cortex. However, direct comparisons of glutamate release patterns have not yet been performed in the cerebellar granular layer. Recently, an optimized fluorescent probe for imaging glutamate neurotransmission, the intensity-based glutamate-sensing fluorescent reporter (iGluSnFR), has been developed, which enables direct and specific reporting of excitatory synaptic glutamate release [31]. To elucidate the spatiotemporal dynamics of glutamatergic signaling in the cerebellum during tactile sensory integration, the optimized glutamate sensor iGluSnFR.A184S was selectively expressed in cerebellar neurons of DBH-Cre mice (Figure 4A). In the presence of the GABA_a_ receptor antagonist SR95531 (20 μM), air puff stimulation with durations of 100 ms, 250 ms, and 500 ms evoked iGluSnFR.A184S fluorescence responses (Figure 4B–D). The mean latency of the fluorescence responses was 24.55 ± 1.02 ms (n = 24), and the response durations were 155.44 ± 4.0 ms (100 ms air puff; n = 8), 305.5 ± 4.6 ms (250 ms air puff; n = 8), and 555.7 ± 5.4 ms (500 ms air puff; n = 8), respectively. The amplitudes of the fluorescence signals were 0.19 ± 0.008, 0.19 ± 0.005, and 0.19 ± 0.008 (field of view ΔF/F (%); n = 8), respectively. The normalized fluorescence amplitudes were 100.0 ± 4.3%, 99.7 ± 2.7%, and 98.6 ± 4.0% (Figure 4C; n = 8), whereas the normalized fluorescence AUC values were 100.0 ± 1.8%, 391.3 ± 40.4%, and 801.48 ± 25.3% (Figure 4D; n = 8), respectively. These results indicate that air puff stimulation evokes glutamate fluorescence signals in the GL of the mouse cerebellum in vivo.

To selectively target hM3Dq, the Gq-coupled excitatory DREADDs, to noradrenergic neurons, we infused an AAV-hSyn-DIO-hM3Dq-mCherry into the LC and an AAV-hSyn-SF-iGluSnFR.A184S-WPRE into cerebellar GL of DBH-Cre mice (Figure 5A). Prior to performing fiber photometry recordings, mice were administered CNO (1 µM) to the cerebellar surface at a flow rate of 0.5 mL/min [32]. Glu-LTP was induced by 20 Hz facial stimulation under control conditions, which was characterized by a sustained increase in both the intensity and AUC of the iGluSnFR.A184S fluorescence response that lasted for more than 50 min (Figure 5B). Between 40 and 50 min after 20 Hz facial stimulation, the normalized intensity of the fluorescence signal (Post) was 132.0 ± 9.5% of baseline (Pre: 100.0 ± 4.0%; F_(1,14)_ = 56.9, p < 0.001; n = 8 mice; Figure 5B,D). The normalized AUC of the fluorescence signal during 40–50 min after 20 Hz stimulation (Post) was 148.5 ± 10.6% of baseline (Pre: 100.0 ± 3.3%; F_(1,14)_ = 55.7, p < 0.001; n = 8 mice; Figure 5B,E). These results indicate that 20 Hz facial stimulation (240 pulses) induces glu-LTP at MF-GrC synapses in Crus II of the mouse cerebellar cortex, suggesting that 20 Hz facial stimulation elicits LTP of glutamate signals at MF terminals.

Notably, chemogenetic activation of the LC noradrenergic neurons was achieved by perfusing CNO (1 µM) onto the cerebellar surface [32]. CNO completely prevented 20 Hz facial stimulation-induced MF-GrC glu-LTP (Figure 5C). The normalized intensity of the fluorescence signal during 40–50 min (post) after 20 Hz facial stimulation was 105.2 ± 5.0% of baseline (pre: 100.0 ± 4.0%; F_(1,14)_ = 0.54, p = 0.47; n = 8 mice; Figure 5C,D), which was significantly lower than that under control conditions (132.0 ± 9.5%; F_(1,14)_ = 16.7, p < 0.001; Figure 5D). Additionally, the normalized AUC of the fluorescence response (Post) after 20 Hz facial stimulation was 101.9 ± 4.0% of baseline (Pre: 100.0 ± 3.6%; F_(1,14)_ = 0.16, p = 0.69; n = 8 mice; Figure 5E), which was significantly lower than that under control conditions (148.5 ± 10.6%; F_(1,14)_ = 27.12, p < 0.001; Figure 5E). In addition, we included a control virus group using AAV2/9-hSyn-DIO-mCherry-WPRE-pA to address the potential concern regarding off-target effects of CNO. The results showed that CNO application failed to produce any detectable changes in MF-GrC synaptic transmission in control virus-expressing mice, indicating that the observed effects were not attributable to nonspecific pharmacological actions of CNO (Supplementary Figure S1). These results indicate that chemogenetic activation of LC noradrenergic neurons prevents facial stimulation-induced MF-GrC glu-LTP, which suggests that upregulation of noradrenergic signaling impairs MF-GrC LTP by inhibiting glutamate release from MF terminals in the mouse cerebellum in vivo.

3.4. Blockade of NMDA Receptor, Optogenetic Activation of LC Noradrenergic Neurons Triggers 20 Hz Stimulation-Induced MF-GrC LTD via Activation of α2A-ARs in the Mouse Cerebellar Cortex

Since 20 Hz facial stimulation-induced MF-GrC LTP depends on NMDA receptor activity in the mouse cerebellum in vivo [3,33], we next investigated the effect of optogenetic activation of LC noradrenergic neurons on facial stimulation-induced MF-GrC synaptic plasticity in the presence of an NMDA receptor antagonist. Blockade of NMDA receptors with D-APV (50 μM) abolished 20 Hz facial stimulation-induced MF-GrC LTP in the control group [3] (Figure 6A–C) but revealed 20 Hz facial stimulation-induced MF-GrC LTD in the optogenetic stimulation of the LC group (Figure 6A,B). In the control group, the normalized amplitude of N (Post) was 103.7 ± 2.7% of baseline (Pre: 100.0 ± 2.30%; F_(1,14)_ = 0.42, p = 0.53; n = 8 recordings; Figure 6C), and the normalized AUC of N (Post) was 100.8 ± 2.9% of baseline (Pre: 100.0 ± 2.6%; F_(1,14)_ = 0.03, p = 0.87; n = 8 recordings; Figure 6D). In the optogenetic stimulation of the LC group (D-APV + optogenetic stimulation of LC), the normalized amplitude of N (Post) was 79.0 ± 3.4% of baseline (Pre: 100.0 ± 2.0%; F_(1,14)_ = 48.71, p < 0.001; n = 8 recordings; Figure 6C), which was significantly lower than that of the control group (103.7 ± 2.7%; F_(1,14)_ = 26.37, p < 0.001; Figure 6C). The normalized AUC of N (Post) was 80.5 ± 3.8% of baseline (Pre: 100.0 ± 2.1%; F_(1,14)_ = 23.70, p < 0.001; n = 8 recordings; Figure 6D), which was significantly lower than that of the control group (100.8 ± 2.9%; F_(1,14)_ = 20.78, p < 0.001; Figure 6D). These results indicate that optogenetic stimulation of the LC elicits NMDA receptor-independent MF-GrC LTD, suggesting that optogenetic activation of LC noradrenergic neurons triggers a long-term decrease in glutamate release from MF terminals in the mouse cerebellar cortex in vivo.

We further examined whether activation of LC noradrenergic neurons triggered MF-GrC LTD accompanied by a change in the paired-pulse ratio (PPR), N2/N1. In the presence of D-APV, activation of LC noradrenergic neurons triggered MF-GrC LTD accompanied by a significant increase in the PPR (Figure 7). The normalized amplitude of N1 was 79.58 ± 2.58% of baseline (100.0 ± 2.51%) during 40–50 min after 20 Hz facial stimulation (F_(1,14)_ = 26.37, p < 0.001, one-way ANOVA, n = 8 recordings; Figure 7C), and the PPR was 117.44 ± 3.15% of baseline (100.0 ± 2.56%) during 40–50 min after 20 Hz facial stimulation (F_(1,14)_ = 45.30, p = 0.001, one-way ANOVA; n = 8 recordings; Figure 7D). These results indicate that activation of LC noradrenergic neurons triggers MF-GrC LTD accompanied with a significant increase in PPR, suggesting that optogenetic activation of LC noradrenergic neurons triggers LTD of glutamate release from MF terminals in the mouse cerebellar cortex in vivo.

Since optogenetic activation of the LC noradrenergic neurons impairs MF-GrC LTP via activation of α2-ARs, we then perfused a selective α2-AR antagonist, YBH (50 μM), onto the cerebellar surface to examine whether optogenetic activation of the LC noradrenergic neurons triggers MF-GrC LTD through activation of α2-ARs in the mouse cerebellar cortex in vivo. Blockade of α2-ARs, 20 Hz facial stimulation paired with optogenetic stimulation of the LC failed to induce MF-GrC LTD in the mouse cerebellar cortex in vivo (Figure 8). In the presence of a mixture of D-APV and YBH, the normalized amplitude of N (Post) during 40–50 min after 20 Hz facial stimulation was 102.0 ± 3.6% of baseline (Pre: 100.0 ± 3.1%; F_(1,10)_ = 0.20, p = 0.66; n = 6 recordings; Figure 8C), and the normalized AUC of N (Post) during 40–50 min after 20 Hz facial stimulation was 100.6 ± 2.2% of baseline (Pre:100.0 ± 2.2%; F_(1,10)_ = 0.02, p = 0.89; n = 6 recordings; Figure 8D). These results indicate that optogenetic activation of the LC noradrenergic neurons triggers MF-GrC LTD through the activation of α2-ARs in the mouse cerebellar cortex in vivo.

Moreover, we perfused a selective α2A-AR antagonist, BRL (50 μM), onto the cerebellar surface to examine whether optogenetic activation of the LC noradrenergic neurons triggers MF-GrC LTD via α2A-ARs in the mouse cerebellar cortex in vivo. Similar to YHB, blockade of α2A-ARs with BRL, 20 Hz facial stimulation paired with optogenetic stimulation of the LC could not induce MF-GrC LTD in the mouse cerebellar cortex (Figure 8A,B). In the presence of a mixture of D-APV and BRL, the normalized amplitude of N (Post) during 40–50 min after 20 Hz facial stimulation was 98.7 ± 3.9% of baseline (Pre: 100.0 ± 2.9%; F_(1,10)_ = 0.38, p = 0.55; n = 6 recordings; Figure 8C), and the normalized AUC of N (Post) during 40–50 min after 20 Hz facial stimulation was 99.1 ± 4.3% of baseline (Pre: 100.0 ± 2.4%; F_(1,10)_ = 0.09, p = 0.76; n = 6 recordings; Figure 8D). These results indicate that optogenetic activation of LC noradrenergic neurons triggers MF-GrC LTD through the activation of α2A-ARs in the mouse cerebellar cortex in vivo.

3.5. Optogenetic Activation of LC Noradrenergic Neurons Triggered Facial Stimulation-Induced MF-GrC LTD via the PKA Signaling Pathway

We further examined whether optogenetic activation of the LC noradrenergic neurons triggers facial stimulation-induced MF-GrC LTD through the PKA signaling cascade by cerebellar surface by applying a potent PKA inhibitor, KT5720 (1 µM). In the presence of D-APV + KT5720, this completely prevented 20 Hz facial stimuli-induced MF-GrC plasticity in both control and optogenetic stimulation of the LC groups (Figure 9A,B). In the control group (D-APV + KT5720), the normalized amplitude of N (post) was 97.7 ± 3.9% of baseline (Pre: 100.0 ± 2.7%; F_(1,14)_ = 0.001, p = 0.99; n = 8 recordings; Figure 9C), and the normalized AUC of N (Post) was 102.7 ± 2.5% of baseline (Pre: 100.0 ± 2.0%; F_(1,14)_ = 0.11, p = 0.74; n = 8 recordings; Figure 9D). In optogenetic stimulation of the LC groups (D-APV + KT5720 + LC optogenetic stimulation), the normalized amplitude of N (post) was 99.2 ± 2.9% of baseline (Pre: 100.0 ± 2.8%; F_(1,14)_ = 0.02, p = 0.89; n = 8 recordings; Figure 9C), which was similar with that of the control group (97.7 ± 3.9%; F_(1,14)_ = 0.02, p = 0.89; Figure 9C). The normalized AUC of N (Post) was 101.2 ± 3.2% of baseline (Pre: 100.0 ± 2.5%; F_(1,14)_ = 0.001, p = 0.99; n = 8 recordings; Figure 9D), which was not significantly different than that of the control group. These results indicate that optogenetic stimulation of the LC triggers facial stimulation-induced MF-GrC LTD through the PKA signaling pathway in the mouse cerebellar cortex.

In addition, we examined whether optogenetic stimulation of the LC triggers facial stimulation-induced MF-GrC LTD through the PKC signaling cascade via cerebellar surface perfusion of the PKC inhibitor chelerythrine (Che; 30 μM). In the presence of D-APV, inhibition of PKC failed to prevent optogenetic stimulation of the LC-triggered MF-GrC LTD in the mouse cerebellar cortex (Figure 10A,B). In the control group (D-APV + chelerythrine), the normalized amplitude of N (post) was 98.7 ± 3.4% of baseline (Pre: 100.0 ± 2.7%; F_(1,14)_ = 0.001, p = 0.94; n = 8 recordings; Figure 10C), and the normalized AUC of N (post) was 99.3 ± 2.9% of baseline (Pre: 100.0 ± 3.1%; F_(1,14)_ = 0.20, p = 0.66; n = 8 recordings; Figure 10D). In the presence of D-APV and chelerythrine, optogenetic stimulation of the LC still triggered MF-GrC LTD (Figure 10A,B). The normalized amplitude of N (Post) was 76.4 ± 3.6% of baseline (Pre: 100.0 ± 2.5%; F_(1,14)_ = 29.83, p < 0.001; n = 8 recordings; Figure 10C), which was significantly lower than that in the control group (98.7 ± 3.4; F_(1,14)_ = 18.65, p < 0.001; Figure 10C). The normalized AUC of N (post) was 79.0 ± 2.1% of baseline (pre: 100.0 ± 2.2%; F_(1,14)_ = 49.44, p < 0.001; n = 8 recordings; Figure 10D), which was significantly lower than that in the control group (99.3 ± 2.9; F_(1,14)_ = 41.14, p < 0.001; Figure 10D). These results demonstrate that optogenetic stimulation of the LC elicits facial stimulation-evoked MF-GrC LTD via the PKA signaling pathway rather than the PKC signaling cascade in the cerebellar cortex of mice in vivo.

In addition, we examined the expression of iGluSnFR.A184S fluorescence and α2A-AR immunofluorescence in the GL of the mouse cerebellar cortex using confocal laser scanning microscopy. α2A-ARs were abundantly expressed in the GL of the mouse cerebellar cortex and were distributed throughout the hemispheric lobules. Rosettes are the terminal enlargements of MFs in the GL (Figure 11A,B) and form cerebellar glomeruli together with dendrites of GrCs and axon terminals of Golgi cells, constituting a distinct synaptic complex. The glutamate sensor iGluSnFR.A184S was mainly distributed in postsynaptic structures of cerebellar glomeruli, whereas α2A-ARs were also abundantly expressed in cerebellar glomeruli, with moderate expression in GrCs. Almost no overlapping yellow fluorescence was observed in the GL (Figure 11B). These results indicate that α2A-ARs are abundantly distributed in cerebellar glomeruli, particularly in MF-GrC presynaptic terminals, suggesting that α2A-AR activation contributes to MF-GrC LTD triggered by optogenetic activation of LC noradrenergic neurons in mice in vivo.

4. Discussion

The present study demonstrates that 20 Hz facial sensory stimulation induces LTP of MF-GrC synaptic transmission in the mouse cerebellar cortex, and this effect is completely prevented by optogenetic activation of LC noradrenergic neurons via α2A-ARs. Similarly, facial stimulation-induced glu-LTP in the granular layer is abolished by chemogenetic activation of LC noradrenergic neurons. Blockade of NMDA receptors, optogenetic activation of LC noradrenergic neurons triggers facial stimulation-induced MF-GrC LTD via activation of α2A-ARs and requires PKA signaling cascades. Immunofluorescence results show that α2A-ARs are abundantly expressed in the granular layer of the cerebellar cortex, particularly in cerebellar glomeruli, but do not colocalize with the glutamate sensor. These results indicate that optogenetic activation of LC noradrenergic neurons suppresses facial stimulation-induced MF-GrC LTP by triggering presynaptic LTD via the α2A-AR/PKA signaling cascade.

4.1. Activation of LC Noradrenergic Neurons Triggers MF-GrC LTD and Impairs LTP via α2-ARs in the Cerebellar Cortex

MFs originating from diverse sources form synapses on the dendrites of GrCs within cerebellar glomeruli and also contact the basolateral dendrites of Golgi cells [34]. Golgi cells are GABAergic interneurons that provide inhibitory input to cerebellar GrCs via GABA_a_ receptors, thereby modulating MF-GrC glutamatergic synaptic transmission. In the cerebellar cortex, Golgi cells regulate GrC excitability primarily through GABA_a_ receptor activation. To isolate facial stimulation-evoked MF-GrC synaptic transmission by eliminating the confounding influence of Golgi cell-mediated inhibition, a GABA_A_ receptor antagonist was applied throughout all experiments. However, blockade of GABA_A_ receptors can alter circuit excitability, MF-GrC synaptic transmission, and the thresholds for long-term plasticity. Our previous work demonstrated that facial stimulation evokes inhibition in the granular layer via GABA_A_ receptor activation; blockade of these receptors abolishes the inhibitory components but enhances the excitatory components of MF-GrC synaptic responses [35]. In vivo studies in rats have shown that, under conditions of intact GABA_A_ receptor-mediated inhibition, 4 Hz facial stimulation induces MF-GrC LTD, whereas the same stimulation protocol induces LTP when inhibition is removed [2]. In contrast, 20 Hz facial stimulation evokes N-methyl-D-aspartate (NMDA) receptor/nitric oxide (NO) cascade-dependent LTP at MF-GrC synapses in the presence of a GABA_a_ receptor blocker in vivo in mice [3]. Collectively, these studies indicate that GABA_a_ receptor-mediated inhibition plays a critical role in shaping facial stimulation-evoked MF-GrC synaptic transmission and plasticity. Although GABA_A_ receptor blockade has inherent limitations, the present experiments necessitated this approach to record pure MF-GrC synaptic responses. Under these conditions, 20 Hz facial stimulation induces NMDA receptor-dependent MF-GrC LTP, a form of sensory stimulation-induced long-term plasticity implicated in cerebellum-related motor learning [3]. The results of the present study demonstrate that pairing 20 Hz air puff stimulation with optogenetic activation of LC noradrenergic neurons prevents the induction of MF-GrC LTP. These findings suggest that activation of LC noradrenergic neurons impairs facial stimulation-induced MF-GrC LTP in the mouse cerebellar cortex.

LC neuronal activity is induced by various stress stimuli: mild stressful stimulation produces moderate NA release, whereas intense stressful stimulation induces robust NA release. It has been demonstrated that photostimulation frequencies of 1, 3, 5, 10, 20, and 40 Hz effectively reproduce distinct levels of endogenous NA release. Compared with 20 and 40 Hz photostimulation, 1–5 Hz stimulation mimics moderate tonic NA release dynamics and results in sustained NA release [25]. The present results showed that 5 Hz optical stimulation of the LC impaired 20 Hz facial stimulation-induced MF-GrC LTP, indicating that 5 Hz optical stimulation sufficiently activates LC noradrenergic neurons and induces sufficient endogenous NA release to modulate MF-GrC synaptic plasticity. It has been demonstrated that NA impairs LTP in various brain regions via activation of α2-ARs, including the amygdala, occipital cortex, and dorsal horn of the spinal cord in rodents [36]. In contrast, α2-AR agonism has been shown to promote LTP at excitatory synapses in the mouse olfactory bulb [34], suggesting that the role of α2-AR in regulating synaptic plasticity is region- and synapse-specific in the brain. Previous studies have demonstrated that α-AR mRNAs are expressed in the cerebellar cortex, with high levels of α2A-AR and α2B-AR mRNAs expressed in the GL [15], and α2A-AR immunoreactivity is also abundantly detected in the GL of rats and monkeys [16,17,18], suggesting that α2-AR activation contributes to the modulation of the GrC network, particularly MF-GrC synaptic transmission and plasticity. In vitro, activation of α2-ARs regulates the short- and long-term plasticity of parallel fiber-Purkinje cell synapses in the mouse cerebellar cortex [20]. Previous work has demonstrated that α2-AR agonists reduce the gain of vestibulo-ocular reflexes and inhibit gain adaptation of the optokinetic response in mice, suggesting that α2-AR activation modulates cerebellar neuronal circuit function [37].

Under in vivo conditions, NA activates α2-ARs to induce suppression of facial stimulation-evoked MF-GrC synaptic transmission [38], whereas it facilitates facial stimulation-induced GABAergic LTD at molecular layer interneuron-Purkinje cell synapses via the PKA signaling cascade in the mouse cerebellar cortex [24]. The present results showed that blockade of α2-ARs in the cerebellar cortex, optogenetic stimulation of the LC noradrenergic neurons failed to prevent facial stimulation-induced MF-GrC LTP in the cerebellar cortex, suggesting that activation of the LC noradrenergic neurons impairs MF-GrC LTP via α2-ARs in vivo in mice. Since 20 Hz facial stimulation-induced MF-GrC LTP was NMDA receptor-dependent under control conditions, and NMDA receptor activity is absent in this context [3,39], we further investigated the effect of α2-AR activation on facial stimulation-induced MF-GrC plasticity in the absence of NMDA receptor activity. Interestingly, following blockade of NMDA receptor-dependent LTP, optogenetic stimulation of the LC noradrenergic neurons triggers facial stimulation-induced MF-GrC LTD in the mouse cerebellar cortex. Furthermore, 20 Hz facial stimulation paired with optogenetic stimulation of the LC failed to induce MF-GrC LTD in the absence of α2-AR activity, indicating that optogenetic activation of the LC noradrenergic neurons triggers MF-GrC LTD through the activation of α2-ARs in the mouse cerebellar cortex in vivo. Moreover, blockade of α2A-ARs with BRL, 20 Hz facial stimulation paired with optogenetic stimulation of the LC could not induce MF-GrC LTD, indicating that optogenetic activation of LC noradrenergic neurons triggers MF-GrC LTD via α2A-ARs in the mouse cerebellar cortex in vivo. In addition, the immunofluorescence experiments show that α2A-ARs are expressed in the GL of the mouse cerebellar cortex, particularly in the presynaptic structures of glomeruli. Therefore, these results suggest that optogenetic activation of LC noradrenergic neurons triggers MF-GrC LTD via α2A-ARs and impairs the NMDA receptor-dependent LTP in the mouse cerebellar cortex in vivo.

4.2. Chemogenetic Activation of LC Noradrenergic Neurons Triggers Facial Stimulation-Induced glu-LTD and Impairs glu-LTP in the Mouse Cerebellar Cortex

Viral and transgenic iGluSnFR tools have potential utility in studies of normal physiology, as well as in neurologic and psychiatric pathologies in which abnormalities in glutamatergic signaling are implicated [40]. More recently, an optimized fluorescent probe for visualizing glutamate neurotransmission was developed, an iGluSnFR [31]. iGluSnFR was engineered in vitro to maximize its fluorescence dynamic range, and its utility for visualizing glutamate release by neurons and astrocytes was validated in increasingly intact neurological systems. Sensors for neurotransmitter concentration enable monitoring of brain activity and provide a more direct measure of regional functional activity that is less dependent on nonlinearities associated with voltage-gated ion channels [40]. Since iGluSnFR can directly and specifically report glutamate release at excitatory synapses [31], we employed iGluSnFR.A184S to visualize facial stimulation-evoked glutamate release at MF-GrC synapses and evaluate 20 Hz facial stimulation-evoked long-term changes in glutamate release in vivo in mice. The results showed that 20 Hz facial stimulation-evoked glutamate release was associated with an increase in fluorescence intensity in the GL of the mouse cerebellar cortex in vivo. The fluorescence density of the iGluSnFR sensor exhibited long-term enhancement at MF-GrC synapses following 20 Hz facial stimulation, indicating that facial stimulation elicited glu-LTP at MF terminals in the mouse cerebellar cortex. Notably, chemogenetic activation of LC noradrenergic neurons induced by CNO administration to the cerebellar surface completely prevented 20 Hz facial stimulation-induced MF-GrC glu-LTP, suggesting that activation of noradrenergic neurons impairs MF-GrC LTP by inhibiting glutamate release from MF terminals in the mouse cerebellar cortex in vivo. The present results also showed that activation of LC noradrenergic neurons triggered MF-GrC LTD accompanied by a significant increase in PPR. Since an increase in PPR has been proposed to reflect a reduction in presynaptic glutamate release probability [3,26,27], these results suggest that optogenetic activation of LC neurons induces presynaptic MF-GrC LTD in the mouse cerebellar cortex in vivo.

4.3. Activation of Noradrenergic Neurons Induces MF-GrC LTD via the α2A-AR/PKA Signaling Cascade

The α2-ARs are known to be Gi/o-coupled metabotropic receptors [41,42,43] that negatively regulate adenylyl cyclase activity and inhibit voltage-gated Ca^2+^ channel activity [42]. Activation of presynaptic α2-ARs inhibits N-type calcium channel activity, neurotransmitter release, and PKA activity, which in turn decreases the phosphorylation of downstream receptors [43]. Activation of α2A-ARs inhibits LTP via the Gi protein-adenylyl cyclase-protein kinase signaling pathway in the medial prefrontal cortex [44]. In the cerebellar cortex, a higher concentration of α2-AR antagonist UK14304 induces suppression of facial sensory stimulation-evoked MF-GrC synaptic transmission [23], whereas the present results showed that optogenetic stimulation of the LC noradrenergic neurons triggers facial stimulation-induced MF-GrC LTD through the PKA signaling pathway in the mouse cerebellar cortex. The effect was prevented by PKA inhibition, indicating that optogenetic activation of the LC noradrenergic neurons modulates MF-GrC synaptic transmission and long-term plasticity via the α2A-AR/PKA signaling cascade. In contrast, inhibition of PKC failed to prevent MF-GrC LTD triggered by optogenetic stimulation of the LC noradrenergic neurons during facial stimulation, indicating that MF-GrC LTD triggered by optogenetic stimulation of the LC noradrenergic neurons is not through the PKC signaling pathway.

4.4. Physiological Significance of Optogenetic Stimulation of the LC Noradrenergic Neuron-Induced Cerebellar MF-GrC LTD

Long-term synaptic plasticity in cerebellar circuits—including that at parallel fiber-Purkinje cell, molecular layer interneuron-Purkinje cell, and MF-GrC synapses—is considered a cellular mechanism underlying cerebellar motor learning and memory [3,8,45]. It is well established that long-term plasticity at MF-GrC synapses plays critical roles in cerebellar learning and motor adaptation [46,47]. Furthermore, noradrenergic fibers originating from the LC project to the cerebellar cortex via multilayered fiber tracts, regulating neuronal activity and synaptic function in the molecular layer, granular layer, and Purkinje cell layer [14]. This noradrenergic innervation is thought to play an important modulatory role in sensory information processing and physiological processes such as motor coordination and motor learning [48,49,50]. That NA modulates cerebellar circuit function via α2-AR activation has been widely investigated under both in vitro and in vivo conditions [20,23,24,51]. However, the results of the present study showed that optogenetic or chemogenetic activation of the LC noradrenergic neurons triggered a novel form of MF-GrC LTD via presynaptic α2A-ARs, which in turn impaired facial stimulation-induced NMDA receptor-dependent LTP in the mouse cerebellar cortex. These results suggest that noradrenergic fibers originating from the LC contribute to the regulation of motor learning by modulating MF-GrC synaptic plasticity via activation of presynaptic α2A-ARs in the cerebellar cortex in vivo. The findings provide novel evidence that LC-derived noradrenergic afferents modulate synaptic plasticity within the cerebellar cortical circuit under in vivo conditions.

4.5. Limitations and Future Directions

The present in vivo experiments were performed under urethane anesthesia. Urethane anesthesia depresses neuronal excitability by activating barium-sensitive potassium leak conductance, without affecting excitatory glutamatergic and GABAergic inhibitory synaptic transmission [52]. Urethane has been widely used in studies of neuromodulatory systems, including LC-mediated noradrenergic signaling [53,54] and in vivo facial stimulation-evoked synaptic transmission in mice [3]. However, urethane anesthesia may suppress chemogenetic or optogenetic activation of LC noradrenergic neurons, leading to reduced modulation of facial stimulation-evoked MF-GrC synaptic transmission and plasticity. Therefore, it is necessary to employ awake, non-anesthetized mice to investigate the modulatory roles of LC noradrenergic neuronal activation on cerebellar MF-GrC synaptic transmission and cerebellum-dependent learning or motor adaptation in future studies. In addition, the present study demonstrated that 20 Hz facial stimulation-induced MF-GrC LTP was impaired by simultaneous activation of LC noradrenergic neurons in the mouse cerebellar Crus II in vivo. The experiments were confined to cerebellar cortical Crus II and whisker pad-evoked sensory inputs in mice. However, given the anatomical and functional heterogeneity of cerebellar lobules, as well as regional differences in LC projection density and neuromodulatory influence, the results cannot be interpreted as universally applicable across all cerebellar regions or sensory modalities in rodents. Future studies will be required to determine whether similar mechanisms operate in other cerebellar lobules or sensory contexts, such as sensory input elicited by limb stimulation in awake animals.

5. Conclusions

Anatomical studies have shown that noradrenergic fibers originating from the LC project to the cerebellar cortex via multilayered fiber tracts, with innervation of neuronal circuit function mediated by distinct subtypes of ARs. The present study highlights that LC-derived noradrenergic afferents regulate synaptic plasticity within the cerebellar cortical circuitry, particularly MF-GrC synaptic plasticity, in intact animals. First, α2A-ARs are abundantly expressed in the GL of the mouse cerebellar cortex and form cerebellar glomeruli together with GrC dendrites and Golgi cell axon terminals, with particularly high expression at MF-GrC presynaptic terminals. Furthermore, optogenetic activation of LC noradrenergic neurons triggers MF-GrC LTD and impairs NMDA receptor-dependent LTP. Moreover, chemogenetic activation of LC noradrenergic neurons completely abrogates facial stimulation-induced MF-GrC glu-LTP in the mouse cerebellar cortex. Together, these results suggest that activation of LC noradrenergic neurons impairs MF-GrC LTP by inhibiting glutamate release from MF terminals via α2A-AR activation in the mouse cerebellar cortex in vivo.

The present results show that optogenetic activation of LC noradrenergic neurons triggers facial stimulation-induced MF-GrC LTD via α2A-AR activation in the mouse cerebellar cortex. Notably, inhibition of PKA, but not PKC, completely prevents the MF-GrC LTD triggered by optogenetic activation of LC noradrenergic neurons, indicating that optogenetic activation of LC noradrenergic neurons modulates MF-GrC synaptic transmission and long-term plasticity via the α2A-AR/PKA signaling cascade. The present findings provide novel evidence that LC-derived noradrenergic afferents modulate synaptic plasticity within the cerebellar cortical circuitry in intact animals.