The experimental procedures were approved by the Animal Care and Use Committee of Yanbian University and were in accordance with the animal welfare guidelines of the U.S. National Institutes of Health. The permit number is SYXK (Ji) 2011–006. Cerebellar slices preparation has been previously described (Xuan et al., 2018). In short, after deeply anaesthetization with halothane, adult (4˗6 weeks old) ICR (Institute of Cancer Research) mice were decapitated immediately. The cerebellum was sectioned and placed in ice-cold artificial cerebrospinal fluid (ACSF) containing (in mM): 125NaCl, 3KCl, 1 MgSO₄, 2 CaCl₂, 1.25 Na₂HPO₄, 25 NaHCO₃, and 10 D-glucose bubbled with 95% O₂ and 5% CO₂ (pH 7.4; 295–305 mOsm). The sagittal slices of cerebellar cortex (300 μm thick) were prepared using a Vibratome (VT 1200s, Leica, Nussloch, Germany). The slices were incubated at least 1 h in a submerged chamber filled with 95%O₂ and 5% CO₂ equilibrated ACSF at room temperature (24–25°C) prior to recording.

Whole-cell patch-clamp recordings from cerebellar PCs were visualized using a 60x water-immersion lens through a Nikon microscopy (Eclipse FN1, Nikon Corp., Tokyo, Japan). Patch electrodes contained a solution of the following (in mM): potassium gluconate 120, HEPES 10, EGTA 1, KCl 5, MgCl2 3.5, NaCl 4, biocytin 8, Na₂ATP 4, and Na₂GTP 0.2 (pH 7.3 with KOH, osmolarity adjusted to 300 mOsm). Patch pipette resistances were 4˗6 MΩ in the bath, with series resistances in the range of 10–20 MΩ. Membrane potentials and/or currents were monitored with an Axopatch 700B amplifier (Molecular Devices, Foster City, CA, United States), filtered at 5 kHz, and acquired through a Digidata 1,440 series analog-to-digital interface on a personal computer using Clampex 10.4 software (Molecular devices, Foster City, CA, United States). Under voltage clamp recording mode, cells were held in voltage-clamp mode at −70 mV. Series resistance was monitored by applying voltage pulses (10 ms, 5 mV), and only neurons with stable series resistance were include in the analysis. SR95531 (10 μM) and kynurenic acid (KYC; 1 mM) were included in the ACSF to block the GABAergic and glutamatergic synaptic transmission. TTX (500 nM) and BaCl₂ (100 μM) were added to the ACSF to block the voltage gated sodium channel and Ba²⁺-sensitive potassium channel, so as to separate the IH current (Cardenas et al., 1999; Qiu et al., 2003). IH current is activated by hyperpolarization current or membrane potential. For some experiments under current-clamp mode, the membrane potential was hyperpolarized to −80 mV by injecting negative currents (−20 to −30 pA) under control conditions.

Mice (n = 4) were deeply anesthetized with an intraperitoneal injection of 7% chloral hydrate (5 ml/kg), and then transcardially perfused by cold phosphate buffer (PBS) at pH 7.4, followed by 4% paraformaldehyde (PFA, Sinopharm Chemical Reagent Co., China) PBS solution. Brain was post-fixed in 4% PFA for 48 h at 4°C, and washed with PBS 5 times at 10 min intervals. Separate the cerebellum from the brain with a razor blade. The cerebellum was exposed to 10% sucrose, 20% sucrose and 30% sucrose in PBS for more than 6 h. After embedding in Tissue-Tek O.C.T. Compound (Beijing zhongshan Jin qiao Biotechnology Co, China), the cerebellum was quickly frozen in −80°C refrigerator for 2 h. Then cerebellum was sectioned into 8 μm slices in the sagittal plane using a freezing microtome (CM1900, Leica, Germany). Slices were stored at 4°C for immunohistochemical experiments.

Microscope slides were permeabilized with 0.3% Triton X-100 in PBS, and then were blocked (10% donkey serum in PBS) and incubated in a primary antibody (rabbit anti-GLP-1 receptor, 1:1,000, Sigma-Aldrich), followed by Cy3 labeled goat anti-rabbit antibody (Life Tech, 1:200) and 4′,6-diamidino-2-phenylindole (DAPI, 1:1,000). The incubation time for the primary antibody was overnight at 4°C. For the secondary antibody and DAPI, the incubation time was 2 h at room temperature. Microscope slides with slices were then washed 3 times in PBS covered with a coverslip and sealed with nail polish. Fluorescence images were acquired by confocal laser-scanning microscope (Nikon C2, Tokyo, Japan) (Wang et al., 2018; Wu et al., 2020). A large region images including the desired region was obtained on Nikon C2 laser confocal system with 10X objective.

GLP-1 (100 nM), Exendin 9–39 (100 nM), Exendin-4 (100 nM), TTX (500 nM), SR95531 (10 μM), kynurenic acid (KYC; 1 mM), ZD7288 (50 μM), KT5720 (500 nM), and H89 (10 μM) were used in the experiments. These drugs were bought from Sigma-Aldrich (Shanghai, China). All drugs were dissolved in solution and kept in frozen in aliquots, and were applied to the cerebellar slices at 0.5 ml/min in ACSF through the peristaltic pump. In the experiments involving PKA inhibitor, application of KT5720 or H89 was began at least 20 min before patch-clamp recording and continuing throughout the experiments.

Electrophysiological data were analyzed using Clampfit 10.3 software. All data are expressed as the mean ± S.E.M. IH was determined by subtracting I-Instant from I-steady at each hyperpolarizing voltage step using the equation: IH = I-steady – I-instant. Student’s paired t-test (SPSS software) was used to determine the level of statistical significance between groups of data. p-values below 0.05 were considered to indicate a statistically significant difference between experimental groups.

We first examined the distribution of GLP1 receptor in the mouse cerebellar cortex using immunohistochemistry. As shown in Figure 1, GLP1 receptor was present in the cerebellum and was most concentrated in the PC layer. This indicates that GLP1 receptor is expressed in mouse cerebellar PCs.

Perfusion of GLP1 (100 nM) increased the membrane excitability of PCs, which was manifested by a significant increase in the frequency of spike firing discharge elicited by depolarization current, accompanied by significant depolarization of membrane potential (Figure 2A). After administration of GLP1 (100 s), the number of spike firing discharge induced by depolarization current (50 pA, 200 ms) was 15.0 ± 0.85, which was significantly higher than that in the ACSF group (ACSF: 10.75 ± 0.80; p < 0.05, n = 8, Figure 2B). The GLP1-induced depolarization of membrane potential was 4.45 ± 0.44 mV, which was significantly different than that in the ACSF group (ACSF: 0.36 ± 0.29 mV; p < 0.05, n = 8, Figure 2C). To clarify whether GLP1 receptor activation can increase the excitability of PCs, we used the GLP1 receptor agonist, exendin-4 (100 nM). Perfusion of exendin-4 (100 nM) increased the frequency of spike firing (Figure 2C). In the presence of exendin-4 (100 s), the frequency of spike firing evoked by depolarization current (50 pA, 200 ms) was 14.25 ± 0.91, which was significantly higher than that in the ACSF group (ACSF: 10.5 ± 0.82; p < 0.05, n = 8, Figure 2D). GLP1 induced depolarization of membrane potential was 3.93 ± 0.33 mV, which was significantly different than that in the ACSF group (p < 0.05 versus ACSF, n = 8, Figure 2E). These results indicate that GLP1 receptor activation enhances the excitability of PCs.

We then used a GLP1 receptor blocker, exendin-9-39 (100 nM), to determine whether GLP1 increased the number of spike discharges by PCs through its receptor. Perfusion of exendin-9-39 had no significant effect on PC excitability, but completely blocked the GLP1-induced increase in spike discharges (Figure 3A). In the presence of exendin-9-39 (100 nM), the number of spike discharges elicited by depolarization current (50 pA, 200 ms) was 10.9 ± 1.28, which was similar to that in the ACSF group (p > 0.05, n = 7, Figure 3B), and the change in membrane potential was 1.11 ± 0.32 mV, which was not significantly different from that in the ACSF group (0.87 ± 0.33 mV; p > 0.05, n = 7, Figure 3C). In the presence of exendin-9-39 and GLP1, the number of spike discharges induced by depolarization current (50 pA, 200 ms) was 10.7 ± 1.46, which was similar to that in the ACSF group (p > 0.05, n = 7, Figure 3B). The change in membrane potential was 1.15 ± 0.29 mV, which was not significantly different from that in the ACSF group (0.87 ± 0.33 mV; p > 0.05, n = 7, Figure 3C). These results indicate that GLP1 activates its receptors on PCs, resulting in membrane depolarization and increased frequency of spike discharge.

Under current-clamp recording conditions, GLP1 significantly depressed after-hyperpolarization potential, resulting in significantly reduced amplitude and area under the curve (AUC) (Figure 4A). In the presence of GLP1, after-hyperpolarization potential amplitude was 10.7 ± 0.96 mV, which was significantly lower than that of baseline (ACSF: 14.5 ± 0.93 mV; p < 0.05, n = 8, Figure 4B), and the AUC of the after-hyperpolarization potential was 897.6 ± 56.0 mV/ms, which was significantly different from that of baseline (ACSF: 1225.3 ± 95.0 mV/ms; p < 0.05, n = 8, Figure 4C). In voltage-clamp recording mode, GLP1 significantly depressed the amplitude of the outward rectified current (IR), resulting in decreased amplitude and AUC of IR (Figure 4D). In the presence of GLP1, the IR amplitude was 39.6 ± 4.0 pA, which was significantly lower than that of baseline (ACSF: 49.1 ± 5.0 pA; p < 0.05, n = 8, Figure 4E); and the AUC of the IR was 722.2 ± 57.0 pA/ms, which was significantly different than that of baseline (ACSF: 874.2 ± 86 pA/ms; p < 0.05, n = 8, Figure 4F). These results indicate that GLP1 depressed the after-hyperpolarization potential of the action potentials in cerebellar PCs.

IH plays a key role in controlling the periodicity of continuous and intermittent action potential firing (Maccaferri and McBain, 1996; Ghamari-Langroudi and Bourque, 2000). IH channels are expressed in cerebellar PCs and participate in modulating simple spike firing activity (Crepel and Penit-Soria, 1986; Roth and Häusser, 2001) and in controlling the steady state of membrane potential (Southan et al., 2000). Activation of G protein-coupled receptors can regulate the activities of IH channels through cAMP, cGMP and/or Ca²⁺ signaling pathways, thereby modulating the excitability of neurons (Pape, 1996; Ludwig et al., 1998; Biel et al., 2002; Qiu et al., 2003). To understand the effect of GLP1 on IH activity, we examined the effect of GLP1 on IH channel activity.

In the presence of TTX (500 nM) and BaCl₂ (100 μM), application of a series of hyperpolarized membrane potentials (−70 to −130 mV, 20 mV/step, 1 s) induced three kinds of membrane current, instant current (I-instant), steady current, and I-tail, which were significantly enhanced by perfusion of GLP1 (100 nM, Figure 5A). The GLP1 induced enhancement of I-instant (Figure 5B), I-steady (Figure 5C) and IH (I-steady − I-instant; Figure 5D) was evoked by hyperpolarizing membrane potential. GLP1 also significantly enhanced I-tail, which reflected the enhancement of IH (Figure 5E). These results indicated that GLP1 facilitates the IH in cerebellar PCs.

To confirm that GLP1 enhanced the activity of the IH channel, we employed an IH channel blocker, ZD7288 (50 μM) and then observed the effect of GLP1 on the hyperpolarizing potential-evoked membrane current (Figure 6). Administration of GLP1 significantly enhanced the I-instant, I-steady, IH current and I-tail evoked by hyperpolarizing potential (−130 mV, 1 s). Application of ZD7288 blocked IH, and abolished the GLP1-enhanced membrane current (Figure 6A). In the presence of ZD7288, the I-instant was 71.9 ± 6.9 pA, which was significantly different than that of the ACSF group (149. 9 ± 10.1 pA; p < 0.05, n = 7, Figure 6B). Additional application of GLP1 failed to increase the I-instant. In the presence of ZD7288 and GLP1, the I-instant was 70.4 ± 5.7 mV, which was not significantly different from that of ZD7288 alone (p > 0.05, n = 7, Figure 6B). In the presence of ZD7288, the I-steady induced by hyperpolarization stimulation was 77.4 ± 7.0 pA, which was significantly different from that of the ACSF group (207.1 ± 15.8 pA; p < 0.05, n = 7, Figure 6C). Additional application of GLP1 failed to increase the I-steady. In the presence of ZD7288 and GLP1, the I-steady was 77.6 ± 5.2 pA, which was not significantly different from that of ZD7288 alone (p > 0.05, n = 7, Figure 6C). In the presence of ZD7288 and GLP1, IH was 7.1 ± 2.4 pA, which was not significantly different from that with ZD7288 alone (5.6 ± 2.2 pA; p > 0.05, n = 7, Figure 6D). In the presence of ZD7288 and GLP1, I-tail was 13.7 ± 3.0 mV, which was similar to that with ZD7288 alone (13.6 ± 8.3 pA; p > 0.05, n = 7, Figure 6E). These results indicate that blockade of IH channels can abolish the GLP1-induced enhancement of the membrane current in cerebellar PCs, indicating that GLP1 enhances the activity of IH channels and increases the excitability of PCs through GLP1 receptors.

To understand the GLP1-induced depolarization through activation of hyperpolarization-activated cationic channels, we observed the effect of GLP1 on the excitability of PCs in the presence of ZD7288 (50 μM). Application of ZD7288 significantly decreased the excitability of PCs, which showed a significant decrease in the number of spike discharges elicited by depolarization current, accompanied by significant hyperpolarization of membrane potential (Figure 7A). In the presence of ZD7288, the number of spike discharges elicited by depolarization current (50 pA, 200 ms) was 8.0 ± 0.77, which was significantly lower than that in the ACSF group (11.8 ± 1.11; p < 0.05, n = 6, Figure 7B), and the change in membrane potential was −4.58 ± 0.45 mV, which was significantly lower than that in the ACSF group (0.54 ± 0.17 mV; p < 0.05, n = 6, Figure 7C). Notably, IH blockade abolished the effect of GLP1 on PC excitability. In the presence of ZD7288 and GLP1 (100 nM), the number of spike firing discharges elicited by depolarization current was 8.2 ± 0.77, which was not significantly different than that with 7,288 alone (ACSF; p < 0.05, n = 6, Figure 7B). The change in membrane potential was −4.55 ± 0.47 mV, which was significantly lower than that in the ACSF group (0.54 ± 0.17 mV; p < 0.05, n = 6), but it was not significantly different from that with ZD7288 alone (p > 0.05, n = 6, Figure 7C). These results indicated that IH blockade abolished the GLP1-induced depolarization of membrane potential, and that IH underlies GLP1-induced depolarization in mouse cerebellar PCs.

We further employed the GLP1 receptor blocker, exendin-9-39 (100 nM), to examine whether GLP1 enhanced IH channel activity via its receptor. Administration of exendin-9-39 did not significantly affect I-instant, I-steady, IH, or I-tail induced by hyperpolarization potential (−130 mV, 1 s). In the presence of exendin-9-39, the mean I-instant was 152.9 ± 13.9 pA (p > 0.05 versus ACSF group, n = 7, Figures 8A,B), the mean I-steady was 231.6 ± 14.9 pA (p > 0.05 versus ACSF group, n = 7, Figure 8C), the mean IH was 78.7 ± 3.1 (p > 0.05 versus ACSF group, n = 7, Figure 8D), and the mean I-tail was 35.3 ± 2.6 pA (p > 0.05 versus ACSF, n = 7, Figure 8E). Moreover, application of exendin-9-39 completely blocked the effect of GLP1 on the hyperpolarization-elicited membrane current. In the presence of exendin-9-39 and GLP1, the mean I-instant was 152.4 ± 14.4 pA (p > 0.05 versus exendin-9-39 alone, n = 7, Figure 8B), the mean I-steady was 231.6 ± 14.9 pA (p > 0.05 versus exendin-9-39 alone, n = 7, Figure 8C), the mean IH was 81.6 ± 3.8 pA (ACSF) (p > 0.05 versus exendin-9-39 alone, n = 7, Figure 8D), and the mean I-tail was 35.4 ± 2.6 pA (p > 0.05 versus exendin-9-39 alone, n = 7, Figure 8E). These results indicate that GLP1 enhances the activity of IH channels and increases the excitability of PCs by activating GLP1 receptors.

Activation of G protein-coupled receptors can regulate the functional activities of IH channels through cAMP, cGMP, and/or Ca²⁺ signaling pathways (Pape, 1996; Ludwig et al., 1998; Biel et al., 2002). We therefore used a selective PKA inhibitor, KT5720 (500 nM), and a general PKA inhibitor, H89 (10 μM), to examine whether the effect of GLP1 on IH channel activity occurs through activation of PKA. To inhibit PKA, cerebellar slices were incubated for 20 min in ACSF containing KT5720 or H89 before patch-clamp recording. In the absence of PKA activity, application of GLP1 failed to increase hyperpolarization-elicited membrane currents (Figure 9A). In the presence of KT5720 and GLP1, the mean value of I-instant was 1156.9 ± 10.8 pA, which was similar to that of the control (KT5720; p > 0.05, n = 8, Figure 9B). The mean value of I-steady was 238.9 ± 11.1 pA, which was not significantly different from that of control (KT5720; p > 0.05, n = 8, Figure 9C). The mean value of IH was 82.0 ± 7.3 pA, which was not significantly different from that of the control (KT5720; p > 0.05, n = 8, Figure 9D). The mean value of I-tail was 33.9 ± 2.0 pA, which was similar to that of the control (KT5720; p > 0.05, n = 8, Figure 9E). In the presence of H89 and GLP1, the mean value of I-instant was 1164.3 ± 10.0 pA, which was similar to that of the control (H89; p > 0.05, n = 8, Figure 9B). The mean value of I-steady was 243.2 ± 7.1 pA, which was not significantly different than that of the control (H89; p > 0.05, n = 8, Figure 9C). The mean value of IH was 81.5 ± 6.5 pA, which was not significantly different from that of the control (H89; p > 0.05, n = 8, Figure 9D). The mean value of IT was 33.7 ± 2.8 pA, which was similar to that of the control (H89; p > 0.05, n = 8, Figure 9E). These results indicate that GLP1 enhanced the activity of IH channels through activating PKA, and that GLP1 facilitates IH and increased excitability of PCs through its receptor and PKA signaling.

In the cerebellum, GLP1 receptors are abundant in cerebellar PCs (Diz-Chaves et al., 2020). Our present results are consistent with those of a previous study (Diz-Chaves et al., 2020); GLP1 receptor immunoreactivity was present in the cerebellar cortex and was abundant in PCs, indicating that GLP1 signaling might play an important role in regulating PC activity (Farkas et al., 2021). We showed that after blockade of glutamate and GABAergic synaptic transmission, GLP1 and its analogs increased the excitability of PCs, indicating that GLP1 enhances the excitability of PCs through a postsynaptic signaling pathway.

Early studies demonstrated that activation of GLP1 receptor in rat hippocampal CA1 neurons regulated the spike firing activity of neurons through synaptic transmission mediated by non-NMDA glutamate receptors (Oka et al., 1999). GLP1 inhibits or excites neurons expressing its receptor in the mouse bed nucleus of the striaterminalis, 40% of which show excitation, and 60% of which show inhibition (Williams et al., 2018). In addition, GLP1 induces depolarization in most hippocampal neurons and hyperpolarization in a few neurons (Cork et al., 2015). In the hypothalamic paraventricular nucleus, activation of GLP1 receptor on CRH neurons up-regulates the function of AMPA receptors through PKA signaling, and enhances excitatory glutamate synaptic transmission (Liu et al., 2017). In the arcuate nucleus, GLP1 analogs act directly on the postsynaptic receptors, leading to membrane depolarization, accompanied by enhanced spontaneous action potential activity (Secher et al., 2014; He et al., 2019). Furthermore, GLP1 induces depolarization of ganglion cell resting membrane potential through the cAMP pathway, accompanied by an increase in intracellular Ca²⁺ concentration, leading to a significant increase in spike firing frequency (Kakei et al., 2002). Moreover, GLP1 receptor activation increases the spike firing frequency of gonadotropin-releasing hormone neurons in male mice and significantly increases the frequency of miniature synaptic postsynaptic currents, indicating that GLP1 receptor activation promotes GnRH neuronal activity (Farkas et al., 2016). Consistent with previous studies (Kakei et al., 2002; Farkas et al., 2016; Liu et al., 2017), we found that GLP1 and its analogue significantly increased the spike firing frequency of PCs in the absence of inhibitory and excitatory synaptic afferents, indicating that GLP1 receptor activation enhances neuronal excitability, leading to membrane depolarization and an increased spike firing rate in cerebellar PCs.

The cerebellum plays a key role in motor regulation and coordination (Ito, 1984, 2006). PCs are the sole output cells of the cerebellar cortex and their axons project to the deep cerebellar nucleus. PCs are GABAergic neurons that exhibit highly autonomic simple spike discharge activity and inhibit the activity of deep cerebellar nucleus neurons by the release of GABA (Ito, 1984, 2002). We show abundant GLP1 receptor immunoreactivity in the PCs of the cerebellar cortex (Diz-Chaves et al., 2020; Farkas et al., 2021) and that GLP1 combines with its receptor to activate IH channels through PKA signaling. This result in PC membrane depolarization and an increase in spike firing rate. In addition, the cerebellar cortex receives preglucagonergic neuron projections (Campos et al., 1994; Cork et al., 2015), which indicates that endogenous GLP1 may be released and participate in motor regulation and coordination by increasing spike firing activity of PCs. Together, the present results provide novel insight into the function of GLP1 in the central nervous system and indicate that GLP1 modulates motor coordination and motor learning by increasing cerebellar PC output.