Neuroscience 256 (2014) 242–251

DEVELOPMENTAL a2-ADRENERGIC REGULATION OF NORADRENERGIC SYNAPTIC FACILITATION AT CEREBELLAR GABAERGIC SYNAPSES M. HIRONO, a,b* S. NAGAO a AND K. OBATA b

synaptic transmission in the developing cerebellar cortex and that a2-ARs temporally restrain the noradrenergic facilitation of sIPSCs after GABAergic synaptogenesis. Ó 2013 IBRO. Published by Elsevier Ltd. All rights reserved.

a

Laboratory for Motor Learning Control, RIKEN Brain Science Institute, Wako, Saitama 351-0198, Japan b Obata Research Unit, RIKEN Brain Science Institute, Wako, Saitama 351-0198, Japan

Key words: noradrenaline, interneuron, IPSC, cerebellar cortex, Purkinje cell, synergistic effect.

Abstract—In the central nervous system, the normal development of neuronal circuits requires adequate temporal activation of receptors for individual neurotransmitters. Previous studies have demonstrated that a2-adrenoceptor (a2-AR) activation eliminates spontaneous action potentials of interneurons in the cerebellar molecular layer (MLIs) and subsequently reduces the frequency of spontaneous inhibitory postsynaptic currents (sIPSCs) in Purkinje cells (PCs) after the second postnatal week. The magnitude of the a2-adrenergic reduction in sIPSC frequency is enhanced during the third postnatal week because of an increase in firing-derived sIPSCs. However, little is known about the effects of a2-AR activation by noradrenaline (NA) on cerebellar GABAergic synaptic transmission that is accompanied by the activation of other AR subtypes, a1- and b-ARs. Here, we developmentally examined the roles of a2-AR activation in the noradrenergic facilitation of sIPSCs in cerebellar PCs. Until the second postnatal week, when substantial inhibitory effects of a2-ARs are absent, NA potentiated sIPSCs and maintained the increased sIPSC frequency, suggesting that NA causes long-lasting facilitation of GABAergic synaptic transmission through a1- and b-AR activation. After the second postnatal week, NA transiently increased the sIPSC frequency, whereas blocking a2-ARs sustained the noradrenergic sIPSC facilitation and increase in the firing rate of MLIs, suggesting that a2-AR activation suppresses the noradrenergic facilitation of GABAergic synaptic transmission. The simultaneous activation of a1- and b-ARs by their specific agonists mimicked the persistent facilitation of sIPSC frequency, which required extracellular signal-regulated kinase 1/2 activation. These findings indicate that NA acts as a neurotrophic factor that strengthens GABAergic

INTRODUCTION Noradrenaline (NA) that is released in the cerebellar cortex serves as a neurotransmitter for modulating cerebellar motor learning under certain conditions, such as states of stress, alert, and active waking (Chandler et al., 2013). NA activates adrenoceptors (ARs), which are classified into one of three major groups, a1, a2, and b, and which modulate synaptic strength (Delaney et al., 2007; Hu et al., 2007). NA modulates cerebellumdependent learning tasks, mostly through the activation of b-ARs (see Cartford et al., 2004 for a review). b2-Adrenergic pathways enhance cerebellar GABAergic synaptic transmission onto interneurons in the molecular layer (MLIs) and Purkinje cells (PCs) (Llano and Gerschenfeld, 1993; Cheun and Yeh, 1996; Kondo and Marty, 1997, 1998; Mitoma and Konishi, 1999; Saitow et al., 2000, 2005). In contrast to b2-ARs, little is known about the physiological roles of a-ARs in the cerebellar cortex. Recent studies have revealed that the activation of a1-ARs in presynaptic MLIs facilitates firing of MLIs as well as enhances the release of GABA at the presynaptic terminals (Herold et al., 2005; Hirono and Obata, 2006). Additionally, our previous studies have shown that a2-AR activation depresses the spontaneous firing of presynaptic MLIs but does not alter the release of GABA at presynaptic terminals, resulting in a partial suppression of spontaneous inhibitory postsynaptic currents (sIPSCs) recorded from PCs of adolescent mice (Hirono and Obata, 2006). Furthermore, we have demonstrated developmental changes in the a2-adrenergic suppression of sIPSCs in PCs through a decrease in the intrinsic firing of MLIs by a reduction in conductance of hyperpolarization-activated cyclic nucleotide-gated (HCN) channels on MLIs, suggesting that the developmental increase in the a2-adrenergic inhibition of sIPSCs is attributed to an age-dependent increase in firing-derived sIPSCs (Hirono et al., 2008). Therefore, compared to the effects of a1- and b-AR activation, a2-AR activation has inhibitory effects on

*Corresponding author. Present address: Organization for Advanced Research and Education, Doshisha University, Kizugawa, Kyoto 6190225, Japan. Tel: +81-774-65-6055. E-mail address: [email protected] (M. Hirono). Abbreviations: ACSF, artificial cerebrospinal fluid; AR, adrenoceptor; BDNF, brain-derived neurotrophic factor; cAMP, cyclic adenosine monophosphate; EGTA, ethylene glycol tetraacetic acid; ERK, extracellular signal-regulated kinase; HCN, hyperpolarizationactivated cyclic nucleotide-gated; HEPES, hydroxyethyl piperazineethanesulfonic acid; ISP, isoproterenol; MEK, mitogen extracellular regulating kinase; mIPSC, miniature inhibitory postsynaptic current; MLI, interneuron in the molecular layer; NA, noradrenaline; PC, Purkinje cell; PE, phenylephrine; PKA, protein kinase A; PKC, protein kinase C; sIPSC, spontaneous inhibitory postsynaptic current; TrkB, tyrosine kinase B; TTX, tetrodotoxin.

0306-4522/13 $36.00 Ó 2013 IBRO. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.neuroscience.2013.10.030 242

243

M. Hirono et al. / Neuroscience 256 (2014) 242–251

GABAergic synaptic transmission at MLI-PC synapses in the cerebellar cortex. However, the physiological roles of a2-ARs in the noradrenergic effects on MLI activity have not yet been fully understood. Here we investigated the contributions of a2-AR activation on IPSC modulations that are induced by NA application to mouse cerebellar slices. We found that NA treatment induced a transient increase in the sIPSC frequency, which was followed by long-lasting facilitation in the cerebellar cortex of mice until the second postnatal week. After the second postnatal week, although NA treatment only elicited a transient increase in the sIPSC frequency, in the presence of a selective a2-AR antagonist the increase in the sIPSC frequency persisted, and the increased firing rate of MLIs continued. Furthermore, the simultaneous activation of a1- and b-ARs mimicked the long-lasting facilitation of the sIPSC frequency. These findings suggest that cerebellar a2-ARs play roles in the presynaptic modulation of GABAergic synaptic transmission from MLIs to the principal neurons, PCs.

EXPERIMENTAL PROCEDURES Slice preparation Cerebellar slices were prepared from C57BL/6 mice [postnatal days (P) 11–25] of either sex as described previously (Hirono and Obata, 2006). All experimental procedures described here were carried out in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Publications No. 80-23) revised 1996. The RIKEN Animal Research Committee approved the procedures, and all efforts were made to minimize the number of animals that were used and their suffering. The mice were deeply anesthetized with halothane and decapitated, and their cerebella were rapidly removed. Parasagittal slices (230 lm thick) of the cerebellar vermis were cut with a vibrating microtome (VT1000S, Leica, Nussloch, Germany) in an ice-cold extracellular solution containing (in mM) 252 sucrose, 3.35 KCl, 21 NaHCO3, 0.6 NaH2PO4, 9.9 glucose, 1 CaCl2, and 3 MgCl2 and gassed with a mixture of 95% O2 and 5% CO2 (pH 7.4). The slices were maintained at room temperature for at least 1 h in a holding chamber, where they were submerged in artificial cerebrospinal fluid (ACSF) containing (in mM) 138.6 NaCl, 3.35 KCl, 21 NaHCO3, 0.6 NaH2PO4, 9.9 glucose, 2 CaCl2, and 1 MgCl2 (bubbled with 95% O2 and 5% CO2 to maintain the pH at 7.4). Electrophysiological recordings and analysis Individual slices were transferred to a recording chamber attached to the stage of a microscope (BX51WI, Olympus, Tokyo, Japan) and superfused with oxygenated ACSF. PCs and MLIs were visually identified under Nomarski optics with a water immersion objective (60, NA 0.90, Olympus). To record sIPSCs as outward currents, PCs were voltage-clamped at 55 to 45 mV with patch pipettes (2–3 MX) filled with the potassium

methanesulfonate-based internal solution containing (in mM) 140 KCH3SO3, 2 KCl, 0.1 CaCl2 1.0 K-EGTA, 10.0 Na-HEPES, 3.0 Mg-ATP, and 0.4 Na-GTP (pH 7.4). To specifically isolate the inhibitory outward currents from the excitatory inward currents, a nonselective ionotropic glutamate receptor antagonist, kynurenic acid (1 mM), was added to the ACSF throughout the recordings. Spontaneous IPSCs were completely abolished by a selective GABAA receptor antagonist SR95531 (10 lM). Miniature inhibitory postsynaptic currents (mIPSCs) were recorded with a CsCl-based internal solution containing (in mM) 140 CsCl, 0.1 CaCl2, 1.0 K-EGTA, 10.0 Na-HEPES, 3.0 Mg-ATP, and 0.4 Na-GTP (pH 7.4) in the presence of tetrodotoxin (TTX, 0.5 lM) and kynurenic acid (1 mM) from PCs voltage-clamped at 70 mV. We recorded extracellular firing from interneurons that were located in the external half of the molecular layer. However, because we did not identify each interneuron as either a basket or stellate cell by using previously reported morphological and physiological criteria (Llano and Gerschenfeld, 1993; Ha¨usser and Clark, 1997), we generically refer to recorded cells as MLIs. For loose cellattached recording, the pipette was gently placed in contact with a MLI, and slight suction was applied. The pipette (2–4 MX) containing ACSF was maintained at 0 mV. The membrane currents were recorded with an amplifier, MultiClamp 700B (Molecular Devices, Foster City, CA, USA), and pCLAMP9.2 software (Molecular Devices), digitized, and stored on a computer disk for off-line analysis. All current signals were filtered at 2–4 kHz and sampled at 5–10 kHz, and sIPSCs and mIPSCs were analyzed with a threshold of 10 pA. Focal stimulation (30–50 V, 0.1–0.2 ms) was applied through a glass microelectrode containing ACSF that was placed within the molecular layer in the cerebellar cortex. Series resistance (8–14 MX) was monitored with 2-mV hyperpolarizing voltage pulse (30–50 ms) every 30 s, and experiments were discarded if the value changed more than 20%. RS79948 and SR95531 were obtained from Tocris Bioscience (Bristol, UK). TTX was obtained from Wako (Osaka, Japan). All other chemicals were from Sigma (St. Louis, MO, USA). All experiments were performed at room temperature (23–26 °C). Spontaneous, evoked and miniature IPSCs, and action potential frequencies were analyzed with the Mini Analysis Program, version 6 (Synaptosoft, Decatur, GA, USA), pCLAMP9.2 software, and Kyplot software (Kyence, Tokyo, Japan). All data are expressed as mean ± standard error of the mean (s.e.m.). The raw values of frequencies (amplitudes) between the control and the tested groups were compared with paired t-test, and, thereafter, the percentages among several groups were statistically analyzed with a nonparametric test, such as the Steel–Dwass test.

RESULTS Effects of NA on sIPSCs at P18–25 In this study, we recorded sIPSCs as outward current responses in cerebellar PCs, which were

244

M. Hirono et al. / Neuroscience 256 (2014) 242–251

voltage-clamped at 55 to 45 mV. The mean frequency and amplitude of sIPSCs were 11.7 ± 1.7 Hz and 21.6 ± 2.0 pA (n = 9), respectively. We first investigated the effects of the bath application of NA (10 lM) for 3 min on sIPSCs. NA increased the sIPSC frequency significantly to 16.4 ± 1.8 Hz (145 ± 6% of control; P < 0.05; n = 9; Fig. 1). In addition, NA increased the amplitude of sIPSCs substantially to 30.8 ± 3.9 pA (142 ± 9% of control; P < 0.01; n = 9; by Kolmogorov–Smirnov test; Fig. 1D). The NA-induced facilitation of the sIPSC frequency and the increase in sIPSC amplitude subsided over 15 min after removing the agonist (Fig. 1A–C). These results are consistent with previous studies that have found that NA facilitates inhibitory synaptic transmission through the AR activation of presynaptic GABAergic interneurons (Llano and Gerschenfeld, 1993; Mitoma and Konishi, 1996, 1999; Kondo and Marty, 1998; Saitow et al., 2000; Herold et al., 2005; Hirono and Obata, 2006).

a2-AR’s suppression of NA-mediated facilitation of sIPSCs Our previous studies have suggested that NA application activates a2-ARs on MLIs and could suppress the NA-mediated enhancement of sIPSCs in PCs (Hirono and Obata, 2006; Hirono et al., 2008). To examine this possibility, we perfused the selective a2-AR antagonist RS79948 (Uhle´n et al., 1998; Clarke and Harris, 2002) onto cerebellar slices that were obtained from mice at P18–25. The pretreatment of cerebellar slices with RS79948 (3 lM) for 10–15 min did not alter the sIPSC frequency (98 ± 2% of control, P = 0.780, n = 10). Furthermore, a 45-min application (including a 10-min pretreatment) of RS79948 had no effect on the sIPSC

frequency (97 ± 4% of control, P = 0.938, n = 4; Fig. 2A), suggesting that the basal sIPSC frequency is not affected by the background activity of a2-ARs in the slice preparation. After 10–15 min of treatment with RS79948, exogenously applied NA (10 lM) increased the frequency of sIPSCs robustly from 9.2 ± 0.9 to 16.2 ± 0.8 Hz (180 ± 14% of control; P < 0.001; n = 5; Fig. 2A, B). The facilitation of the sIPSC frequency was maintained 30 min after the onset of NA application in the presence (131 ± 7% of control; n = 5), but not in the absence (96 ± 2% of control; n = 9) of RS79948 (P < 0.01, by Steel–Dwass test; Fig. 2A, B). The mean amplitude of sIPSCs were increased by NA in the presence of RS79948 from 19.1 ± 0.7 to 27.5 ± 1.0 pA (145 ± 8% of control; P < 0.01; n = 5), while the amplitude enhancement subsided within 30 min after the onset of NA application (102 ± 5% of control; P = 0.719; n = 5) (Fig. 2C). To further confirm these effects of RS79948, we applied another selective a2-AR antagonist, RX821002 (Clarke and Harris, 2002; Happe et al., 2004) to cerebellar slices and examined the effects of NA on sIPSCs. The treatment of cerebellar slices with RX821002 (1 lM) for 45 min did not alter the sIPSC frequency (101 ± 14% of control; n = 5). In the presence of RX821002, NA (10 lM) increased the sIPSC frequency from 10.2 ± 1.2 to 18.1 ± 0.6 Hz (183 ± 28% of control; P < 0.001; n = 6). The sIPSC frequency remained enhanced 30 min after the onset of NA application (138 ± 12% of control; P < 0.05; n = 6) (P < 0.05 compared to the control value by Steel– Dwass test; Fig. 2B). NA also increased the mean amplitude of sIPSCs from 20.6 ± 1.0 to 28.4 ± 0.9 pA (139 ± 5% of control; P < 0.001; n = 6), while the enhancement subsided 30 min after the onset of NA

Fig. 1. NA transiently increases the frequency of sIPSCs in PCs. (A) Typical current recording showing an increase in the sIPSC frequency by NA (10 lM). The PC was voltage-clamped at 50 mV. (B) Detailed sIPSCs recorded before (a), during (b), and 30 min after (c) NA application. (C) Time course of the sIPSC frequency (n = 9). Frequency of sIPSC was measured every 30 s. (D) Cumulative probability fraction of the sIPSC amplitude obtained from the same cell as in (A). Amplitude distribution significantly shifted to the right by NA (⁄⁄⁄P < 0.001, by Kolmogorov–Smirnov test).

M. Hirono et al. / Neuroscience 256 (2014) 242–251

245

Fig. 2. a2-AR antagonists maintain NA-mediated facilitation of sIPSCs in PCs. (A) Time courses of the sIPSC frequency. NA (10 lM) was applied for 3 min in the presence (red circles) and absence (open circles, the same as Fig. 1C) of the a2-AR antagonist, RS79948 (3 lM). In the presence of RS79948, the NA-mediated increase in the sIPSC frequency did not subside. Long-term treatment with RS79948 did not change the basal frequency of sIPSCs (gray circles). (B) Mean effect of NA on the sIPSC frequency under the control condition (n = 9) and in the presence of 3 lM RS79948 (RS, n = 5) or 1 lM RX821002 (RX, n = 6) during and 30 min after NA application. The sIPSC frequency was still increased in the presence of the antagonists 30 min after the onset of NA application (⁄P < 0.05, ⁄⁄P < 0.01, by Steel–Dwass test). (C) Mean effect of NA on the sIPSC amplitude under the control condition (n = 9) and in the presence of 3 lM RS79948 (RS, n = 5) or 1 lM RX821002 (RX, n = 6). (D) Time course of the evoked IPSC amplitude (n = 4). The amplitude subsided after washout of NA. (E) Average responses of four consecutively evoked IPSCs that were recorded at the time points indicated by a–c in (D). (F) Mean effect of NA on the evoked IPSC amplitude in the presence of RS79948 during and 30 min after NA application (⁄⁄P < 0.01, n = 4). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

application (100 ± 7% of control; P = 1.00; n = 6) (Fig. 2C). Taken together, these results indicate that the activation of a2-ARs through NA reduces the noradrenergic facilitation of sIPSCs potentially mediated by a1- and b-AR activation. In the presence of RS79948, we examined the effects of NA on evoked IPSCs in PCs. NA (10 lM) potentiated the amplitude of evoked IPSCs from 89 ± 4 to 139 ± 8 pA (157 ± 6% of control; P < 0.01; n = 4; Fig. 2D F) in association with an increase in the sIPSC frequency (Fig. 2A). However, the evoked IPSC amplitude recovered to the basal level after washout of NA (93 ± 7% of control; P = 0.336; n = 4; Fig. 2D F), while the sIPSC frequency remained relatively enhanced. No effect of a2-AR activation on NA-mediated facilitation of mIPSCs We examined whether blocking a2-ARs altered the NA-mediated facilitation of mIPSCs recorded from PCs. To record mIPSCs from PCs as inward current responses, we used the CsCl-based internal solution and voltageclamped PCs at 70 to 65 mV. In the presence of TTX (0.5 lM) and kynurenic acid (1 mM), the perfusion of NA (10 lM) increased the mIPSC frequency from 7.3 ± 2.1 to 12.2 ± 2.9 Hz (186 ± 18% of control;

P < 0.01; n = 6). The frequency of mIPSCs returned to the basal level within 30 min after the onset of NA application (95 ± 3% of control; P = 0.434; n = 6) (Fig. 3A, C). On the other hand, the mean amplitude of mIPSCs was not affected by NA application (from 72 ± 20 to 71 ± 17 pA; 104 ± 10% of control; P = 0.908; n = 6; Fig. 3D), as has been reported previously (Llano and Gerschenfeld, 1993). The pretreatment of cerebellar slices with the selective a2-AR antagonist RS79948 (3 lM) for 15 min did not alter the mIPSC frequency (102 ± 3% of control, P = 0.803, n = 6), and the subsequent application of NA increased the mIPSC frequency from 7.5 ± 1.0 to 13.4 ± 1.3 Hz (183 ± 12% of control; P < 0.001; n = 6; Fig. 3B, C). The mean amplitude of mIPSCs was not affected by NA (from 56 ± 9 to 62 ± 9 pA; 113 ± 4% of control; P = 0.130; n = 6; Fig. 3D). The extent of the increase in the mIPSC frequency was the same in the absence and presence of RS79948 (186 ± 18 vs. 183 ± 12%; P = 0.241, by Steel–Dwass test; Fig. 3C). Even in the presence of RS79948, the effect of NA on the mIPSC frequency reversed after the removal of NA (103 ± 7% of control; P = 0.794; n = 6; Fig. 3A C). These results suggest that a2-AR activation does not control the NA-mediated facilitation of GABA release at presynaptic terminals.

246

M. Hirono et al. / Neuroscience 256 (2014) 242–251

Fig. 3. No effect of a2-AR antagonist RS79948 on the NA-mediated facilitation of mIPSCs recorded from PCs. (A, B) Detailed current records for mIPSCs before, during, and 30 min after NA (10 lM) application in the absence (A) and presence of 3 lM RS79948 (B). (C) Time courses of the frequency of mIPSCs recorded from PCs in the absence (n = 6) and presence of RS79948 (n = 6). (D) No effect of NA was observed on the mIPSC amplitude in the absence (n = 6) or in the presence of RS79948 (n = 6) either during or 30 min after NA application.

Blocking a2-ARs induces NA-mediated long-lasting facilitation of MLI firing We examined the effects of NA on the spontaneous firing of MLIs with loose cell-attached recordings. Under control conditions, the application of NA (10 lM) increased the firing rate of MLIs transiently from 7.2 ± 1.3 to 11.5 ± 1.5 Hz (172 ± 14% of control; P < 0.001; n = 8). The facilitation subsided 30 min after the onset of NA application (105 ± 3% of control; P = 0.495; n = 8) (Fig. 4A, C, D). In the presence of RS79948 (3 lM), NA increased the firing rate from 9.3 ± 1.6 to 14.8 ± 1.7 Hz (180 ± 20% of control; P < 0.001; n = 8). In contrast to the NA-mediated short facilitation of MLI firing under control conditions, the increased firing rate continued 30 min after the onset of NA application (135 ± 8% of control; P < 0.01; n = 8) (P < 0.01 compared to the control value by Steel– Dwass test; Fig. 4B D). Therefore the long-lasting facilitation of sIPSCs induced by NA in the presence of the selective a2-AR antagonist was attributed to the noradrenergic increase in the firing rate of MLIs under blocking a2-ARs.

Synergistic effects of a1- and b-AR activation on sIPSCs To examine whether the simultaneous activation of

a1- and b-ARs induces a larger and longer increase in the sIPSC frequency, we applied selective agonists for

a1- and b-ARs at lower concentrations and then examined the effects of both agonists. A selective a1-AR agonist, phenylephrine (PE) (3 lM, 2 min), caused a small increase in the sIPSC frequency from 14.9 ± 1.3 to

18.0 ± 0.9 Hz (123 ± 5% of control; P < 0.01; n = 6; Fig. 5A, B). A selective b-AR agonist, isoproterenol (ISP) (10 nM, 2 min), also caused a small increase in the sIPSC frequency from 16.3 ± 1.2 to 19.6 ± 5 Hz (121 ± 6% of control; P < 0.05; n = 4; Fig. 5A, B). The simultaneous application of PE and ISP facilitated the sIPSC frequency remarkably from 10.7 ± 1.5 to 17.9 ± 1.4 Hz (182 ± 20% of control; P < 0.001; n = 7; Fig. 5A, B). The percentage of the increase in the sIPSC frequency by the application of both PE and ISP was significantly larger than the value that was obtained by the application of each agonist (P < 0.05, by Steel–Dwass test; Fig. 5B) and larger than the value estimated by the additive effects of the agonists, suggesting that the activation of both a1- and b-ARs evokes synergistic effects on sIPSCs in PCs. Furthermore, the simultaneous application of PE and ISP maintained the increase in the sIPSC frequency 30 min after the onset of the application (114 ± 5% of control; P < 0.01; n = 7; Fig. 5A, C). Because intracellular signaling following a1- and b-AR activation can activate extracellular signal-regulated kinase 1/2 (ERK1/2) (Yuan et al., 2002; Sweatt, 2004), we examined whether ERK1/2 was involved in the synergistic effects. We pretreated the cerebellar slices with a mitogen extracellular regulating kinase (MEK) inhibitor, PD98059 (10 lM, 15 min), and applied PE and ISP simultaneously. Even in the presence of PD98059, the increase in the sIPSC frequency was observed during the perfusion of both PE and ISP (145 ± 9% of control; P < 0.001; n = 9; Fig. 5A, B). However, PD98059 significantly blocked the long-lasting facilitation of the sIPSC frequency (P < 0.05, by Steel– Dwass test; Fig. 5A, C). When we applied only ISP at a

247

M. Hirono et al. / Neuroscience 256 (2014) 242–251

Fig. 4. a2-AR antagonist RS79948 induces NA-mediated long-lasting facilitation of spontaneous action potentials of MLIs. Action potentials were recorded by loose cell-attached recordings from MLIs. (A, B) Spontaneous firing recorded before, during, and 30 min after NA (10 lM) application in the absence (A) and the presence of 3 lM RS79948 (B). (C) Time courses of the firing rate of MLIs in the absence (n = 8) and presence of RS79948 (n = 8). (D) NA caused the long-lasting increase in the firing rate of MLIs in the presence of RS79948 (⁄⁄P < 0.01, by Steel–Dwass test, n = 8).

higher concentration (20 lM) for 2 min, the sIPSC frequency increased extremely (163 ± 12% of control; P < 0.01; n = 5). The facilitation subsided within 30 min after the onset of the agonist application (98 ± 11% of control; P = 0.325; n = 5). In the presence of PD98059 (10 lM), ISP (20 lM) still transiently increased the sIPSC frequency (135 ± 8% of control; P < 0.01; n = 4). There was no significant difference in the extent of the transient facilitation with or without the inhibitor (P = 0.086, by Steel–Dwass test). These results suggest that the simultaneous activation of a1- and b-ARs could induce the long-lasting facilitation of the sIPSC frequency through ERK1/2 activation. Effects of NA on sIPSCs at P11–14 Our previous study demonstrated that the activation of

a2-ARs in MLIs decreases the frequency of sIPSCs recorded from PCs after the second postnatal week but not younger than that (Hirono et al., 2008). We hypothesized that, at P11–14, NA could induce the longlasting increase in the sIPSC frequency through a1- and b-AR activation. Therefore, we examined the effects of NA on sIPSCs at P11–14. NA (10 lM, 3 min) increased the sIPSC frequency from 10.6 ± 1.2 to 16.4 ± 1.1 Hz (160 ± 10% of control; P < 0.001; n = 7), and the facilitation of the sIPSC frequency continued 30 min after the onset of NA (121 ± 3% of control; P < 0.001; n = 7) (Fig. 6A, B). At P11–14, NA changed the mean amplitude of sIPSCs from 28.0 ± 2.5 to 40.2 ± 4.2 pA (147 ± 17% of control; P < 0.05; n = 7), while the enhancement of current amplitude recovered to the

basal level 30 min after NA application (96 ± 8% of control; P = 0.626; n = 7) (Fig. 6C). To examine whether the noradrenergic long-lasting increase in the sIPSC frequency at P11–14 was elicited by the simultaneous activation of a1- and b-ARs in MLIs, we applied the b-AR antagonist, propranolol, or the a1-AR antagonist, corynanthine. In the presence of propranolol (10 lM), NA caused a small increase in the sIPSC frequency transiently from 8.5 ± 1.2 to 10.7 ± 1.6 Hz (125 ± 6% of control; P < 0.05; n = 7), and the NA-mediated facilitation of sIPSCs subsided (Fig. 6A, B). The mean amplitude of sIPSCs was not altered by NA (from 22.3 ± 2.5 to 21.8 ± 2.1 pA; 99 ± 5% of control; P = 0.666; n = 7; Fig. 6C). In the presence of corynanthine (10 lM), NA increased the sIPSC frequency transiently from 11.2 ± 1.6 to 15.7 ± 1.9 Hz (141 ± 8% of control; P < 0.01; n = 4), and the NA-mediated facilitation of sIPSCs returned to the basal level (Fig. 6A, B). The mean amplitude of sIPSCs was increased by NA from 27.7 ± 2.7 to 42.5 ± 7.0 pA (150 ± 10% of control; P < 0.05; n = 4), while the mean amplitude recovered to the basal level 30 min after the onset of NA application (111 ± 6% of control; P = 0.102; n = 4) (Fig. 6C). These results suggest that NA signals induce the long-lasting facilitation of GABAergic synaptic transmission at P11–14 through the simultaneous activation of a1- and b-ARs on MLIs.

DISCUSSION Our findings demonstrate that a2-AR activation suppresses the NA-mediated long-lasting facilitation of

248

M. Hirono et al. / Neuroscience 256 (2014) 242–251

by attenuating presynaptic spike-generation at somatodendritic domains or the axon hillock of MLIs and/or attenuating spike-invasion into fine branches of the axon. The a2-AR is most commonly linked to Gi/o proteins, the activation of which inhibits adenylyl cyclase and reduces intracellular cyclic adenosine monophosphate (cAMP) concentration (Bylund, 1995). The reduction of intracellular cAMP concentration decreases conductance of HCN channels (Carr et al., 2007; Barth et al., 2008; Hirono et al., 2008) and subsequently eliminates spontaneous firing of MLIs (Hirono et al., 2008). This could be one of mechanisms underlying the shift of long-term to short-term sIPSC facilitation that is triggered by the synergistic effects of a1- and b-AR activation in presynaptic GABAergic neurons. Synergistic sIPSC facilitation by a1- and b-AR activation

Fig. 5. Simultaneous activation of a1- and b-ARs induces a longlasting increase in the sIPSC frequency. (A) Time courses of the sIPSC frequency. A 2-min treatment of the a1-AR agonist phenylephrine (PE, 3 lM, n = 6) or the b-AR agonist isoproterenol (ISP, 10 nM, n = 4) transiently increased the sIPSC frequency. Simultaneous activations of a1- and b-ARs for 2 min caused a long-lasting facilitation of the sIPSC frequency (n = 7). A 15-min pretreatment with the MEK inhibitor PD98059 (PD, 10 lM, n = 9) inhibited the long-lasting increase in sIPSCs that was induced by the application of both PE and ISP. (B) The percentage of the increase in the sIPSC frequency that was induced during a 2-min application of both PE and ISP was significantly larger than that obtained by the application of only PE or ISP (⁄P < 0.05, by Steel–Dwass test). (C) Application of both PE and ISP sustained the increased sIPSC frequency 30 min after the onset of the agonists. PD98059 (PD, 10 lM, n = 9) blocked the long-lasting effects of PE and ISP on the sIPSC frequency (⁄P < 0.05, by Steel–Dwass test).

GABAergic transmission at synapses between MLIs and PCs after the second postnatal week. NA facilitates sIPSCs in PCs through the presynaptic activation of both a1- and b-ARs, and their simultaneous activation synergistically causes the long-lasting facilitation of sIPSCs. This effect may contribute to the maturation of GABAergic synaptic connections between MLIs and PCs. After maturation of GABAergic transmission, presynaptic a2-AR activation attenuates the long-lasting facilitation of sIPSCs that is induced by the simultaneous activation of both a1- and b-ARs, resulting in the noradrenergic temporal facilitation of inhibitory transmission onto PCs. Presynaptic suppression of IPSCs by a2-AR activation Our previous studies have shown that the a 2-AR agonist clonidine does not elicit any change in the paired-pulse ratio of evoked IPSCs, the mIPSC frequency, or the postsynaptic GABA receptor sensitivity (Hirono and Obata, 2006; Hirono et al., 2008). These studies demonstrate that the activation of presynaptic a2-ARs of MLIs suppresses GABAergic transmission

The firing pattern of PCs is regulated by inhibitory transmission between MLIs and PCs (Ha¨usser and Clark, 1997; Mittmann et al., 2005; Wulff et al., 2009), which can be facilitated preferentially by presynaptic b-AR activation when NA diffuses into the cerebellar cortex (Llano and Gerschenfeld, 1993; Cheun and Yeh, 1996; Kondo and Marty, 1997, 1998; Mitoma and Konishi, 1999; Saitow et al., 2000). The b-AR agonist ISP increases the mIPSC frequency (Mitoma and Konishi, 1999), and the ISP-mediated enhancement of evoked IPSCs is blocked by the selective b2-AR antagonist ICI118,551 (Saitow et al., 2000), indicating that b2-AR activation contributes to the ISP-mediated enhancement of IPSCs. Additionally, NA activates presynaptic a1-ARs, and facilitates inhibitory synaptic transmission onto PCs through an increase in the release probability of GABA at the nerve terminals of MLIs (Crepel et al., 1987; Parfitt et al., 1988; Herold et al., 2005; Hirono and Obata, 2006). The presynaptic a1-AR activation induces not only an increase in [Ca2+]i (Kirischuk et al., 1996a,b; Kulik et al., 1999), but also protein kinase C (PKC) activation (Kirischuk et al., 1996a; Docherty, 1998), leading to phosphorylation of presynaptic proteins that are involved in vesicle fusion and modulates synaptic transmitter release (Leenders and Sheng, 2005). Therefore, as we observed here, simultaneous activation of a1- and b2-ARs that are located at presynaptic terminals of MLIs can synergistically boost GABAergic transmitter release. Our previous study has indicated that a1-AR activation increases the firing rate of MLIs (Hirono and Obata, 2006), suggesting that a1-ARs are expressed not only in presynaptic terminals, but also in somatodendritic sites of MLIs. Thus NA could induce cross-talk between protein kinase A (PKA) and PKC in somatodendritic sites of MLIs through the simultaneous activation of a1- and b2-ARs. In sensory neurons of Aplysia, 5-HT facilitates synaptic transmission by the presynaptic simultaneous activation of PKA and PKC (Byrne and Kandel, 1996). MLIs express A-type K+ channels in their somatodendritic sites (Kollo et al., 2006). The A-type K+ channels are phosphorylated by ERK1/2, and

M. Hirono et al. / Neuroscience 256 (2014) 242–251

249

Fig. 6. Effects of NA on sIPSCs recorded from PCs of mice at P11–14. (A) Time courses of the frequency of sIPSCs. Mice at P11–14 exhibited a long-lasting increase in the sIPSC frequency after a treatment of 10 lM NA for 3 min (n = 7). In the presence of the b-AR antagonist propranolol (10 lM, n = 7) or the a1-AR antagonist corynanthine (10 lM, n = 4), the NA-induced facilitation of sIPSCs subsided after washout of NA. (B) Mean effect of NA on the sIPSC frequency recorded under the control condition (n = 7) and in the presence of propranolol (10 lM; ⁄P < 0.05; by Steel– Dwass test; n = 7) or corynanthine (10 lM, n = 4) from mice at P11–14. The NA-mediated long-lasting facilitation of sIPSCs was blocked by propranolol or corynanthine (⁄P < 0.05; by Steel–Dwass test). (C) Mean effect of NA on the amplitude of sIPSCs recorded from mice at P11–14. Propranolol blocked the NA-mediated transient increase in the sIPSC amplitude, while corynanthine did not (⁄P < 0.05; by Steel–Dwass test).

the voltage-dependence of the activation is shifted in a depolarizing direction, suggesting that the firing rate of MLIs could be increased. Because both of PKA and PKC can activate ERK1/2 (Yuan et al., 2002; Sweatt, 2004), the simultaneous activation of a1- and b2-ARs facilitates ERK1/2 activity synergistically and enhances the excitability of MLIs. Alternatively, large-conductance potassium (BK) channels might be inhibited by ion channel palmitoylation and phosphorylation induced by PKA and PKC activation (Zhou et al., 2012). In the present study, we demonstrate that the activation of a1- and b2-ARs synergistically and extremely facilitates sIPSCs in PCs and causes long-lasting facilitation; whereas each activation of a1- or b2-ARs causes only a transient and small increase in sIPSCs. We also show that the synergistic effects of a1- or b2-AR activation require ERK1/2 activation in MLIs. Therefore, the simultaneous activation of a1- and b2-ARs can elicit cross-talk of downstreams of PKA and PKC activation, resulting in the long-lasting higher excitability of MLIs. Functional implication of NA in the cerebellar cortical circuits NA serves as a neurotrophic factor in normal cortical development because the deletion of NA at birth results in modified cortical development with the formation of abnormal numbers of synapses (Felten et al., 1982; Blue and Molliver, 1987; Winzer-Serhan and Leslie, 1999). Additionally, noradrenergic excitatory effects on neurons and astrocytes can increase the expression of

trophic factors such as brain-derived neurotrophic factor (BDNF) (Juric et al., 2008; Francis et al., 2012). In the cerebellar cortex, tyrosine kinase B (TrkB) signaling is required for the establishment of GABAergic synapses (Seil and Drake-Baumann, 2000; Rico et al., 2002), and the activation of TrkB receptors by BDNF contributes to an increase in the amplitude of GABAergic miniature IPSCs in PCs (Boxall, 2000; Drake-Baumann, 2005). However, it remains unknown how BDNF regulates MLI firing. The artificial activation and inhibition of a2-ARs during development affects the normal formation of neuronal circuits, synaptic connectivity, and neuronal responses (Kreider et al., 2004; Mondaca et al., 2004; Sanders et al., 2005). In the cerebellar neuronal network and in cerebellar functions, a2-ARs play a role in information processing because a2A-knockout mice exhibit impaired motor coordination skills (La¨hdesma¨ki et al., 2002). After neurodevelopment, one plausible role of a2-ARs in MLIs is to prevent the overexcitation that is induced by a1- and b2-AR activation when NA diffuses into the cerebellar cortex. a2-AR activation could cause a temporal limitation in the excitatory effects of NA on MLIs. This study suggests that a balance of activation in all of the AR subtypes expressed in presynaptic MLIs plays an important role in regulating the patterns of PC firing, resulting in proper cerebellar signaling processing. NA plays a crucial role in increasing the signal-tonoise ratio as it pertains to attention (Berridge and Waterhouse, 2003; Hurley et al., 2004). In the principal neurons of the dorsal cochlear nucleus, NA

250

M. Hirono et al. / Neuroscience 256 (2014) 242–251

simultaneously reduces spontaneous inhibitory inputs and enhances evoked inhibitory currents through a2-AR activation, suggesting the noradrenergic enhancement of feed-forward inhibition (Kuo and Trussell, 2011). In the cerebellar cortex, MLIs form feed-forward inhibition onto PCs, which regulates the precise timing of PC firing triggered by parallel fiber inputs (Mittmann et al., 2005). Therefore, after the maturation of the neuronal circuits in the cerebellar cortex, a2-ARs that are activated profoundly by NA may enhance feed-forward inhibition and facilitate motor coordination and learning during arousal, attention, and excitation.

Acknowledgements—We thank Drs. N. Suzuki and F. Saitow for their invaluable comments and critical reading of this manuscript. This work was supported by the Special Postdoctoral Researchers Program (M.H.) from RIKEN.

REFERENCES Barth AM, Vizi ES, Zelles T, Lendvai B (2008) a2-Adrenergic receptors modify dendritic spike generation via HCN channels in the prefrontal cortex. J Neurophysiol 99:394–401. Berridge CW, Waterhouse BD (2003) The locus coeruleus– noradrenergic system: modulation of behavioral state and statedependent cognitive processes. Brain Res Brain Res Rev 42:33–84. Blue ME, Molliver ME (1987) 6-Hydroxydopamine induces serotonergic axon sprouting in cerebral cortex of newborn rat. Brain Res 429:255–269. Boxall AR (2000) GABAergic mIPSCs in rat cerebellar Purkinje cells are modulated by TrkB and mGluR1-mediated stimulation of Src. J Physiol 524:677–684. Bylund DB (1995) Pharmacological characteristics of a2 adrenergic receptor subtypes. Ann NY Acad Sci 763:1–7. Byrne JH, Kandel ER (1996) Presynaptic facilitation revisited: state and time dependence. J Neurosci 16:425–435. Carr DB, Andrews GD, Glen WB, Lavin A (2007) a2-Noradrenergic receptors activation enhances excitability and synaptic integration in rat prefrontal cortex pyramidal neurons via inhibition of HCN currents. J Physiol 584:437–450. Cartford MC, Gould T, Bickford PC (2004) A central role for norepinephrine in the modulation of cerebellar learning tasks. Behav Cogn Neurosci Rev 3:131–138. Chandler DJ, Nicholson SE, Zitnik G, Waterhouse BD (2013) Norepinephrine and synaptic transmission in the cerebellum. In: Manto M, Gruol DL, Schmahmann JD, Koibuchi N, Rossi F, editors. Handbook of the cerebellum and cerebellar disorders. New York: Springer. p. 895–914. Cheun JE, Yeh HH (1996) Noradrenergic potentiation of cerebellar Purkinje cell responses to GABA: cyclic AMP as intracellular intermediary. Neuroscience 74:835–844. Clarke RW, Harris J (2002) RX821002 as a tool for physiological investigation of a2-adrenoceptors. CNS Drug Rev 8:177–192. Crepel F, Debono M, Flores R (1987) a-Adrenergic inhibition of rat cerebellar Purkinje cells in vitro: a voltage-clamp study. J Physiol 383:487–498. Delaney AJ, Crane JW, Sah P (2007) Noradrenaline modulates transmission at a central synapse by a presynaptic mechanism. Neuron 56:880–892. Docherty JR (1998) Subtypes of functional a1- and a2-adrenoceptors. Eur J Pharmacol 361:1–15. Drake-Baumann R (2005) Rapid modulation of inhibitory synaptic currents in cerebellar Purkinje cells by BDNF. Synapse 57:183–190.

Felten DL, Hallman H, Jonsson G (1982) Evidence for a neurotropic role of noradrenaline neurons in the postnatal development of the rat cerebellar cortex. J Neurocytol 11:119–135. Francis BM, Yang J, Hajderi E, Brown ME, Michalski B, McLaurin J, Fahnestock M, Mount HTJ (2012) Reduced tissue levels of noradrenaline are associated with behavioral phenotypes of the TgCRND8 mouse model of Alzheimer’s disease. Neuropsychopharmacology 37:1934–1944. Happe HK, Coulter CL, Gerety ME, Sanders JD, O’Rourke M, Bylund DB, Murrin LC (2004) Alpha-2 adrenergic receptor development in rat CNS: an autoradiographic study. Neuroscience 123: 167–178. Ha¨usser M, Clark BA (1997) Tonic synaptic inhibition modulates neuronal output pattern and spatiotemporal synaptic integration. Neuron 19:665–678. Herold S, Hecker C, Deitmer JW, Brockhaus J (2005) a1-Adrenergic modulation of synaptic input to Purkinje neurons in rat cerebellar brain slices. J Neurosci Res 82:571–579. Hirono M, Obata K (2006) a-Adrenoceptive dual modulation of inhibitory GABAergic inputs to Purkinje cells in the mouse cerebellum. J Neurophysiol 95:700–708. Hirono M, Matsunaga W, Chimura T, Obata K (2008) Developmental enhancement of a2-adrenoceptor-mediated suppression of inhibitory synaptic transmission onto mouse cerebellar Purkinje cells. Neuroscience 156:143–154. Hu H, Real E, Takamiya K, Kang MG, Ledoux J, Huganir RL, Malinow R (2007) Emotion enhances learning via norepinephrine regulation of AMPA-receptor trafficking. Cell 131:160–173. Hurley LM, Devilbiss DM, Waterhouse BD (2004) A matter of focus: monoaminergic modulation of stimulus coding in mammalian sensory networks. Curr Opin Neurobiol 14:488–495. Juric DM, Loncar D, Carman-Krzan M (2008) Noradrenergic stimulation of BDNF synthesis in astrocytes: mediation via a1and b1/b2-adrenergic receptors. Neurochem Int 52:297–306. Kirischuk S, Matiash V, Kulik A, Voitenko N, Kostyuk P, Verkhratsky A (1996a) Activation of P2-purino-, a1-adreno and H1-histamine receptors triggers cytoplasmic calcium signaling in cerebellar Purkinje neuron. Neuroscience 73:643–647. Kirischuk S, Tuschik S, Verkhratsky A, Kettenmann H (1996b) Calcium signaling in mouse Bergmann glial cells mediated by a1adrenoceptors and H1 histamine receptors. Eur J Neurosci 8:1198–1208. Kollo M, Holderith NB, Nusser Z (2006) Novel subcellular distribution pattern of A-type K+ channels on neuronal surface. J Neurosci 26:2684–2691. Kondo S, Marty A (1997) Protein kinase A-mediated enhancement of miniature IPSC frequency by noradrenaline in rat cerebellar stellate cells. J Physiol 498:165–176. Kondo S, Marty A (1998) Differential effects of noradrenaline on evoked, spontaneous and miniature IPSCs in rat cerebellar stellate cells. J Physiol 509:233–243. Kreider ML, Seidler FJ, Cousins MM, Tate CA, Slotkin TA (2004) Transiently overexpressed a2-adrenoceptors and their control of DNA synthesis in the developing brain. Dev Brain Res 152:233–239. Kulik A, Haentzsch A, Lu¨ckermann M, Reichelt W, Ballanyi K (1999) Neuron-glia signaling via a1 adrenoceptor-mediated Ca2+ release in Bergmann glial cells in situ. J Neurosci 19:8401–8408. Kuo SP, Trussell LO (2011) Spontaneous spiking and synaptic depression underlie noradrenergic control of feed-forward inhibition. Neuron 71:306–318. La¨hdesma¨ki J, Sallinen J, MacDonald E, Kobilka BK, Fagerholm V, Scheinin M (2002) Behavioral and neurochemical characterization of a2A-adrenergic receptor knockout mice. Neuroscience 113:289–299. Leenders AG, Sheng ZH (2005) Modulation of neurotransmitter release by the second messenger-activated protein kinases: implications for presynaptic plasticity. Pharmacol Ther 105:69–84. Llano I, Gerschenfeld HM (1993) b-Adrenergic enhancement of inhibitory synaptic activity in rat cerebellar stellate and Purkinje cells. J Physiol 468:201–224.

M. Hirono et al. / Neuroscience 256 (2014) 242–251 Mitoma H, Konishi S (1996) Long-lasting facilitation of inhibitory transmission by monoaminergic and cAMP-dependent mechanism in rat cerebellar GABAergic synapses. Neurosci Lett 217:141–144. Mitoma M, Konishi S (1999) Monoaminergic long-term facilitation of GABA-mediated inhibitory transmission at cerebellar synapses. Neuroscience 88:871–883. Mittmann W, Koch U, Ha¨usser M (2005) Feed-forward inhibition shapes the spike output of cerebellar Purkinje cells. J Physiol 563:369–378. Mondaca M, Herna´ndez A, Pe´rez H, Valladares L, Sierralta W, Ferna´ndez V, Soto-Moyano R (2004) a2-Adrenoceptor modulation of long-term potentiation elicited in vivo in rat occipital cortex. Brain Res 1021:292–296. Parfitt KD, Freedman R, Bickford-Wimer PC (1988) Electrophysiological effects of locally applied noradrenergic agents at cerebellar Purkinje neurons: receptor specificity. Brain Res 462:242–251. Rico B, Xu B, Reichardt LF (2002) TrkB receptor signaling is required for establishment of GABAergic synapses in the cerebellum. Nat Neurosci 5:225–233. Saitow F, Satake S, Yamada J, Konishi S (2000) b-Adrenergic receptor-mediated presynaptic facilitation of inhibitory GABAergic transmission at cerebellar interneuron-Purkinje cell synapses. J Neurophysiol 84:2016–2025. Saitow F, Murakoshi T, Suzuki H, Konishi S (2005) Metabotropic P2Y purinoceptor-mediated presynaptic and postsynaptic enhancement of cerebellar GABAergic transmission. J Neurosci 25:2108–2116. Sanders JD, Happe HK, Murrin LC (2005) A transient expression of functional a2-adrenergic receptors in white matter of the developing brain. Synapse 57:213–222.

251

Seil FJ, Drake-Baumann R (2000) TrkB receptor ligands promote activity-dependent inhibitory synaptogenesis. J Neurosci 20:5367–5373. Sweatt JD (2004) Mitogen-activated protein kinases in synaptic plasticity and memory. Curr Opin Neurobiol 14:311–317. Uhle´n S, Dambrova M, Nasman J, Schioth HB, Gu Y, WikbergMatsson A, Wikberg JES (1998) [3H]RS79948-197 binding to human, rat, guinea pig and pig a2A-, a2B- and a2C-adrenoceptors. Comparison with MK912, RX821002, rauwolscine and yohimbine. Eur J Pharmacol 343:93–101. Winzer-Serhan UH, Leslie FM (1999) Expression of a2A adrenoceptors during rat neocortical development. J Neurobiol 38:259–269. Wulff P, Schonewille M, Renzi M, Viltono L, Sassoe-Pognetto M, Badura A, Gao Z, Hoebeek FE, van Dorp S, Wisden W, Farrant M, De Zeeuw CI (2009) Synaptic inhibition of Purkinje cells mediates consolidation of vestibulo-cerebellar motor learning. Nat Neurosci 12:1042–1049. Yuan LL, Adams JP, Swank M, Sweatt JD, Johnston D (2002) Protein kinase modulation of dendritic K+ channels in hippocampus involves a mitogen-activated protein kinase pathway. J Neurosci 22:4860–4868. Zhou X, Wulfsen I, Korth M, McClaffery H, Lukowski R, Shipston MJ, Ruth P, Dobrev D, Wieland T (2012) Palmitoylation and membrane association of the stress axis regulated insert (STREX) controls BK channel regulation by protein kinase C. J Biol Chem 287:32161–32171.

(Accepted 14 October 2013) (Available online 21 October 2013)