Address correspondence and reprint requests to Il-Sung Jang, Department of Pharmacology, School of Dentistry, Kyungpook National University, 188-1, Samduk 2 ga-dong, Jung-gu, Daegu 700-412, Republic of Korea. E-mail: firstname.lastname@example.org
The adenosinergic modulation of GABAergic spontaneous miniature inhibitory postsynaptic currents (mIPSCs) was investigated in mechanically dissociated rat tuberomammillary nucleus (TMN) neurons using a conventional whole-cell patch clamp technique. Adenosine (100 μM) reversibly decreased mIPSC frequency without affecting the current amplitude, indicating that adenosine acts presynaptically to decrease the probability of spontaneous GABA release. The adenosine action on GABAergic mIPSC frequency was completely blocked by 1 μM DPCPX, a selective A1 receptor antagonist, and mimicked by 1 μM CPA, a selective A1 receptor agonist. This suggests that presynaptic A1 receptors were responsible for the adenosine-mediated inhibition of GABAergic mIPSC frequency. CPA still decreased GABAergic mIPSC frequency even either in the presence of 200 μM Cd2+, a general voltage-dependent Ca2+ channel blocker, or in the Ca2+-free external solution. However, the inhibitory effect of CPA on GABAergic mIPSC frequency was completely occluded by 1 mM Ba2+, a G-protein coupled inwardly rectifying K+ (GIRK) channel blocker. In addition, the CPA-induced decrease in mIPSC frequency was completely occluded by either 100 μM SQ22536, an adenylyl cyclase (AC) inhibitor, or 1 μM KT5720, a specific protein kinase A (PKA) inhibitor. The results suggest that the activation of presynaptic A1 receptors decreases spontaneous GABAergic transmission onto TMN neurons via the modulation of GIRK channels as well as the AC/cAMP/PKA signal transduction pathway. This adenosine A1 receptor-mediated modulation of GABAergic transmission onto TMN neurons may play an important role in the fine modulation of the excitability of TMN histaminergic neurons as well as the regulation of sleep-wakefulness.
Histaminergic neurons within the tuberomammillary nucleus (TMN) release histamine onto various brain areas including the cortex, and have important roles in the homeostatic regulation of sleep-wakefulness (Brown et al. 2001). The TMN receives both glutamatergic and GABAergic innervation, and these synaptic activities eventually have influence on the excitability of TMN histaminergic neurons. Especially, both GABAergic and galaninergic innervation from the ventrolateral preoptic nucleus (VLPO) to the TMN plays pivotal roles in the regulation of sleep-wakefulness (Sherin et al. 1996, 1998; Steininger et al. 2001). For example, lesions of the VLPO have been known to cause insomnia (Lu et al. 2000). In addition, the electrical stimulation of the pre-optic area releases GABA onto the TMN and leads to a hyperpolarization of TMN neurons (Yang and Hatton 1997). The resultant reduction of the excitability in TMN neurons would cause sleep (Sherin et al. 1998). In addition to GABA and histamine, a variety of neurotransmitters and/or neuromodulators, such as acetylcholine, noradrenaline and serotonin, are involved in the regulation of sleep-wakefulness (Koyama and Kayama 1993).
On the other hand, adenosine has been proposed to be one of the endogenous substances that cause sleep because adenosine and its analogues, when administrated to experimental animal models, induce sleep. In addition, the extracellular adenosine concentration increases in the cortex and basal ganglia during sleep deprivation, and decreases during the recovery after sleep deprivation (Porkka-Heiskanen et al. 2000; Basheer et al. 2004). Several lines of evidence have suggested that both A1 and A2A receptors are closely related to sleep induction (for review, Basheer et al. 2004; Huang et al., 2005). Although a previous study performed by in vivo microdialysis has shown that CGS 21680, a selective A2A receptor agonist, induces sleep by increasing GABA release in the TMN, which inhibits the histaminergic arousal system (Hong et al. 2005), it is still unknown whether adenosine A1 receptors affect GABAergic transmission onto TMN histaminergic neurons. In the present study, therefore, we have investigated the adenosine-mediated presynaptic modulation of GABAergic transmission in acutely dissociated rat TMN neurons.
Materials and methods
All experiments complied with the guiding principles for the care and use of animals approved by the Council of the Physiological Society of Korea and the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Every effort was made to minimize both the number of animals used and their suffering.
Wistar rats (12–15 days old) were decapitated under pentobarbital anesthesia (50 mg/kg, i.p.). The brain was dissected and transversely sliced at a thickness of 400 μm in the incubation solution (see Solutions) saturated with 95% O2 and 5% CO2 using a microslicer (Vibratome® 1000; Warner Instruments, Hamden, CT, USA). Slices containing the tuberomammillary nucleus (TMN) were kept in the incubation solution saturated with 95% O2 and 5% CO2 at room temperature (22–24°C) for at least 1 h before the mechanical dissociation. For dissociation, slices were transferred into a 35 mm culture dish (Primaria 3801; Becton Dickinson, Rutherford, NJ, USA) containing the standard external solution (see Solutions), and the TMN region was identified under a binocular microscope (SMZ-1; Nikon, Tokyo, Japan). Details of the mechanical dissociation have been described previously (Rhee et al. 1999; Akaike and Moorhouse 2003). Briefly, mechanical dissociation was accomplished using a custom-built vibration device and a fire-polished glass pipette oscillating at about 50–60 Hz (0.3–0.5 mm) on the surface of the TMN region. Slices were removed and the mechanically dissociated neurons were allowed to settle and adhere to the bottom of the dish for 15 min. These dissociated neurons lost most distal processes but retained a short portion (∼ 50 μm in a length) of their proximal dendrites (Fig. 1a). Since this study was carried out in a reduced preparation compared with a slice preparation, it should be noted that the basal levels of adenosine and histamine around TMN neurons, if any, should be different from the intact neuronal circumstance. Therefore, the effects of these neurotransmitters on the electric activity of TMN neurons expected in intact animals may be overlooked in this study.
To determine whether the large neurons (> 20 μm in a somatic diameter) used for electrophysiological recordings in the present experiments indeed belong to TMN histaminergic neurons, immunohistochemical examinations were performed using an anti-histidine decarboxylase (HDC) antibody. Neurons were mechanically dissociated on the glass coverslips coated with polyethylenimine in a 35-mm culture dish. After most of the neurons had settled down and adhered to the coverslips, each coverslip was moved to a parafilm sheet for immunohistochemistry. Neurons were washed with phosphate-buffered saline (PBS, pH 7.4) and fixed with 4% paraformaldehyde in PBS for 30 min. After treatment with 0.2% Triton-X 100 for 5 min, neurons were incubated with PBS containing rabbit anti-HDC antibody (1 : 1200; American Research Products, Inc., Belmont, USA) in a moist chamber at 4°C overnight. The remaining staining procedures consisted of incubation with a biotinylated gout anti-rabbit secondary antibody and streptavidin (1 : 200; Santa Cruz Biotechnology, Inc., CA, USA). Images of the neurons were collected with a digital camera (Carl Zeiss, Germany). All procedures were performed at room temperature (22–24°C).
All electrical measurements were performed using conventional whole-cell patch recordings and a standard patch-clamp amplifier (Axopatch 200B; Axon Instruments; Union City, CA, USA). Neurons were voltage clamped at a holding potential (VH) of 0 mV. Patch pipettes were made from borosilicate capillary glass (1.5 mm outer diameter, 0.9 mm inner diameter; G-1.5; Narishige, Tokyo, Japan) in two stages on a pipette puller (P-97; Sutter Instrument Co., Novato, CA, USA). The resistance of the recording pipettes filled with the internal solution was 4–6 MΩ. The liquid junction potential and pipette capacitance were compensated for. Neurons were viewed under phase contrast on an inverted microscope (TE-2000; Nikon). Membrane currents were filtered at 1 kHz (E-3201A Decade Filter; NF Electronic Instruments, Tokyo, Japan), digitized at 4 kHz, and stored on a computer equipped with pCLAMP 10.02 (Axon Instruments). When recording, 10 mV hyperpolarizing step pulses (30 ms in duration) were periodically delivered to monitor the access resistance. All experiments were performed at room temperature (22–24°C).
Spontaneous mIPSCs (mIPSCs) were counted and analyzed using the MiniAnalysis program (Synaptosoft, Inc., Decatur, GA), as described previously (Jang et al. 2002). Briefly, mIPSCs were screened automatically using an amplitude threshold of 10 pA and then visually accepted or rejected based upon their rise and decay times. To quantify the effects of adenosine receptor agonists on GABAergic mIPSCs, the average values of both the frequency and amplitude of mIPSCs during the control period (5–10 min) were calculated for each recording. Then the average values of both the frequency and amplitude of all the events during the application of adenosine receptor agonists (5 min) were normalized to control values. Each external solution containing various drugs was incubated for more than 10 min before the application of adenosine receptor agonists, and the average values of both the frequency and amplitude of mIPSCs during each drug condition (10 min) were collected after the stable occurrence of mIPSCs. The inter-event intervals and amplitudes of a large number of synaptic events obtained from the same neuron were examined by constructing cumulative probability distributions and compared using the Kolmogorov-Smirnov (K-S) test with Stat View software (SAS Institute, Inc., Cary, NC, USA). The effects of these different conditions were quantified as a relative change in mIPSC frequency compared to the control values. The continuous curve of the concentration-response relationship of adenosine was fitted using a least-squares fit to the following equation:
where I is the adenosine-induced inhibition ratio of mIPSC frequency and C is the corresponding adenosine concentration. IC50 and n denote the half-inhibitory concentration and the Hill coefficient, respectively. Numerical values are provided as the mean ± standard error of the mean (SEM) using values normalized to the control. Significant differences in the mean amplitude and frequency were tested using Student’s two-tailed paired t-test, using absolute values rather than normalized ones. Values of p < 0.05 were considered significant.
The ionic composition of the incubation solution consisted of (in mM) 124 NaCl, 2 KCl, 1 KH2PO4, 26 NaHCO3, 2 CaCl2, 1 MgCl2 and 10 glucose saturated with 95% O2 and 5% CO2. The pH was about 7.45. The standard external solution was (in mM) 150 NaCl, 5 KCl, 2 CaCl2, 1 MgCl2, 10 glucose and 10 Hepes, and was adjusted to a pH of 7.4 with Tris-base. The Ca2+-free external solution was (in mM) 150 NaCl, 5 KCl, 3 MgCl2, 2 EGTA, 10 glucose and 10 Hepes, and was adjusted to a pH of 7.4 with Tris-base. For recording mIPSCs, these external solutions routinely contained 300 nM tetrodotoxin (TTX), 10 μM 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX) and 50 μM DL-2-amino-5-phosphonovaleric acid (APV) to block voltage-dependent Na+ channels and ionotropic glutamate receptors, respectively. The ionic composition of the internal (pipette) solution consisted of (in mM) 140 Cs-methanesulfonate, 5 TEA-Cl, 5 CsCl, 2 EGTA, 2 ATP-Mg and 10 Hepes with a pH adjusted to 7.2 with Tris-base.
The drugs used in this study were APV, forskolin, muscimol, EGTA, 9-(tetrahydro-2-furanyl)-9H-purin-6-amine (SQ22536), ATP-Mg (from Sigma, St. Louis, MO, USA) and TTX, bicuculline, CNQX, KT5720, N6-cyclopentyladenosine (CPA), 4-[2-[[6-amino-9-(N-ethyl-β-D-ribofuranuronamidosyl)-9H-purin-2-yl]amino]ethyl]benzenepropanoic acid hydrochloride (CGS21680), 8-cyclopentyl-1,3-dipropylxanthine (DPCPX), 4-ethylphenylamino-1,2-dimethyl-6-methylaminopyrimidium chloride (ZD7288), (from Tocris, Bristol, UK). All solutions containing drugs were applied using the ‘Y–tube system’ for rapid solution exchange (Akaike and Harata 1994).
GABAergic mIPSCs in mechanically dissociated TMN neurons
After a brief mechanical dissociation of the TMN region, large (> 20 μm in a somatic diameter) and small neurons (10–15 μm) were observed (Fig. 1a). Large neurons displayed a variety in the shape of the somata including fusiform, triangular, and multipolar forms, and were comparable to those previously described types (Takeshita et al. 1998). To confirm whether these large neurons belong to TMN histaminergic neurons, their immunoreactivity against HDC, a marker of histaminergic neurons, was examined. As shown in Fig. 1a, large neurons were HDC-positive, whereas small ones were HDC-negative, indicating that large neurons were TMN histaminergic neurons. Therefore, these large neurons were used in all subsequent electrophysiological experiments. In the presence of 300 nM TTX, 10 μM CNQX and 50 μM APV, which block voltage-dependent Na+ channels and ionotropic glutamate receptors, prominent spontaneous outward currents were observed from these isolated TMN neurons using a conventional whole-cell recording technique under voltage-clamp conditions (at a holding potential of 0 mV). Under these conditions, the application of 10 μM bicuculline, a selective GABAA receptor antagonist, completely and reversibly blocked all spontaneous outward currents (n = 12) (Fig. 1b). Thus the spontaneous events were identified as GABAergic spontaneous miniature inhibitory postsynaptic currents (mIPSCs) mediated by GABAA receptors.
Adenosine activates presynaptic A1 receptors to decrease spontaneous GABA release on to TMN neurons
To examine whether the activation of presynaptic adenosine receptors affects spontaneous GABA release on to TMN neurons, the effect of adenosine on GABAergic mIPSCs were observed. The application of 100 μM adenosine decreased mIPSC frequency in the majority of TMN neurons (12 of 17 neurons; 70.6%) tested (Fig. 2a and b). As shown in Fig. 2b, adenosine shifted the distribution of the inter-event interval to the right (p < 0.01; K-S test) without affecting the distribution of the current amplitude (p = 0.75; K-S test). The pooled data indicate that adenosine decreased the mean mIPSC frequency to 66.1 ± 6.0% of the control (n = 12, p < 0.05), but had no significant effect on the mean current amplitude (94.9 ± 3.0% of the control, n = 12, p = 0.17) (Fig. 2b insets). In addition, adenosine inhibited spontaneous GABA release in a concentration-dependent manner (a half-inhibitory concentration was 1.1 μM), in which adenosine even at a 1 μM concentration significantly decreased GABAergic mIPSC frequency to 76.3 ± 7.3% of the control (n = 6, p < 0.05) (Fig. 2c). The results suggest that adenosine acts presynaptically to decrease the probability of spontaneous GABA release. Adenosine can activate a variety of receptors including A1, A2A, A2B and A3 receptor subtypes (for review; Fredholm et al. 2001). Since A1 receptors are known to inhibit neurotransmitter release at a variety of central synapses (Scanziani et al. 1992; Wu and Saggau 1994; Chen and van den Pol 1997; Okada et al. 2001; Jeong et al. 2003), the effect of DPCPX, a selective A1 receptor antagonist, on the adenosine-induced inhibition of GABAergic mIPSC frequency was examined. The adenosine (100 μM)-induced inhibition of mIPSC frequency was completely attenuated in the presence of 100 nM DPCPX (119.5 ± 18.6% of the DPCPX condition, n = 10, p = 0.11) (Fig. 2d).
Furthermore, CPA (1 μM), a selective A1 receptor agonist, mimicked adenosine-induced inhibition of mIPSC frequency (53.4 ± 3.5% of the control, n = 15, p < 0.01), without affecting the mean current amplitude (92.7 ± 3.5% of the control, n = 15, p = 0.17) (Fig. 3a and b). The effect of CGS21680, a selective A2A receptor agonist, on spontaneous GABA release was also tested. As shown in Fig. 3b, 1 μM CGS21680 had no effect on GABAergic mIPSC frequency (84.9 ± 6.0% of the control, n = 5, p = 0.08). In addition, CPA did not affect the muscimol-induced currents (99.9 ± 0.7% of the control, n = 5, p = 0.87) (Fig. 3c and d), indicating that the activation of A1 receptors does not change the sensitivity of postsynaptic GABAA receptors. These results suggest that the adenosine-induced decrease in spontaneous GABA release is mediated by presynaptic A1 receptors. In all subsequent experiments, CPA was used to activate presynaptic A1 receptors based on its selectivity for A1 receptors.
Mechanisms underlying A1 receptor-mediated inhibition of spontaneous GABA release
The activation of presynaptic metabotropic receptors inhibits multiple types of voltage-dependent Ca2+ channels (VDCCs) and Ca2+ influx from the extracellular space, and decreases neurotransmitter release at various synapses (for review; Wu and Saggau 1997). To test the involvement of presynaptic VDCCs in the A1 receptor-mediated inhibition of spontaneous GABA release, the effect of Cd2+, a general VDCC blocker, on the CPA-induced decrease in mIPSC frequency was observed. The application of 200 μM Cd2+ significantly decreased both mIPSC frequency and amplitude to 53.7 ± 15.7% (n = 6, p < 0.05) and 64.9 ± 5.9% of the control (n = 6, p < 0.05), respectively (Fig. 4a and b). After 10 min pre-treatment of Cd2+, CPA (1 μM) still decreased mIPSC frequency to 60.7 ± 4.9% of the Cd2+ condition (n = 6, p < 0.05, Fig. 4a and b). The effect of Ca2+-free external solution on the CPA-induced decrease in mIPSC frequency was also observed. Exposure of TMN neurons to the Ca2+-free external solution significantly decreased GABAergic mIPSC frequency to 37.2 ± 14.2% of the control (n = 5, p < 0.05), with a small decrease in the mean current amplitude (64.5 ± 9.2% of the control, n = 5, p < 0.05) (Fig. 4c and d). This result indicates that Ca2+ influx from the extracellular space partially contribute to the generation of mIPSCs (see also Jang et al. 2001; Nakamura et al. 2003). In the Ca2+-free external solution, CPA still decreased mIPSC frequency to 62.1 ± 9.6% of the Ca2+-free condition (n = 5, p < 0.05) (Fig. 4c and d). These results suggest that the A1 receptor-mediated inhibition of spontaneous GABA release is not related to the Ca2+ influx passing through presynaptic VDCCs.
The activation of adenosine A1 receptors is also known to open G-protein-coupled inwardly rectifying K+ (GIRK) channels to hyperpolarize presynaptic nerve terminals (Yang et al. 2004). Therefore, the effect of Ba2+, which is known to block GIRK channels (Gerber et al. 1989; Birnstiel et al. 1992), on the CPA-induced decrease in GABAergic mIPSC frequency was observed. Ba2+ at 1 mM concentration greatly increased GABAergic mIPSC frequency to 438.4 ± 102.1% of the control (n = 7, p < 0.05), without affecting the mean current amplitude (108.1 ± 22.5% of the control, n = 7, p = 0.75) (Fig. 5a and b). In the presence of Ba2+, however, CPA failed to decrease mIPSC frequency (136.4 ± 14.0% of the Ba2+ condition, n = 7, p = 0.08) (Fig. 5a and b). The results suggest that the A1 receptor-mediated inhibition of spontaneous GABA release is related to the activation of pre-synaptic GIRK channels.
In general, A1 receptor activation is negatively coupled to cAMP formation by inhibiting adenylyl cyclase (AC) via Gi/o protein. A decrease in intracellular cAMP concentration can regulate the activity of protein kinase A (PKA), or directly affect spontaneous neurotransmitter release (Katsurabayashi et al. 2004). Therefore, the effect of forskolin, an AC activator, on the CPA-induced decrease in GABAergic mIPSC frequency was observed. Pre-treatment with 10 μM forskolin significantly increased mIPSC frequency to 164.8 ± 23.5% of the control (n = 7, p < 0.05) (Fig. 6a and b). The results suggest that an increase in cAMP concentration within presynaptic terminals indeed increases the probability of spontaneous GABA release. Under these conditions, CPA (1 μM) failed to decrease mIPSC frequency (104.8 ± 9.1% of the forskolin condition, n = 7, p = 0.66) (Fig. 6a and b). The effect of SQ22536, an AC inhibitor (Turcato and Clapp 1999), was also observed. SQ22536 (100 μM) itself decreased mIPSC frequency (67.6 ± 8.9% of the control, n = 5, p < 0.05) (Fig. 6c and d), indicating that presynaptic AC within GABAergic terminals is somewhat tonically active. Under these conditions, CPA (1 μM) failed to decrease mIPSC frequency (104.8 ± 7.6% of the SQ22536 condition, n = 5, p = 0.61) (Fig. 6c and d). The results suggest that the A1 receptor-mediated inhibition of spontaneous GABA release is closely related to the reduction of cAMP formation by inhibiting AC.
cAMP is known to regulate neurotransmitter release via PKA-dependent and/or PKA-independent signal transduction pathways at various central synapses (for review; Seino and Shibasaki 2005). Therefore, the effect of KT5720, a specific PKA inhibitor, on the CPA-induced decrease in GABAergic mIPSC frequency was observed. KT5720 (1 μM) hardly affected GABAergic mIPSC frequency (111.9 ± 26.9% of the control, n = 5, p = 0.69) (Fig. 7a and b). Under these conditions, CPA (1 μM) failed to decrease mIPSC frequency (123.0 ± 36.0% of the KT5720 condition, n = 5, p = 0.62) (Fig. 7a and b). The effect of ZD7288, a hyperpolarization and cyclic nucleotide-activated (HCN; also called Ih) channel blocker, on the CPA-induced decrease in GABAergic mIPSC frequency was also observed. ZD7288 (30 μM) itself slightly but not significantly decreased mIPSC frequency (75.1 ± 19.6% of the control, n = 4, p = 0.33) (Fig. 7c and d). However, CPA still decreased GABAergic mIPSC frequency to 48.6 ± 3.6% of the ZD7288 condition (n = 4, p < 0.01) even in the presence of ZD7288 (Fig. 7c and d). The results suggest that the A1 receptor-mediated inhibition of spontaneous GABA release is closely related to the PKA rather than HCN channels.
Presynaptic A1 receptors on GABAergic terminals in TMN neurons
A growing body of evidence has suggested that A1 receptors are involved in the presynaptic inhibition of neurotransmitter release at central synapses. Although the activation of A1 receptors are well known to inhibit excitatory glutamatergic transmission (Scanziani et al. 1992; Thompson et al. 1992; Wu and Saggau 1994; Ulrich and Huguenard 1995; Calabresi et al. 1997; Kimura et al. 2003; Yang et al. 2007), A1 receptor-mediated inhibition of GABAergic transmission is not so frequently reported (Ulrich and Huguenard 1995; Chen and van den Pol 1997; Bagley et al. 1999; Jeong et al. 2003). The present results provide an additional example demonstrating that functional adenosine A1 receptors are present on GABAergic nerve terminals projecting to TMN neurons, and that their activation decreases the probability of spontaneous GABA release. Several lines of evidence support the conclusion that presynaptic A1 receptor are responsible for the adenosine-induced decrease in mIPSC frequency. First, adenosine decreased mIPSC frequency without affecting the current amplitude, indicating that adenosine acts presynaptically to decrease spontaneous GABA release. Furthermore, no postsynaptic currents were observed during the application of adenosine or CPA. Second, adenosine actions on mIPSC frequency were mimicked by a selective A1 receptor agonist and blocked by a selective A1 receptor antagonist. Finally, mechanically dissociated TMN neurons used in this study have cell-free presynaptic nerve terminals (see also Akaike and Moorhouse 2003).
On the other hand, although a recent study demonstrated that a selective A2A receptor agonist increases GABA release in the TMN (Hong et al. 2005), the present results suggest that A2A receptors are not involved in the modulation of GABAergic transmission in TMN neurons, because adenosine did not change GABAergic mIPSC frequency even after the blockade of A1 receptors. Also a selective A2A receptor agonist had no effect on GABAergic mIPSC frequency. One explanation for this discrepancy is that A2A receptors might be selectively expressed on the soma rather than axon terminals of GABAergic neurons projecting to TMN neurons, so that the activation of somatic A2A receptors increases the firing rate of GABAergic neurons, and thus increases GABA release on to TMN neurons. However the detailed subcellular localization of A2A receptors within the TMN and preoptic area is poorly understood. In this stage, the results could not evaluate the role of somatic A2A receptors in GABAergic transmission because mechanically dissociated TMN neurons have functional cell-free pre-synaptic nerve terminals. More studies including the subcellular localization of A2A receptors would be needed to evaluate the involvement of A2A receptors in the regulation of GABAergic transmission in the TMN.
Possible mechanisms underlying the A1 receptor-mediated inhibition of spontaneous GABA release
Adenosine A1 receptors are coupled to Gi/o protein (Scholz and Miller 1992; Greif et al. 2000), and have three possible modes of action in causing presynaptic inhibition for neurotransmitter release: the reduced production of cAMP by inhibiting of AC, the reduced Ca2+ influx by inhibiting presynaptic VDCCs, and the increased K+ conductance by activating GIRK channels (Wu and Saggau 1997; Dunwiddie and Masino 2001). Likewise, the mechanisms underlying the A1 receptor-mediated inhibition of neurotransmitter release are very diverse among the central synapses. For example, the activation of A1 receptors inhibits neurotransmitter release by reducing the Ca2+ influx through presynaptic N- and/or P/Q-type VDCCs (Wu and Saggau 1994; Okada et al. 2001; Gundlfinger et al. 2007). In the present study, however, presynaptic VDCCs are unlikely to contribute to the A1 receptor-mediated inhibition of spontaneous GABA release, because CPA still decreased GABAergic mIPSC frequency either in the Ca2+ free external solution or in the presence of Cd2+, a general VDCC blocker. These results provide additional evidence that A1 receptors act on the synaptic release machinery downstream of Ca2+ influx.
In addition to the modulation of presynaptic VDCCs, the activation of presynaptic A1 receptors is known to decrease the probability of neurotransmitter release by inhibiting AC and the resultant reduction of cAMP/PKA signal transduction pathway, as shown in previous studies (Bagley et al. 1999; Jeong et al. 2003; Yang et al. 2004). In fact, cAMP-dependent PKA activation is known to regulate the number of readily releasable vesicles, without affecting either the number of morphologically docked vesicle or the number of active synaptic terminals (Trudeau et al. 1996), as many presynaptic proteins related to synaptic release such as synaptosomal-associated proteins (SNAPs), rabphilin 3A and Rab3-interacting molecule 1α (RIM1α) are known to be phosphorylated by PKA (Turner et al., 1999). In TMN neurons, the CPA-induced decrease in mIPSC frequency was also completely occluded by ether SQ22536 or KT5720. These results indicate that the activation of presynaptic A1 receptors is likely to decrease GABAergic mIPSC frequency via the AC/cAMP/PKA signal transduction pathway. On the other hand, cAMP can affect neurotransmitter release via a PKA-independent mechanism. For example, the activation of HCN channels by cAMP at neuromuscular junctions has been shown to facilitate neurotransmitter release (Beaumont and Zucker 2000). However, the involvement of HCN channels in the CPA-induced decrease in mIPSC frequency would be negligible because CPA still decreased mIPSC frequency even in the presence of ZD7288.
Presynaptic K+ channels play an important role in the regulation of neurotransmitter release (Meir et al. 1999), and especially GIRK channels are known to be regulated a number of G-protein-coupled receptors including A1 receptors (Gerber et al. 1989; Yang et al. 2004). In the present study, since the A1 receptor-mediated inhibition of mIPSC frequency was completely occluded after the blockade of GIRK channels, presynaptic GIRK channels would be other targets of the A1 receptor-mediated inhibition of spontaneous GABA release onto TMN neurons.
The somnogenic effect of adenosine has been proposed to be mediated by both A1 and A2A receptor subtypes (for reviews; Basheer et al. 2004). The administration of CGS 21680, a selective A2A receptor agonist, into the lateral ventricle in wild type mice induces rapid eye movement (REM) and non-REM sleep in a dose- and time-dependent manner, but this effect disappears in A2A receptor knock out mice (Urade et al. 2003). In addition, whereas CGS 21680 decreases histamine release in both the frontal cortex and medial pre-optic area in a dose-dependent manner, it increases GABA release in the TMN but not the frontal cortex (Hong et al. 2005). Furthermore, adenosine, via the activation of A2A receptors, stimulates a subpopulation of VLPO GABAergic neurons (Gallopin et al. 2005). On the contrary, the A1 receptor-mediated somnogenic effect is likely to be related to the cholinergic system of the basal forebrain. For example, the activation of A1 receptors stimulates Ca2+ release in cholinergic but not non-cholinergic neurons (Basheer et al. 2002), and the administration of A1 receptor antisense oligonucleutides into the basal forebrain reduces non-REM sleep (Thakkar et al. 2003). However, such a somnogenic action of A1 receptors has been implicated by a recent report showing that caffeine, which blocks A1 and A2A receptors similarly, promotes wakefulness in wild type and A1 receptor knock out mice, but not in A2A receptor knock out mice (Huang et al. 2005).
In the brain, the extracellular adenosine concentration has been estimated to be in the range of 30–300 nM under normal physiological conditions (Daly and Fredholm 1998; Porkka-Heiskanen et al. 2000), and it has been shown to increase with increased neuronal activity or metabolism (Winn et al. 1980; Van Wylen et al. 1986). In addition, sleep deprivation or prolonged wakefulness is known to increase the extracellular adenosine concentration in the basal forebrain as well as the cortex (Porkka-Heiskanen et al. 1997, 2000). The results suggest that adenosine increased during wakefulness might promote sleep by inhibiting the basal forebrain arousal system. Although the change in adenosine level in the TMN during sleep-wakefulness remains to be elucidated at present, extracellular adenosine might regulate the excitability of TMN neurons by activating A1 receptors expressed on GABAergic terminals, because the activation of A1 receptors by extracellular adenosine, at as low as 1 μM concentration, could inhibit spontaneous GABA release. Alternatively, adenosine may inhibit TMN neurons by activating A1 receptors on the cell body of TMN neurons, although a previous study has shown that adenosine has no effect on membrane response in TMN neurons (Furukawa et al. 1994).
The present study has shown that the adenosine-induced inhibition of spontaneous GABA release onto TMN neurons was primarily mediated by presynaptic A1 receptors, but not A2A receptors. Although the exact origin of the GABAergic terminals on mechanically dissociated TMN neurons is not clear at present, some portions of GABAergic terminals might originate from the VLPO. Despite of the origin of GABAergic nerve terminals, our present findings should be incompatible with the somnogenic action of adenosine as described above, because the disinhibition, that is the inhibition of inhibitory GABAergic transmission, of TMN neurons by presynaptic A1 receptors would eventually increase the excitability of TMN histaminergic neurons. Therefore, it can be proposed that, at least in part, pre-synaptic A1 receptors on GABAergic nerve terminals might play a role in preventing TMN neurons from being over-inhibited. In conclusion, A1 receptors expressed on GABAergic nerve terminals projecting to TMN neurons may play an important role in the fine modulation of the excitability of the TMN neurons as well as the regulation of sleep-wakefulness.
We thank Prof. Harold Martin (Kyungpook National University) for correcting the English. This study was supported by the BioMedical Research Institute grant from Kyungpook National University Hospital (2007) to J.-K. Choi. D.-S. Yeom and J.-H. Cho equally contributed to this study.