Skip to main content
NCBI home page
The Journal of Physiology logoLink to The Journal of Physiology
. 1999 Nov 1;520(Pt 3):815–825. doi: 10.1111/j.1469-7793.1999.00815.x

Adenosine-induced presynaptic inhibition of IPSCs and EPSCs in rat hypothalamic supraoptic nucleus neurones

Stéphane H R Oliet 1, Dominique A Poulain 1
PMCID: PMC2269632  PMID: 10545146

Abstract

  1. The effects of adenosine on synaptic transmission in magnocellular neurosecretory cells were investigated using whole-cell patch-clamp recordings in acute rat hypothalamic slices that included the supraoptic nucleus.

  2. Adenosine reversibly reduced the amplitude of evoked inhibitory (IPSCs) and excitatory (EPSCs) postsynaptic currents in a dose-dependent manner (IC50≈ 10 μm for both types of current).

  3. Depression of IPSCs and EPSCs by adenosine was reversed by the application of the A1 adenosine receptor antagonist 8-cyclopentyl-1,3-dimethylxanthine (CPT; 10 μm).

  4. When pairs of stimuli were given at short intervals, adenosine inhibitory action was always less effective on the second of the two responses than on the first, resulting in an increased paired-pulse facilitation and suggesting a presynaptic site of action. This observation was confirmed by analysis of spontaneous miniature synaptic currents whose frequency, but not amplitude or kinetics, was reversibly reduced by 100 μM adenosine.

  5. CPT had no effect on synaptic responses evoked at a low frequency of stimulation (0.05–0.5 Hz), indicating the absence of tonic activation of A1 receptors under these recording conditions. However, CPT inhibited a time-dependent depression of both IPSCs and EPSCs induced during a 1 Hz train of stimuli.

  6. Taken together, these results suggest that adenosine can be released within the supraoptic nucleus at a concentration sufficient to inhibit the release of GABA and glutamate via the activation of presynaptic A1 receptors. By its inhibitory feedback action on the major afferent inputs to oxytocin and vasopressin neurones, adenosine could optimally adjust electrical and secretory activities of hypothalamic magnocellular neurones.

Magnocellular neurosecretory cells of the hypothalamus, which synthesise vasopressin or oxytocin, project their axons to the neurohypophysis where their neurohormones are secreted directly into the bloodstream. Whereas vasopressin is known to play a key role in body-fluid homeostasis, oxytocin is principally involved in lactation and parturition. The release of these peptides at neurohypophysial terminals is correlated to distinct patterns of electrical activity generated at the level of magnocellular neurone cell bodies located in the supraoptic (SON) and paraventricular nuclei of the hypothalamus (Bicknell, 1988; Poulain & Theodosis, 1988). This neuronal activity is itself under the control of excitatory and inhibitory afferent inputs arising from many brain areas (see Anderson et al. 1990). As elsewhere in the brain, glutamate and GABA are the main excitatory and inhibitory neurotransmitters in the hypothalamus including those within the magnocellular nuclei (Theodosis et al. 1986; Van den Pol et al. 1990; Decavel & Van den Pol, 1990). Electrophysiological recordings have indicated the presence of functional GABA and glutamate receptors on SON neurones (Randle & Renaud, 1987; Hu & Bourque, 1991). Furthermore, ultrastructural studies have revealed that GABAergic and glutamatergic terminals, respectively, account for about 40 and 20 % of all synapses on SON neurones (Gies & Theodosis, 1994; El Majdoubi et al. 1996, 1997). Modulation of these inputs, therefore, should represent a very potent way to regulate the activity of hypothalamic magnocellular neurones.

A wide range of substances have been shown to serve as neuromodulators in the brain. Among them is adenosine, which modulates synaptic transmission in several brain areas (see for example Proctor & Dunwiddie, 1987; Ulrich & Huguenard, 1995; Shen & Johnson, 1997). This nucleoside, produced by the metabolic breakdown of ATP, is present in all cells and can be released under physiological conditions (Mitchell et al. 1993; Manzoni et al. 1994). In the central nervous system, three main types of adenosine receptor (A1, A2 and A3) have been described, all of which are coupled to G-proteins (Greene & Haas, 1991; Jacobson, 1998). Few data are available regarding the A3 receptor. On the other hand, activation of the A1 type inhibits adenylate cyclase activity, while activation of the A2 receptor increases it (Van Calker et al. 1979). Most of the reported effects of adenosine on synaptic transmission are mediated through the activation of presynaptic A1 receptors (Lambert & Teyler, 1991; Scanziani et al. 1992; Ulrich & Huguenard, 1995; Shen & Johnson, 1997).

In this study, we have investigated a possible role for adenosine in regulating the synaptic activity of magnocellular supraoptic neurones using whole-cell patch-clamp recordings in acute rat hypothalamic slices. We provide evidence that functional adenosine receptors are present in the SON, that their activation depresses GABA and glutamate release, and that these effects are mediated via presynaptic A1 receptors. In addition, we show that an activity-dependent release of endogenous adenosine can be induced within the SON, depressing both excitatory and inhibitory transmission.

METHODS

All experiments were carried out according to French/European guidelines.

Slice preparation

Male and female Sprague-Dawley rats (1–2 months old) were killed by decapitation using a guillotine. The brain was quickly removed and placed in ice-cold artificial cerebrospinal fluid (ACSF; see below) saturated with 5 % CO2 and 95 % O2. Coronal slices (300 μm) were cut with a vibratome (Campden Instruments Ltd, UK) from a block of tissue containing the hypothalamus. Slices including the supraoptic nucleus were hemisected along the midline and allowed to recover for at least 1 h before recording. A slice was then transferred into a recording chamber where it was submerged and continuously perfused (1–2 ml min−1) with ACSF. The composition of the ACSF was (mM): 123 NaCl, 2.5 KCl, 1 Na2HPO4, 26.2 NaHCO3, 1.3 MgCl2, 2.5 CaCl2 and 10 glucose (pH 7.4; 295 mosmol kg−1). All experiments were carried out at room temperature.

Patch-clamp recording

Magnocellular neurosecretory cells were visually identified using infrared differential interference contrast microscopy. Patch-clamp recording pipettes (3–5 MΩ) were filled with a solution (adjusted to pH 7.1 with CsOH) containing (mM): 130 CsCl, 8 NaCl, 10 Hepes, 0.2 EGTA, 10 glucose, 2 Mg-ATP, 0.3 GTP and 5 QX-314. Membrane currents were recorded using an EPC7 List amplifier (List, Germany). Signals were filtered at 2 kHz and digitised at 5 kHz via a DigiData 1200 interface (Axon Instruments, Inc.). Series resistance (4–15 MΩ) was monitored and cells were excluded from data analysis if more than a 15 % change occurred during the course of the experiment. All cells were held at −60 mV in voltage-clamp mode. To evoke synaptic responses, either a bipolar stainless steel or a concentric stimulating electrode connected to an isolated stimulator (Digitimer Ltd, UK) was placed in the hypothalamic region dorsolateral to the recording site as described previously (Kombian et al. 1996). Synaptic responses were evoked at 0.05 Hz, unless otherwise stated, using square pulse of 0.1 ms duration, and analysed on-line using data acquisition and analysis software (Acquis, BioLogic, France). Spontaneous unitary synaptic currents (miniature synaptic currents) obtained in the presence of tetrodotoxin (TTX; 0.5 μM) were stored on videotape via a pulse-code modulator (Cygnus Technology, Inc., USA), detected and analysed off-line using Axograph (Axon Instruments, Inc.). To study paired-pulse facilitation (PPF), two synaptic responses (S1 and S2) were evoked by a couple of stimuli given at short intervals (40 ms for EPSCs and 60 ms for IPSCs to take into account the longer decay of GABAergic responses). PPF was expressed as the amplitude ratio of the second synaptic response to the first synaptic response (S2/S1).

Data were compared statistically with either the non-parametric Kolmogorov-Smirnov test or Student's paired t test. Significance was assessed at P < 0.05. All data are reported as means ±s.e.m.

Drugs

All drugs were bath applied. Appropriate stock solutions were made and diluted with ACSF just before application. Drugs used were bicuculline methiodide, 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), D-(−)-2-amino-5-phosphonopentanoic acid (AP5), adenosine, 2-hydroxysaclofen, 8-cyclopentyl-1,3-dimethylxanthine (CPT; Research Biomedicals International), QX-314 chloride (Almonde Labs) and TTX (Sigma).

RESULTS

Adenosine inhibits evoked synaptic responses

Stimulation of the region dorsolateral to the SON elicited synaptic responses previously shown to be mediated by the activation of GABAA and non-NMDA glutamate receptors (Randle et al. 1986; Gribkoff & Dudek, 1990). In the presence of CNQX (20 μM) and AP5 (50 μM), blockers of non-NMDA and NMDA glutamate receptors, respectively, pure GABAA receptor-mediated IPSCs were recorded. Because of the chloride-based intracellular solution used in our experiments, inward IPSCs were recorded when the cells were held at −60 mV. To test whether activation of adenosine receptors could modulate GABAA receptor-mediated synaptic responses, adenosine (100 μM) was added to the bathing solution. As illustrated in Fig. 1A, adenosine caused a significant and reversible inhibition of evoked IPSCs in all cells tested (n = 13). On average, adenosine reduced the amplitude of evoked IPSCs by 60.0 ± 4.9 %. In five cells, bicuculline methiodide (BMI; 10 μM) was added at the end of the experiment, resulting in a complete inhibition of the IPSCs, thus confirming that the recorded synaptic responses were entirely mediated via GABAA receptors.

Figure 1. Adenosine inhibits the amplitude of evoked synaptic currents.

Figure 1

Open in a new tab

A, bath application of adenosine (100 μM) dramatically reduced evoked IPSCs. Complete recovery was obtained after drug washout. IPSCs were completely abolished by the addition of bicuculline methiodide (BMI; 10 μM) at the end of the experiment. The insets are averaged sample records of 5 consecutive sweeps taken at the indicated times. B, evoked EPSCs were reversibly depressed by adenosine (100 μM). Bath application of CNQX (20 μM) at the end of the experiment completely inhibited the synaptic responses. Insets as for A.

When similar experiments were carried out in the absence of CNQX and AP5, but in the presence of BMI to isolate glutamatergic responses, bath application of 100 μM adenosine also resulted in a reversible reduction (56.5 ± 6.0 %) of EPSCs in all cells tested (n = 10). The EPSCs recorded in these conditions were mediated purely by non-NMDA glutamate receptors since the addition of CNQX completely abolished the responses in six out of six cells (Fig. 1B).

The effect of adenosine on both IPSCs and EPSCs had a slow onset (15 s to 5 min), lasted several minutes after application, and was reversible upon washout of the drug.

The use of varying concentrations of adenosine (1–50 μM) allowed us to show that the adenosine-mediated inhibition of GABAergic and glutamatergic synaptic responses was dose dependent. Evoked IPSCs were reduced, on average, by 3.9 ± 2.1 % (n = 3 cells), 21.0 ± 12.3 % (n = 5) and 59.9 ± 8.8 % (n = 4) for 1, 10 and 50 μM adenosine, respectively. Similarly, evoked EPSCs were depressed by 2.0 ± 0.3 % (n = 4), 17.8 ± 4.2 % (n = 4) and 58.0 ± 10.0 % (n = 6) for 1, 10 and 50 μM adenosine, respectively. As illustrated in Fig. 2, fitting these data with a Boltzmann function revealed values for half-maximal inhibition (IC50) of 12.7 μM for IPSCs and 12.8 μM for EPSCs. The threshold dose for adenosine-mediated effects was 1 μM in both cases. These similarities in IC50, maximal effect and threshold dose suggest that the same type of adenosine receptor was involved in the modulation of both IPSCs and EPSCs.

Figure 2. Adenosine-mediated effects on synaptic currents are dose dependent.

Figure 2

Open in a new tab

Data points are means (n = 3–13) of the percentage inhibition induced by the indicated adenosine concentration on IPSCs (•) and EPSCs (^). Fitting the two sets of data with a Boltzmann function revealed IC50 values of 12.7 μM for IPSCs and 12.8 μM for EPSCs.

The effect of adenosine is mediated via A1 receptors

In order to identify the receptors involved in the modulation of IPSCs and EPSCs by adenosine, we used the specific A1 receptor antagonist CPT. In this set of experiments, adenosine (100 μM) was first applied to induce inhibition of evoked synaptic responses. Subsequently, CPT (10 μM) was added to the bathing solution containing adenosine. The A1 receptor antagonist reversed the action of adenosine on both IPSCs and EPSCs (Fig. 3Aa andBa). On average, evoked IPSC amplitude returned to 93.9 ± 9.6 % of control (n = 6; Fig. 3Ab) in the presence of adenosine plus CPT. Similarly, CPT caused the EPSC amplitude to return to 83.6 ± 6.8 % of control (n = 5; Fig. 3Bb). These results suggest that most of the effects induced by adenosine on IPSCs and EPSCs are mediated via A1 receptors. CPT by itself did not cause a significant change in the amplitude of GABAA and non-NMDA glutamate receptor-mediated synaptic currents, indicating an absence of tonic activation of A1 receptors by ambient adenosine under our conditions.

Figure 3. Adenosine effects are mediated by A1 receptors.

Figure 3

Open in a new tab

Aa, bath application of CPT (open bar; 10 μM) reversed the inhibition of evoked IPSCs induced by adenosine (filled bar; 100 μM). A summary of this effect obtained in 6 different supraoptic neurones is illustrated in Ab. IPSC amplitude was normalized to the value obtained under control conditions. Ba, in the presence of adenosine (100 μM), CPT (10 μM) caused the evoked EPSCs to return to baseline values. A summary of this effect obtained in 5 different cells is illustrated in Bb.

Adenosine affects paired-pulse facilitation

To determine the site of action of adenosine, we first examined the effects of adenosine on paired-pulse facilitation (PPF) in a subset of cells. As shown in Fig. 4, under control conditions, the amplitude of the two IPSCs (S1 and S2) was similar, indicating that the second synaptic response (S2) was not facilitated by the first IPSC (S1; Fig. 4Aa) evoked 60 ms earlier. As expected, application of adenosine (100 μM; Fig. 4Ab) inhibited the first IPSC by 47 %. However, the second synaptic response was depressed by only 18 %, indicating that the inhibitory effect of adenosine on the amplitude of the second of the two IPSCs was significantly less than that on the first (Fig. 4Aa and Ab; P = 0.004). As illustrated in Fig. 4Ac, the PPF ratio (S2/S1; see Methods) obtained for IPSCs was on average close to 1 under control conditions (1.14 ± 0.02; n = 6), and increased to 2.14 ± 0.20 in the presence of 100 μM adenosine.

Figure 4. Adenosine affects PPF.

Figure 4

Open in a new tab

Aa, averaged sample records of 10 consecutive sweeps obtained under control conditions (left; thick trace) and in the presence of adenosine (100 μM) in the bath (right; thin trace). Paired-pulse interval was 60 ms. Ab, superimposition of the 2 traces normalised to the first IPSC obtained under control conditions reveals that PPF is increased in the presence of adenosine. Mean PPF ratios obtained in 6 different neurones under control conditions and in the presence of adenosine are plotted in Ac. B, similar experiments to those illustrated in A, obtained for evoked EPSCs. PPF ratios plotted in Bc were obtained from 5 different supraoptic neurones.

We used the same protocol on EPSCs (Fig. 4Ba and Bb). As observed for IPSCs, the inhibitory effect of adenosine on the amplitude of the second EPSC was clearly less than that on the first (Fig. 4Ba and Bb; P < 0.0001). In the presence of adenosine, the PPF ratio increased on average from 1.23 ± 0.17 to 1.99 ± 0.19 (n = 9; Fig. 4Bc).

Since PPF reflects a change in the probability of transmitter release, its modulation suggests that adenosine acts presynaptically to depress both IPSCs and EPSCs.

The frequency, but not the amplitude, of miniature synaptic currents is sensitive to adenosine

To investigate further the location of adenosine receptors, we next analysed the action of adenosine on the amplitude and frequency distributions of miniature IPSCs and EPSCs (mIPSCs and mEPSCs). These spontaneous unitary events were recorded in the presence of TTX in the bathing solution to prevent the occurrence of action potential-driven synaptic currents and, therefore, ensure that monoquantal events were recorded. At the concentration used (0.5 μM), TTX was found to completely abolish evoked synaptic transmission. As for evoked responses, mIPSCs were isolated by blocking glutamate receptors with AP5 and CNQX, whereas blocking GABAA receptors with bicuculline isolated mEPSCs.

Typically, the frequency of mIPSCs ranged between 3.1 and 8.3 Hz under control conditions (n = 5) while the mean amplitude varied between −61.7 and −102.0 pA (500–800 events for each cell). When adenosine (100 μM) was added to the bathing solution, the inhibitory synaptic activity was dramatically reduced (Fig. 5A). When the amplitude distributions of mIPSCs were compared in a particular cell (Fig. 5B), no significant difference was found between the distribution obtained under control conditions and that obtained in the presence of adenosine (P = 0.17). Moreover, the kinetics of mIPSCs were not affected by application of adenosine as illustrated in Fig. 5B (inset), where the means of 50 consecutive unitary events obtained under control conditions and in the presence of adenosine are superimposed. Conversely, adenosine produced a significant shift towards the right in the event interval distribution (Fig. 5C; P < 0.0001), reflecting a decrease of mIPSC frequency (−51 %). In five neurones, adenosine was seen to decrease the frequency of mIPSCs, on average, by 47.2 ± 8.1 % (Fig. 5D), without affecting the mean mIPSC amplitude (quantal size, q; +1.6 ± 3.7 %).

Figure 5. Effects of adenosine on mIPSCs.

Figure 5

Open in a new tab

A, consecutive traces (12 s each) showing typical mIPSCs before, during and after application of 100 μM adenosine. B, cumulative plots of amplitude distributions before (thick line) and during (thin line) adenosine application for the above experiment were similar. Inset shows superimposed mIPSC means obtained in the absence and in the presence of adenosine. C, adenosine caused the event interval distribution to be shifted towards the right, indicating a decrease in mIPSC frequency. Pooled data from 5 cells are shown in D, illustrating the percentage change in mean amplitude (q) and frequency (Hz) of mIPSCs caused by adenosine.

A similar analysis was carried out on mEPSCs. Under control conditions, mEPSC frequency varied between 2.0 and 2.8 Hz (n = 6). The mean mEPSC amplitude, calculated in six cells, ranged from −14.6 to −26.4 pA. As for mIPSCs, adenosine appeared to be equally efficient in depressing the excitatory synaptic activity in all neurones tested as illustrated in Fig. 6A. In this experiment, adenosine did not affect either the amplitude or the kinetics of mEPSCs (Fig. 6B). On the other hand, adenosine induced a significant shift towards the right of the event interval distribution (Fig. 6C; P < 0.001). On average, mEPSC frequency was depressed by 45.1 ± 9.0 % (Fig. 6D; n = 6) while the quantal size remained unchanged (−4.8 ± 2.1 %).

Figure 6. Effects of adenosine on mEPSCs.

Figure 6

Open in a new tab

A, consecutive traces (12 s each) showing typical mEPSCs before, during and after application of 100 μM adenosine. B, amplitude distributions before (thick line) and during (thin line) adenosine application for the above experiment were similar. Inset shows superimposed means of 50 consecutive mEPSCs obtained in the absence and presence of adenosine. C, the event interval distribution was shifted towards the right during adenosine application, indicating that a decrease in the frequency of mEPSCs has occurred. Pooled data from 5 cells are shown in D, illustrating the percentage change in mean amplitude and frequency of mEPSCs caused by adenosine.

It is usually accepted that modifications of miniature synaptic current frequency correspond to a change in neurotransmitter release (Redman, 1990). These findings, therefore, are consistent with a presynaptic location for the adenosine receptors, whose activation results in an inhibition of GABA and glutamate release. Conversely, since the amplitude distributions and the kinetics of mIPSCs and mEPSCs were unaffected, it is highly likely that GABAA and non-NMDA glutamate receptors are not directly modulated by adenosine.

Activity-dependent release of adenosine

As indicated earlier, the presence of CPT in the bath did not affect basal glutamatergic and GABAergic transmission, suggesting that adenosine receptors were not activated under our conditions by ambient adenosine. Previous studies in the hippocampus have shown, however, that adenosine release can be induced during sustained stimulation of synaptic inputs (Mitchell et al. 1993; Manzoni et al. 1994). We therefore tested whether we could induce endogenous adenosine release in the SON by using frequencies of stimulation higher than 0.05 Hz. These experiments were carried out in the presence of saclofen (20 μM) to prevent a possible activation of GABAB receptors, which could occur during stimulation at such frequencies and result in an adenosine-independent synaptic depression (Kombian et al. 1996; Mouginot et al. 1998). Likewise, 50 μM AP5 was added to the external solution to prevent the induction of NMDA-dependent plasticity that could develop during these stimulations as described in other brain areas (Malenka & Nicoll, 1993).

If adenosine is released during sustained afferent stimulation, activation of presynaptic A1 receptors should lead to the progressive depression of evoked synaptic currents. We found that neither inhibitory nor excitatory transmission was affected when the stimulation rate was increased up to 0.5 Hz (not shown). However, we did record a reliable and reproducible time-dependent decrease of IPSCs (n = 5) and EPSCs (n = 6) within the course of a 200 s train of stimuli given at 1 Hz (Fig. 7Aa and Ba). This depression started to develop after 60 s and reached a maximal level of inhibition at 130 s (70.1 ± 13.6 % of control for IPSCs and 68.2 ± 11.5 % for EPSCs). After allowing the synaptic currents to recover their initial values in response to a frequency of stimulation of 0.05 Hz (2–10 min), CPT (10 μM) was added to the bathing solution. In the same cells, the depression of IPSCs and EPSCs elicited during a 1 Hz train was then abolished (Fig. 7Ab and Bb); after 130 s, the mean IPSC and EPSC amplitudes were 102 ± 15 and 104 ± 10 % of control, respectively. The depression reappeared when the same stimulation protocol was applied after washout of CPT (n = 3 for IPSCs and n = 4 for EPSCs). These results suggest, therefore, that adenosine is released during a 1 Hz train of stimuli, and induces a delayed inhibition of both IPSCs and EPSCs.

Figure 7. Endogenous release of adenosine depresses synaptic responses.

Figure 7

Open in a new tab

Aa, data obtained from 5 different cells, pooled to illustrate the depression occurring when IPSCs were evoked at 1 Hz. Ab, in the same cells, the depression was abolished in the presence of CPT (10 μM) in the external solution. In this set of experiments, data were averaged every 4 s for the purpose of clarity, and normalised to the mean IPSC value obtained during the first 50 s under control conditions. Insets illustrate means of 5 consecutive IPSCs recorded at the indicated time. B, similar experiment to that in A, carried out on evoked EPSCs.

DISCUSSION

Presynaptic inhibition of GABA and glutamate release

Our observations provide direct evidence that adenosine inhibits both GABAergic and glutamatergic afferent transmission to hypothalamic magnocellular neurones. Since we did not record cells from a preferential location in the SON, and since all the cells tested were sensitive to adenosine (1–100 μM), it is very likely that adenosine modulates afferent inputs to both oxytocin- and vasopressin-secreting neurones.

Adenosine has been reported to exert a strong inhibitory modulation of glutamatergic (Lambert & Teyler, 1991; Li & Perl, 1994; Obrietan et al. 1995) or GABAergic (Hu & Li, 1997; Chen & Van den Pol, 1997) synaptic transmission in various central areas. In addition, there are also examples where adenosine was found to inhibit both excitatory and inhibitory transmission (Ulrich & Huguenard, 1995; Shen & Johnson, 1997). In all these studies, the site of action of adenosine was located presynaptically.

In the SON, we found that adenosine-induced inhibition of GABAergic and glutamatergic synaptic transmission was mediated through activation of A1 receptors. This conclusion is based on the observation that adenosine-mediated effects were completely reversed by the specific A1 receptor antagonist CPT. Such findings are in agreement with the presence of A1 receptor mRNA in the rat hypothalamus (Reppert et al. 1991; Mahan et al. 1991). It has been reported that postsynaptic A1 receptors may modulate SON neurone activity under certain circumstances (Noguchi et al. 1998). Although we cannot rule out the possibility that our recording conditions could have masked or inhibited postsynaptic effects, our experiments clearly demonstrated that adenosine inhibited excitatory and inhibitory synaptic currents by acting presynaptically. First, adenosine affected PPF, a type of facilitation known to result from a change in the probability of transmitter release. Second, adenosine reduced the frequency of mIPSCs and mEPSCs. Third, neither the amplitude nor the kinetics of mIPSCs and mEPSCs were affected by adenosine. Finally, no changes in input resistance were detected throughout our experiments. All these observations are consistent with a presynaptic action of adenosine on GABA and glutamate release.

Mechanism of action

The mechanisms leading to inhibition by adenosine of spontaneous and evoked release at glutamate and GABA synapses are likely to be the same since the magnitude and the time course of the two effects were comparable. Because PPF was affected by adenosine, adenosine must have inhibited synaptic transmission by depressing the probability of release (Pr) of GABA and glutamate rather than by reducing the number of functional synapses. How this could be achieved remains to be determined but one possible mechanism may involve the inhibition of voltage-gated Ca2+ channels by adenosine (see also Greene & Hass, 1991; Chen & Van den Pol, 1997; Noguchi et al. 1998). Reducing Ca2+ entry in the presynaptic axon terminal would prevent the coupling between action potentials and exocytosis of transmitter vesicles (Augustine et al. 1987). Nevertheless, such a mechanism appears unlikely to account for the action of adenosine on the frequency of mIPSCs and mEPSCs, since action potential-dependent transmitter release is completely abolished in the presence of TTX. Under these conditions, presynaptic voltage-gated Ca2+ channels would not be involved unless they were tonically active. This appears unlikely for several reasons. First, it was recently reported that perfusion of Ca2+-free medium did not affect the amplitude or the frequency of mEPSCs (Inenaga et al. 1998) and mIPSCs (Brussard et al. 1999) in SON neurones. Second, we found that the frequency and the amplitude of mIPSCs were unaffected in the presence of cadmium (30 μM), an inhibitor of voltage-gated Ca2+ channels, whereas evoked transmitter release was completely abolished (S. H. R. Oliet & R. Piet, unpublished observations).

Another mechanism may be that activation of A1 receptors induces an inhibition of transmitter release by affecting the protein-protein interaction cascade involved in exocytosis. Interestingly, a similar process may underlie the depression of mEPSC and mIPSC frequency induced by the activation of presynaptic GABAB receptors in the SON (Kabashima et al. 1997). Whether a common mechanism is involved following the activation of A1 adenosine and GABAB receptors remains to be determined. Both types of receptor are coupled to G-proteins and could eventually lead to the activation of the same second messenger system.

Release of endogenous adenosine

Until now, there have been few reports describing endogenous release of adenosine in the central nervous system. In the CA1 region of the hippocampus, high frequency stimulation led to release of adenosine, which resulted in a heterosynaptic inhibition of glutamate release (Mitchell et al. 1993; Manzoni et al. 1994). In these studies, it was proposed that tetanic stimulation induced a massive release of adenosine, which upon diffusion could depress synaptic transmission at neighbouring synapses. Using a much lower frequency of stimulation (1 Hz), we were able here to induce a delayed (50–60 s) homosynaptic depression at glutamatergic and GABAergic synapses, an effect that was prevented by the A1 receptor antagonist CPT. We conclude, therefore, that during a 1 Hz stimulus train, but not at lower frequencies, the Pr of GABA and glutamate from nerve terminals is modulated by presynaptic A1 receptors responding to endogenous adenosine.

Such experiments, however, do not reveal the site of adenosine release and/or production. Adenosine is found throughout the brain and can be released by neurones or glial cells (White & Hoehn, 1991). Enzymatic cleavage of ATP also produces adenosine in the extracellular space. In our experiments, therefore, endogenous adenosine, or its precursor, could originate from diverse sources, including stimulated nerve terminals, glial cells or the magnocellular neurones themselves. ATP is present in secretory granules (reviewed in Aunis, 1998) which can undergo somato-dendritic exocytosis in the magnocellular nuclei (reviewed in Ludwig, 1998). However, since cells were voltage clamped at −60 mV and since a blocker of voltage-gated sodium channels (QX-314) was present in the patch pipette solution, it is unlikely that action potential-dependent somato-dendritic release occurred in the recorded cells. We cannot rule out the possibility, nonetheless, that there was somato-dentritic release from neighbouring neurones, a release which would then modulate adjacent GABAergic and glutamatergic synapses. ATP can also be co-released from cholinergic and catecholaminergic terminals (Zimmerman, 1994). Such a mechanism could occur during our stimulation protocol since catecholaminergic inputs are present in the SON (Sawchenko & Swanson, 1982) and their activation has been shown to induce an ATP receptor-mediated response (Day et al. 1993). Interestingly, it has been reported that ATP and UTP, but not adenosine, acting through P2 purinoreceptors, depolarised supraoptic neurones via the activation of postsynaptic non-selective cationic channels (Hiruma & Bourque, 1995). However, since we detected no inward current or changes in input resistance during the 1 Hz stimulation, it is unlikely that such a mechanism occurred in our experiments.

As for the delay in the depression induced during 1 Hz stimulus trains, it may reflect (i) a build up of adenosine that would eventually reach a sufficient concentration to activate adenosine receptors, (ii) a period of diffusion between the site of release and the site of action, and/or (iii) the time necessary for breakdown of ATP into adenosine.

Functional role of adenosine in SON neurones

Glutamate and GABA are the major excitatory and inhibitory neurotransmitters governing the hypothalamo-neurohypophysial system (Theodosis et al. 1986; Van den Pol et al. 1990; El Madjoubi et al. 1997). Thus, depression of either GABA or glutamate release will have significant functional consequences on the electrical activity of supraoptic neurones and, therefore, on neurohypophysial secretion. Since stimulation of GABAergic and glutamatergic inputs evoked an activity-dependent release of adenosine, adenosine may be physiologically relevant whenever afferent synaptic activity is high. On the one hand, GABAergic activity would be limited by adenosine, thus preventing too strong an inhibition of magnocellular cells, and ensuring a minimal level of secretory activity. On the other hand, adenosine released during increased glutamatergic activity, as for example during suckling (Jourdain et al. 1998) or hyperosmotic challenges (Richard & Bourque, 1995), could prevent magnocellular neuronal activity from increasing too much. Such a mechanism could account, at least in part, for a noteworthy feature of the hypothalamo-neurohypophysial system, namely that its patterns of activity are optimally adjusted to maximize secretory output (Dutton & Dyball, 1979; Poulain & Wakerley, 1982). Thus strong continuous firing leads rapidly to secretory ‘fatigue’, whereas periodic activity made up of a brief burst of action potentials is highly efficient (Bicknell, 1988; Bourque, 1990). By its presynaptic inhibitory action, therefore, adenosine could play a role in maintaining action potential discharge at an optimum rate for excitation-secretion coupling in the neurohypophysis.

Acknowledgments

We thank Dr D. T. Theodosis for her helpful comments on the manuscript. This work was partly supported by a grant from the Conseil Régional d'Aquitaine (970301209).

References

  1. Anderson WA, Bruni JE, Kaufman A. Afferent connections of rat's supraoptic nucleus. Brain Research Bulletin. 1990;24:191–200. doi: 10.1016/0361-9230(90)90205-e. [DOI] [PubMed] [Google Scholar]
  2. Augustine GJ, Charlton MP, Smith SJ. Calcium action in synaptic transmitter release. Annual Review of Neuroscience. 1987;10:633–693. doi: 10.1146/annurev.ne.10.030187.003221. [DOI] [PubMed] [Google Scholar]
  3. Aunis D. Exocytosis in chromaffin cells of the adrenal medulla. International Review of Cytology. 1998;181:213–320. doi: 10.1016/s0074-7696(08)60419-2. [DOI] [PubMed] [Google Scholar]
  4. Bicknell RJ. Optimizing release from peptide hormone secretory nerve terminals. Journal of Experimental Biology. 1988;139:51–65. doi: 10.1242/jeb.139.1.51. [DOI] [PubMed] [Google Scholar]
  5. Bourque CW. Intraterminal recordings from the rat neurohypophysis in vitro. The Journal of Physiology. 1990;421:247–262. doi: 10.1113/jphysiol.1990.sp017943. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Brussard AB, Devay P, Leyting-Vermeulen JL, Kits KS. Changes in properties and neurosteroid regulation of GABAergic synapses in the supraoptic nucleus during the mammalian female reproductive cycle. The Journal of Physiology. 1999;516:513–524. doi: 10.1111/j.1469-7793.1999.0513v.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chen G, Van den Pol AN. Adenosine modulation of calcium currents and presynaptic inhibition of GABA release in suprachiasmatic and arcuate nucleus neurones. Journal of Neurophysiology. 1997;77:3035–3047. doi: 10.1152/jn.1997.77.6.3035. [DOI] [PubMed] [Google Scholar]
  8. Day TA, Sibbald JR, Khanna S. ATP mediates an excitatory noradrenergic neuron input to supraoptic vasopressin cells. Brain Research. 1993;607:341–344. doi: 10.1016/0006-8993(93)91528-z. [DOI] [PubMed] [Google Scholar]
  9. Decavel C, Van den Pol AN. GABA: a dominant neurotransmitter in the hypothalamus. Journal of Comparative Neurology. 1990;302:1019–1037. doi: 10.1002/cne.903020423. [DOI] [PubMed] [Google Scholar]
  10. Dutton A, Dyball REJ. Phasic firing enhances vasopressin release from the neurohypophysis. The Journal of Physiology. 1979;290:433–440. doi: 10.1113/jphysiol.1979.sp012781. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. El Majdoubi M, Poulain DA, Theodosis DT. The glutamatergic innervation of oxytocin- and vasopressin-secreting neurones in the rat supraoptic nucleus and its contribution to lactation-induced synaptic plasticity. European Journal of Neuroscience. 1996;8:1377–1389. doi: 10.1111/j.1460-9568.1996.tb01600.x. [DOI] [PubMed] [Google Scholar]
  12. El Majdoubi M, Poulain DA, Theodosis DT. Lactation-induced plasticity in the supraoptic nucleus augments axodendritic and axosomatic GABAergic and glutamatergic synapses: an ultrastructural analysis using the disector method. Neuroscience. 1997;80:1137–1147. doi: 10.1016/s0306-4522(97)00193-0. [DOI] [PubMed] [Google Scholar]
  13. Gies U, Theodosis DT. Synaptic plasticity in the rat supraoptic nucleus during lactation involves GABA innervation and oxytocin neurones: a quantitative immunocytochemical analysis. Journal of Neuroscience. 1994;14:2861–2869. doi: 10.1523/JNEUROSCI.14-05-02861.1994. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Greene RW, Haas HL. The electrophysiology of adenosine in the mammalian central nervous system. Progress in Neurobiology. 1991;36:329–341. doi: 10.1016/0301-0082(91)90005-l. [DOI] [PubMed] [Google Scholar]
  15. Gribkoff VK, Dudek FE. Effects of excitatory amino acid antagonists on synaptic responses of rat supraoptic neurones in slices of rat hypothalamus. Journal of Neurophysiology. 1990;63:60–71. doi: 10.1152/jn.1990.63.1.60. [DOI] [PubMed] [Google Scholar]
  16. Hiruma H, Bourque CW. P2 purinoceptor-mediated depolarization of rat supraoptic neuroscretory cells in vitro. The Journal of Physiology. 1995;489:805–811. doi: 10.1113/jphysiol.1995.sp021093. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Hu B, Bourque CW. Functional N-methyl-D-aspartate and non-N-methyl-D-aspartate receptors are expressed by rat supraoptic neuroscretory cells in vitro. Journal of Neuroendocrinology. 1991;3:509–514. doi: 10.1111/j.1365-2826.1991.tb00311.x. [DOI] [PubMed] [Google Scholar]
  18. Hu HZ, Li ZW. Modulation by adenosine of GABA-activated current in rat dorsal root ganglion neurones. The Journal of Physiology. 1997;501:67–75. doi: 10.1111/j.1469-7793.1997.067bo.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Inenaga K, Honda E, Hirakawa T, Nakamura S, Yamashita H. Glutamatergic synaptic inputs to mouse supraoptic neurones in calcium-free medium in vitro. Journal of Neuroendocrinology. 1998;10:1–7. doi: 10.1046/j.1365-2826.1998.00662.x. [DOI] [PubMed] [Google Scholar]
  20. Jacobson KA. Adenosine A3 receptors: novel ligands and paradoxical effects. Trends in Pharmacological Sciences. 1998;19:184–191. doi: 10.1016/s0165-6147(98)01203-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Jourdain P, Israel J-M, Dupouy B, Oliet SHR, Allard M, Vitiello S, Theodosis DT, Poulain DA. Evidence for a hypothalamic oxytocin-sensitive pattern-generating network governing oxytocin neurones in vitro. Journal of Neuroscience. 1998;18:6641–6649. doi: 10.1523/JNEUROSCI.18-17-06641.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kabashima N, Shibuya I, Ibrahim N, Ueta Y, Yamashita H. Inhibition of spontaneous EPSCs and IPSCs by presynaptic GABAB receptors on rat supraoptic magnocellular neurones. The Journal of Physiology. 1997;504:113–126. doi: 10.1111/j.1469-7793.1997.113bf.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kombian SB, Zidichouski A, Pittman QJ. GABAB receptors presynaptically modulate excitatory synaptic transmission in the rat supraoptic nucleus in vitro. Journal of Neurophysiology. 1996;76:1166–1179. doi: 10.1152/jn.1996.76.2.1166. [DOI] [PubMed] [Google Scholar]
  24. Lambert NA, Teyler TJ. Adenosine depresses excitatory but not fast inhibitory synaptic transmission in area CA1 of the rat hippocampus. Neuroscience Letters. 1991;122:50–52. doi: 10.1016/0304-3940(91)90190-5. [DOI] [PubMed] [Google Scholar]
  25. Li J, Perl ER. Adenosine inhibition of synaptic transmission in the substantia gelatinosa. Journal of Neurophysiology. 1994;72:1611–1621. doi: 10.1152/jn.1994.72.4.1611. [DOI] [PubMed] [Google Scholar]
  26. Ludwig M. Dendritic release of vasopressin and oxytocin. Journal of Neuroendocrinology. 1998;10:881–895. doi: 10.1046/j.1365-2826.1998.00279.x. [DOI] [PubMed] [Google Scholar]
  27. Mahan LC, McVittie LD, Smyk-Randall EM, Nakata H, Monsma FJ, Geren CR, Sibley DR. Cloning and expression of an A1-adenosine receptor from rat brain. Molecular Pharmacology. 1991;40:1–7. [PubMed] [Google Scholar]
  28. Malenka RC, Nicoll RA. NMDA-receptor-dependent synaptic plasticity: multiple forms and mechanisms. Trends in Neurosciences. 1993;16:141–147. doi: 10.1016/0166-2236(93)90197-t. [DOI] [PubMed] [Google Scholar]
  29. Manzoni OJ, Manabe T, Nicoll RA. Release of adenosine by activation of NMDA receptors in the hippocampus. Science. 1994;265:2098–2101. doi: 10.1126/science.7916485. [DOI] [PubMed] [Google Scholar]
  30. Mitchell JB, Lupica CR, Dunwiddie TV. Activity-dependent release of endogenous adenosine modulates synaptic responses in the rat hippocampus. Journal of Neuroscience. 1993;13:3439–3447. doi: 10.1523/JNEUROSCI.13-08-03439.1993. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Mouginot D, Kombian SB, Pittman QJ. Activation of presynaptic GABAB receptors inhibits IPSCs in rat magnocellular neurones in vitro. Journal of Neurophysiology. 1998;79:1508–1517. doi: 10.1152/jn.1998.79.3.1508. [DOI] [PubMed] [Google Scholar]
  32. Noguchi J, Shibuya I, Tanaka N, Kabashima N, Ueta Y, Yamashita H. Adenosine inhibits N- and P/Q-type voltage-dependent Ca2+ channels in rat supraoptic neurones. Folia Endocrinologica Japonica. 1998;74:439. [Google Scholar]
  33. Obrietan K, Belousov AB, Heller HC, Van den Pol AN. Adenosine pre- and postsynaptic modulation of glutamate-dependent calcium activity in hypothalamic neurones. Journal of Neurophysiology. 1995;74:2150–2162. doi: 10.1152/jn.1995.74.5.2150. [DOI] [PubMed] [Google Scholar]
  34. Poulain DA, Theodosis DT. Electrical activity of peptidergic neurones and its relation to hormone release. In: Pickering BT, Wakerley JB, Summerlee AJS, editors. Neurosecretion: Cellular Aspects of the Production and Release of Neuropeptides. New York: Plenum Press; 1988. pp. 181–189. [Google Scholar]
  35. Poulain DA, Wakerley JB. Electrophysiology of hypothalamic magnocellular neurones secreting oxytocin and vasopressin. Neuroscience. 1982;7:773–808. doi: 10.1016/0306-4522(82)90044-6. [DOI] [PubMed] [Google Scholar]
  36. Proctor W, Dunwiddie T. Pre- and postsynaptic actions of adenosine in the in vitro rat hippocampus. Brain Research. 1987;426:187–190. doi: 10.1016/0006-8993(87)90441-0. [DOI] [PubMed] [Google Scholar]
  37. Randle JCR, Bourque CW, Renaud LP. Characterization of spontaneous and evoked inhibitory postsynaptic potentials in rat supraoptic neurosecretory neurones in vitro. Journal of Neurophysiology. 1986;56:1703–1717. doi: 10.1152/jn.1986.56.6.1703. [DOI] [PubMed] [Google Scholar]
  38. Randle JCR, Renaud LP. Action of gamma-aminobutyric acid on rat supraoptic nucleus neurosceretory neurones in vitro. The Journal of Physiology. 1987;56:1703–1717. doi: 10.1113/jphysiol.1987.sp016592. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Redman S. Quantal analysis of synaptic potentials in neurones of the central nervous system. Physiological Reviews. 1990;70:165–198. doi: 10.1152/physrev.1990.70.1.165. [DOI] [PubMed] [Google Scholar]
  40. Reppert SM, Weaver DR, Stehle JH, Rivkees SA. Molecular cloning and characterization of a rat A1-adenosine receptor that is widely expressed in brain and spinal cord. Molecular Endocrinology. 1991;5:1037–1048. doi: 10.1210/mend-5-8-1037. [DOI] [PubMed] [Google Scholar]
  41. Richard D, Bourque CW. Synaptic control of rat supraoptic neurones during osmotic stimulation of the organum vasculosum lamina terminalis in vitro. The Journal of Physiology. 1995;489:567–577. doi: 10.1113/jphysiol.1995.sp021073. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Sawchenko PE, Swanson LW. The organization of noradrenergic pathways from the brainstem to the paraventricular and supraoptic nuclei in the rat. Brain Research Reviews. 1982;4:275–325. doi: 10.1016/0165-0173(82)90010-8. [DOI] [PubMed] [Google Scholar]
  43. Scanziani M, Capogna M, Gähwiler BH, Thompson SM. Presynaptic inhibition of miniature excitatory synaptic currents by baclofen and adenosine in the hippocampus. Neuron. 1992;9:919–927. doi: 10.1016/0896-6273(92)90244-8. [DOI] [PubMed] [Google Scholar]
  44. Shen KZ, Johnson SW. Presynaptic GABAB and A1-adenosine receptors regulate synaptic transmission to rat substantia nigra reticulata neurones. The Journal of Physiology. 1997;505:153–163. doi: 10.1111/j.1469-7793.1997.153bc.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Theodosis DT, Paut L, Tappaz ML. Immunocytochemical analysis of the GABAergic innervation of oxytocin- and vasopressin-secreting neurones in the rat supraoptic nucleus. Neuroscience. 1986;19:207–222. doi: 10.1016/0306-4522(86)90016-3. [DOI] [PubMed] [Google Scholar]
  46. Ulrich D, Huguenard JR. Purinergic inhibition of GABA and glutamate release in the thalamus: implications for thalamic network activity. Neuron. 1995;15:909–918. doi: 10.1016/0896-6273(95)90181-7. [DOI] [PubMed] [Google Scholar]
  47. Van Calker D, Mueller MJ, Hamprecht B. Adenosine regulates via two different types of receptors, the accumulation of cyclic AMP in cultured brain cells. Journal of Neuroscience. 1979;33:999–1005. doi: 10.1111/j.1471-4159.1979.tb05236.x. [DOI] [PubMed] [Google Scholar]
  48. Van den Pol AN, Wuarin JP, Dudek FE. Glutamate, the dominant excitatory transmitter in neuroendocrine regulation. Science. 1990;250:1276–1278. doi: 10.1126/science.1978759. [DOI] [PubMed] [Google Scholar]
  49. White TD, Hoehn K. Release of adenosine and ATP from nervous tissue. In: Stone WT, editor. Adenosine in the Nervous System. London: Academic Press; 1991. pp. 173–195. [Google Scholar]
  50. Zimmermann H. Signalling via ATP in the nervous system. Trends in Neurosciences. 1994;17:420–426. doi: 10.1016/0166-2236(94)90016-7. [DOI] [PubMed] [Google Scholar]

Articles from The Journal of Physiology are provided here courtesy of The Physiological Society

RESOURCES