The Journal of Neurophysiology Vol. 81 No. 1 January 1999, pp. 247-255
Copyright ©1999 by the American Physiological Society

A₁ Adenosine Receptors Modulate Respiratory Activity of the Neonatal Mouse Via the cAMP-Mediated Signaling Pathway

S. L. Mironov, K. Langohr, and D. W. Richter

II Department of Physiology, University of Göttingen, 37073 Göttingen, Germany

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Mironov, S. L., K. Langohr, and D. W. Richter. A₁ adenosine receptors modulate respiratory activity of the neonatal mouse via the cAMP-mediated signaling pathway. J. Neurophysiol. 81: 247-255, 1999. The effects of adenosine and its analogs on the function of the respiratory center were studied in the spontaneously active rhythmic slice of neonatal and juvenile mice (4-14 days old). Whole cell, spontaneous postsynaptic currents (sPSCs) and single channel K_ATP currents were recorded in inspiratory neurons of the pre-Bötzinger complex. Adenosine (50-600 µM) inhibited the respiratory rhythm. This was accompanied by increase in the activity of K_ATP channels in cell-attached patches. The A₁ adenosine receptor agonist, 2-chloro-N⁶-cyclopentyladenosine (CCPA, 0.3-2 µM), inhibited the respiratory rhythm, sPSCs, and enhanced activity of K_ATP channels. The A₁ adenosine receptor antagonist, 8-cyclopentyl-1,3-dipropylxanthine (DPCPX, 1-3 µM), showed opposite effects and occluded the CCPA actions. Agents specific for A₂ adenosine receptors (CGS 21860 and NECA, both applied at 1-10 µM) were without effect. Elevation of intracellular cAMP concentration ([cAMP]_i) by 8-Br-cAMP (200-500 µM), forskolin (0.5-2 µM), or isobutylmethylxantine (IBMX, 30-90 µM) reinforced the rhythm, whereas NaF (100-800 µM) depressed it. The open probability of single K_ATP channels in cell-attached patches decreased after application of forskolin and increased in the presence of NaF. [cAMP]_i elevation reversed the effects of A₁ receptors both on the respiratory rhythm and K_ATP channels. A₁ receptors and [cAMP]_i modified the hypoxic respiratory response. In the presence of A₁ agonists the duration of hypoxic augmentation shortened, and depression of the respiratory rhythm occurred earlier. Elevation of [cAMP]_i prolonged augmentation and delayed the development of the depression. We conclude that A₁ adenosine receptors modulate the respiratory rhythm via inhibition of intracellular cAMP production and concomitant activation of K_ATP channels.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Adenosine is an ubiquitous neuromodulator that acts through specific membrane receptors: A₁, A₂ (further subdivided into high-affinity A_2a and low-affinity A_2b types) and A₃ receptors (Fredholm 1995; Olsson and Pearson 1990). All these receptors are coupled to G-proteins, which can inhibit or stimulate adenylyl cyclase. Resulting changes in intracellular cAMP levels can cause up- or down-regulation of various ion channels (Gerber and Gähwiler 1994; Kuroda et al. 1976). Modulation of ion channels can be also directly accomplished by , subunits of G-proteins via membrane-delimited pathways (Dolphin et al. 1986; Olsson and Pearson 1990).

In the CNS, adenosine and its analogues depress neuronal firing rates (Dunwiddie 1985; Greene and Haas 1989; Meghji 1991; Olsson and Pearson 1990) by acting postsynaptically through an A₁ receptor-controlled increase of voltage- or Ca²⁺-dependent K⁺ conductances (Gerber and Gähwiller 1994; Greene and Haas 1985). This involves cAMP- and G-protein-mediated signaling pathways (Thompson et al. 1993). Presynaptic terminals also have A₁ receptors, which suppress transmitter release through inhibition of Ca²⁺ currents (Dolphin et al. 1986; Scholz and Miller 1991) and by potentiating K⁺ conductances (Trussell and Jackson 1987).

Adenosine receptors are differently expressed in the brain. High densities of A₁ receptors are found in areas that are essential for respiratory control (Reppert et al. 1991). Several lines of evidence also point to a critical role of adenosine in the hypoxic depression of ventilation. Systemic administration of adenosine receptor agonists depresses respiration in vivo, which is antagonized by theophylline or aminophylline (Eldridge et al. 1985; Ginsborg and Hirst 1972). 8-cyclopentyl-1,3-dipropylxanthine (DPCPX), an antagonist to A₁ adenosine receptors, enhances the respiratory rhythm in cats even during normoxia, revealing tonic activation of A₁ receptors (Schmidt et al. 1995). Activation of A₁ adenosine receptors and hypoxia produces similar depression of synaptic interactions. Because extracellular levels of adenosine are increased during prolonged hypoxia (Lutz 1992; Nagel et al. 1993), it was suggested that adenosine receptors contribute to the failure of synaptic interaction observed during hypoxia (Richter et al. 1991). A₁ receptor antagonists, such as theophylline (Eldridge et al. 1985) and DPCPX (Schmidt et al. 1995), delay such hypoxic depression of respiratory activity.

The aim of the present study was to investigate the mechanisms of the adenosinergic modulation of respiratory neurons in vitro and their possible involvement in the hypoxic reactions of the respiratory network. The use of isolated respiratory center (Smith et al. 1991) allowed us to study the influence of adenosine on the function of the respiratory network that is not contaminated by indirect effects originating from failure of cardiovascular functions, and hence brain stem metabolic supplies, or from peripheral and central chemoreceptors. Based on the current knowledge about the action of adenosine in the CNS, the study was focused on spontaneous synaptic currents and K⁺ channels in inspiratory neurons. It was found that adenosinergic effects on the respiratory rhythm could be mimicked by activation of A₁ receptors and were associated with activation of K_ATP channels. Manipulations of the intracellular cAMP concentration ([cAMP]_i) altered the respiratory rhythm and modulated the open probability of single K_ATP channels in cell-attached patches. Because the effects caused by activation of A₁ receptors were neutralized by [cAMP]_i elevation, we suggest that adenosine's effects on the respiratory rhythm can be explained by changes in [cAMP]_i, which alter the function of K_ATP channels.

	METHODS

Abstract Introduction Methods Results Discussion References

Slice preparations

Mice (NMRI) of both sexes (P4-P14) were anesthetized with ether and decapitated at the C3-C4 spinal level. The brain and upper cervical spinal cord were isolated in ice-cold artificial cerebrospinal fluid (ACSF) that was saturated with carbogen (95% O₂-5% CO₂). Following a transverse cut of the neuroaxis at the level of the inferior colliculus, the cerebellum was removed. The isolated brain stem was glued with cyano-acrylate on an agar block with its rostral end directed upwards. Brain stem slicing was started from the rostral end with the neuroaxis inclined by 20° to the plane of the blade. Such configuration kept most transverse projections from pre-Bötzinger complex (preBötC) to the XII nucleus and their axons to XII rootlets intact (Ramirez et al. 1996). On sectioning the brain stem, the caudal end of the aquaeduct was reached. Thereafter slices were cut 100-200 µm thick until cytoarchitectonic landmarks such as inferior olive, nucleus of the solitary tract, hypoglossal nucleus and nucleus ambiguus were visible while the facial nucleus disappeared, indicating that the rostral boundary of the preBötC was reached. The next cut was made 650-750 µm thick to obtain one slice containing the functional respiratory center. This slice was transferred into the recording chamber, put onto the nylon mesh and fixed with a horseshoe-shaped holder, and the XII rootlet was drawn into a suction electrode. The concentration of extracellular K⁺ in ACSF saturated with carbogen at 29°C was elevated to 8-10 mM within 30-60 min to activate the respiratory network (Smith et al. 1991). Under such conditions regular rhythmic activity that was stable up to 14 h was established.

Electrophysiological recordings

Activity from XII rootlets was recorded with suction electrodes, amplified 5,000-10,000 times, band-pass (0.25-1.5 kHz) filtered, rectified, and integrated (Paynter filter with a time constant of 50-100 ms). Hypoglossal activity was taken as index of the central respiratory rhythm (Smith et al. 1991). Intracellular recordings were obtained from preBötC inspiratory neurons using patch electrodes manufactured from borosilicate glass with filament (GC150F, Clark Instruments, UK). They had tip openings of 1.5-2 µm and resistances of 2-4 M. Intracellular signals were amplified with a patch-clamp amplifier EPC-7 (ESF Friedland, Germany). Membrane currents were filtered at 3 kHz (3 dB), digitized at 5 kHz, and stored for analysis with the use of an IBM-compatible PC. Data analysis was performed with the use of home-written programs in Turbo-Pascal 7.0. Data are presented as means ± SE. Statistical significance was determined by using Student's t-tests. Results were considered significant if P < 0.05.

Respiratory neurons were searched in the area attributed to the preBötC (Smith et al. 1991). A "blind patching" technique (Blanton et al. 1989) was used, and inspiratory neurons were identified by their discharge patterns, showing correlation of spontaneous action potentials with inspiratory output of hypoglossal nerve. In about 80% of trials, gentle sucking led to formation of a gigaseal with a resistance of 3-12 G [mean 6 ± 2 (SE) G, n = 127]. Thereafter the spikes became larger and were seen as large deflections in the cell-attached mode (Fig. 1). After rupturing the patch, the whole cell configuration was obtained, and the discharges were blocked revealing the underlying spontaneous excitatory synaptic currents (Fig. 3)(see also Mironov and Richter 1998; Mironov et al. 1998).

View larger version (22K): [in this window] [in a new window]   FIG. 1. Response of KATP channels and respiratory center to adenosine. Im, top: recordings of the membrane current in a cell-attached patch XII, middle: integrated hypoglossal nerve activity. Patch command potential was 0 mV. Two arrows denoted as "c" and "o1" indicate the closed and 1st open channel levels, respectively. Adenosine was applied at the beginning of the trace. Note correlation of spike discharges in the inspiratory neuron (sharp, vertical deflections) with the inspiratory bursts in hypoglossal nerve. Adenosine inhibited both spike discharges and the inspiratory XII bursts, and in parallel, the channel activity increased. Bottom: representative traces of activity of single KATP channels measured in the presence of adenosine at times indicated.

View larger version (42K): [in this window] [in a new window]   FIG. 3. A1 receptor agonist inhibits inspiratory dicharges. Time courses of the membrane current (Im) and the hypoglossal (XII) nerve activity were recorded after addition of 1 µM CCPA. Note the increase in holding current. The whole cell recordings were made at holding potential of 40 mV with the use of intracellular solution contained K+ gluconate. Spontaneous IPSCs and EPSCs are seen as brief upward and downward deflections from the baseline. Bottom: expanded portions of the currents trace (top, *).

Only data obtained from inspiratory neurons are presented here. In comparison with other cells, e.g., expiratory and tonically active nonrespiratory neurons, the spikes recorded in cell-attached mode were absent in interburst intervals or after rhythmic activity was blocked. Therefore all single channel data were measured during the intervals between inspiratory bursts. The current and patch command potentials in cell-attached patches are presented as inside the cell minus outside, according to conventions for intracellular recordings. The open probability, p_open, was obtained by dividing the mean current by the unitary current and the number of channels. The latter was determined as a maximal number of simultaneous openings seen during a given measurement. These estimates of the number of active channels were also confirmed by binomial analysis. They, however, may not correspond to the total number of channels present in the patch because there were also "sleepy" channels that could be activated by diazoxide (Fig. 5).

View larger version (42K): [in this window] [in a new window]   FIG. 5. KATP channels are potentiated by adenosine. Recordings were made after addition of 20 µM diazoxide (A), 200 µM adenosine (B), and 90 µM glibenclamide (C) to the bath. Horizontal dotted lines indicate the closed (top) and dashed lines indicate several open channel levels. The time after the drug addition is indicated near each trace (0 min corresponds to the control). The patch was held at 0 mV.

The properties of K_ATP channels in inspiratory neurons were described previously (Mironov et al. 1998). In the present study the channels were identified by their conductance, gating pattern, and pharmacology. The channels were chosen for analysis if they had conductance of ~75 pS, and they could be activated by hypoxia. When these requirements were fulfilled, A₁ agonists/antagonists and/or [cAMP]_i-modulating agents were applied. For the pharmacological tests, diazoxide and glibenclamide were used. The ATP-sensitivity of 75 pS-channels was verified in nine inside-out patches as described previously (Mironov et al. 1998). Briefly, in the end of the experiment the cell-attached patch was excised from the cell. To compare the measurements of K_ATP channels in inside-out and in cell-attached patches, K⁺ concentration in a bath was raised to 133 mM by replacing NaCl. This produced symmetric K⁺ distribution and zeroed a K⁺ reversal potential. Addition of submillimolar ATP inhibited K_ATP channels (Fig. 9B). However, because of channel run-down in inside-out patches (Mironov et al. 1998), the reversibility of ATP block could not be established.

View larger version (44K): [in this window] [in a new window]   FIG. 9. cAMP-signaling pathway mediates the action of adenosine. Changes in hypoglossal nerve discharge (A) and activity of KATP channels (B) induced by subsequent additions of 1 µM CCPA and 250 µM 8-Br-cAMP as indicated. Drugs were added at the beginning of each trace. B, left and middle: recordings made in a cell-attached patch at a potential of 0 mV. Right: activity of KATP channels in inside-out patch (holding potential 60 mV). After patch excision, K+ concentration in extracellular solution was raised to 133 mM and 0.5 mM ATP was added. Horizontal dotted lines: closed level (top). Dashed lines: open channel levels. Time in seconds after the drug addition is indicated near each trace (0 s corresponds to the control). Note a reversal of CCPA effects after addition of membrane-permeable cAMP-analogue.

In whole cell measurements, series resistance (R_s), input resistance (R_m), and whole cell capacitance were measured with the use of a 10 mV command depolarization. R_s was 13.4 ± 1.5 M. The mean input resistance was 390 ± 42 M, and the mean capacitance was 32 ± 2 pF. The resting potentials of inspiratory neurons measured in current-clamp and using an intracellular solution containing K⁺-gluconate ranged from 52 to 68 mV (mean 60 ± 4 mV, n = 21). For cell-attached recordings, we therefore assumed that the average membrane potential was 60 mV.

In the experiments in which hypoxia was tested, the oxygen tissue pressure was measured with the use of oxygen-sensitive electrodes (tip diam 10-20 µm) that were connected to a chemical microsensor meter (Diamond Electro-Tech, Ann Arbor, MI). Before measurements were taken, the PO₂-sensitive electrode was calibrated in the perfusion chamber with the use of two artificial cerebrospinal fluid (ACSF) solutions saturated with carbogen or a gas mixture containing 95% N₂-5% CO₂ and 1 mM Na₂S₂O₃ to determine a zero oxygen level. The oxygen-sensitive electrode was placed about 100 µm below the slice surface, within a layer (100 ± 25 µm), where most inspiratory neurons were measured. To prevent loss of dissolved gases, the perfusing solution was delivered through stainless steel tubing. To produce hypoxic conditions in the slice, the bubbling gas mixture was changed from carbogen to 95% N₂-5% CO₂. Fifteen to 20 s after onset of hypoxia, mean PO₂ changed from 200 ± 50 to 5 ± 3 mmHg (n = 31). The responses to hypoxia were readily reproducible and could be performed 20-30 times for a given slice with a complete restoration of respiratory rhyhmic activity, even when O₂ pressure had been reduced for 20-30 min.

Solutions and drugs

ACSF contained (in mM) 128 NaCl, 3 KCl, 1.5 CaCl₂, 1.0 MgSO₄, 21 NaHCO₃, 0.5 NaH₂PO₄, and 30 D-glucose (pH 7.4 was adjusted with NaOH). Solutions with elevated K⁺ (8-10 mM) were obtained by replacing NaCl with KCl. The pipette solution for cell-attached recordings contained (in mM) 125 KCl, 15 NaCl, 2 MgCl₂, 2 ATP, 10 N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES), 1 CaCl₂, 3 1,2-bis-(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid (BAPTA); pH was adjusted to 7.4 with KOH and osmolarity to 285-290 mOsm. For intracellular recordings, KCl was replaced by K⁺ gluconate. All salts used for extra- and intracellular solutions were from Sigma (Deisenhofen, Germany). Adenosine agonists and antagonists, forskolin, IBMX, and 8-Br-cAMP were obtained from Research Biochemical International (Cologne, Germany).

The bath solution was recycled into a 200 ml reservoir equilibrated with carbogen. All drugs were added from corresponding stock solutions, which were prepared freshly in ACSF or dimethyl sulfoxide (DMSO). The residual DMSO concentration was <0.1%, and the vehicle at 0.1% concentration had no effect. Depending on the perfusion rate (ranged from 25 to 45 ml/min), the drugs were delivered to the experimental chamber within 8-12 s. Complete washout of drugs was achieved by the perfusing of 400-500 ml fresh solution with the same K⁺ concentration.

	RESULTS

Abstract Introduction Methods Results Discussion References

A₁ receptors modulate the respiratory rhythm

One to two minutes after the application of adenosine (50-600 µM) to the bath, both the amplitude and the frequency of the hypoglossal rhythm decreased (Fig. 1). Under control conditions, the mean interval between XII inspiratory bursts was 5.0 ± 0.4 s (N = 26). In the presence of adenosine, it increased to 7.2 ± 0.4 s (50 µM, N = 7), 12.1 ± 0.6 s (200 µM, N = 8), and 18.2 ± 0.7 s (500 µM, N = 11), respectively. Here and below, N and n correspond to the number of slices and inspiratory neurons examined, respectively.

To identify the type of receptor responsible for this action, we applied the pharmacological agents, which act on specific adenosine receptor subtypes. Submicromolar concentrations of the A₁ receptor agonist, 2-chloro-N⁶-cyclopentyladenosine (CCPA), mimicked the action of adenosine on the respiratory rhythm (Fig. 2, first trace). In the presence of CCPA, the mean interval between hypoglossal inspiratory bursts was 7.0 ± 0.4 s (0.3 µM, N = 7), 11.2 ± 0.8 s (0.9 µM, N = 8), 16.1 ± 0.8 s (2 µM, N = 7), and 31.2 ± 1.7 s (4 µM, N = 9), respectively, compared with the control value of 5.1 ± 0.3 s (N = 21). In the presence of CCPA the amplitude of respiratory bursts decreased to 91 ± 2 (0.3 µM), 51 ± 6 (0.9 µM), 11 ± 3 (2 µM), and 3 ± 1% (4 µM) of control.

View larger version (24K): [in this window] [in a new window]   FIG. 2. Effects of A1 receptor agonists (CCPA) and antagonists (DPCPX) on the hypoglossal nerve discharge. Drugs were applied at the beginning of the trace. Note that to suppress respiratory bursts in the presence of DPCPX, CCPA concentration had to be raised 60-fold.

The respiratory bursts were augmented by A₁ receptors antagonist, DPCPX. This likely reveals a tonic inhibition of respiratory activity by endogenous adenosine. In the presence of 1 µM DPCPX, the interval between inspiratory bursts decreased from 6.5 ± 0.5 s to 3.5 ± 0.3 s, and their amplitude increased to 145 ± 8% of control (n = 11). In the presence of DPCPX, submicromolar concentrations of CCPA were no longer effective, and only at concentrations >10 µM the rhythm was suppressed again (n = 4, Fig. 2, third trace).

Neither of the agonists for A₂ receptors (Fredholm 1995), 2 - p - (2 - carboxyethyl)phenethylamino - 5' - N - ethylcarboxamidoadenosine (CGS 21860) (A_2a) or 5'-N-ethylcarboxamido-adenosine (NECA) (A_2b), affected the rhythm, nor prevented the inhibitory action of adenosine even when applied at concentrations up to 10 µM (both at n = 6, data not shown).

Changes in spontaneous synaptic currents induced by activation of A₁ receptors

In whole cell mode inspiratory neurons revealed spontaneous inhibitory (sIPSC) and excitatory (sEPSC) postsynaptic currents. They had different voltage-dependency: sEPSCs (and synaptic drives, see below) reversed at 0 ± 3 mV and sIPSCs changed sign at 59 ± 5 mV (n = 23). Thus at a holding potential of 40 mV, spontaneous IPSCs and EPSCs could be observed as brief upward and downward deflections from the baseline (Fig. 3). In inspiratory neurons, sEPSCs are assembled into synaptic drive currents (SDCs), which correlate with inspiratory bursts recorded from XII nerve (Fig. 3). sIPSCs were generated by both GABA_A- and glycine-receptors, as they were inhibited by 10 µM bicuculline and 1 µM strychnine. Either antagonist produced only a partial blockade, however, and only their combined application led to a complete inhibition of sIPSCs (n = 13, data not shown). CNQX (4 µM) abolished both sEPSCs and SDCs, indicating that they were generated by AMPA/kainate receptors (n = 11, data not shown).

As Fig. 3 shows, 1 µM CCPA inhibited all types of synaptic currents, but with different efficacy, corresponding to the following sequence: SDCs > sEPSCs > sISPCs. The changes in sIPSCs and sEPSCs induced by 1 µM CCPA were analyzed at 25 and 75 mV, respectively, because at these potentials the contribution of oppositely directed synaptic currents was minimal. For five inspiratory neurons 5 min after CCPA addition, the mean amplitude and frequency of synaptic currents were 16 ± 3 pA and 6 ± 3 Hz versus 31 ± 2 pA and 25 ± 2 Hz in control (sIPSCs), 12 ± 2 pA and 3 ± 2 Hz versus 32 ± 2 pA and 24 ± 3 Hz (sEPSCs). In the presence of 1 µM CCPA, the mean amplitude of synaptic drives recorded at 50 mV decreased to 7 ± 3 pA from the control value of 37 ± 6 pA.

K_ATP channels are activated by A₁ receptors

Figure 1 shows that bath application of adenosine suppressed both spiking activity of neurons and inspiratory output of the hypoglossal nerve. This was accompanied by a gradual increase in the activity of single K_ATP channels (for their identification see below and Mironov et al. 1998). Adenosine (200 µM) increased the open probability (p_open) from 0.06 ± 0.02 to 0.28 ± 0.12 (n = 9).

CCPA enhanced the activity of the channels, which was additionally increased by hypoxia (Fig. 4). CCPA (1 µM) increased p_open from 0.13 ± 0.07 to 0.28 ± 0.09 and hypoxia increased it further to 0.42 ± 0.12 (n = 11). Addition of adenosine to the bath was no longer effective after CCPA was applied (n = 4, data not shown), which was likely because of occlusion of A₁ receptors.

View larger version (43K): [in this window] [in a new window]   FIG. 4. Channel recordings made in a cell-attached patch show changes in the activity of KATP channels induced by A1 receptor agonist (CCPA) and subsequent hypoxia. Left: I-V relationships were obtained using a ramp protocol in control (A), 5 min after application of 1 µM CCPA (B), and 4 min after subsequent hypoxia (C). Patch command voltage is indicated on the x-axis. Values given do not take into account the resting potential (60 mV, as measured after establishing the whole cell configuration). Right: histograms of open time distribution. They were obtained for cell-attached patch at holding potential of 0 mV and were not corrected for missed events. Time distributions were approximated by single exponentials with time constants given in graphs. Mean open time is equal to the time constant of the exponent. For 5 inspiratory neurons its mean values were 1.28 ± 0.06 ms (control), 1.29 ± 0.07 ms (1 µM CCPA), 1.31 ± 0.06 ms (hypoxia).

As Fig. 4 shows, in the presence of CCPA and during hypoxia the channel conductance (75 pS for this patch) did not change. Its mean value was 75 ± 5 pS (n = 34). Therefore changes in channel activity must involve modification of the gating mechanism. Because of the presence of several channels in cell-attached patches, only the open times could be quantified. Their distribution was monoexponential (Mironov et al. 1998). CCPA and hypoxia did not change the mean open time (Fig. 4), therefore the observed effects in p_open stem from changes in the closed time distribution.

Bath application of specific activator of K_ATP channels, diazoxide, increased the channel activity and recruited additional (sleepy) channels (Fig. 5A). Subsequent addition of 200 µM adenosine to the bath further enhanced the activity of K_ATP channels (Fig. 5B). Diazoxide (20 µM) increased p_open from 0.11 ± 0.06 to 0.34 ± 0.09 and adenosine (200 µM) raised it further to 0.44 ± 0.12 (n = 7). In the experiment shown in Fig. 5C, 90 µM glibenclamide, a blocker of K_ATP channels, inhibited the channel activity. Similar effects were observed for four other cells. For cells pretreated with diazoxide, the activity of K_ATP channels could be further enhanced by 0.5-3 µM CCPA (n = 8, data not shown). In the presence of 1 µM CCPA, p_open increased from 0.23 ± 0.05 to 0.38 ± 0.09 (n = 4).

View larger version (41K): [in this window] [in a new window]   FIG. 6. Changes in intracellular cAMP levels modulate the respiratory rhythm recorded from hypoglossal (XII) nerve. A: drug application starts at the beginning of each trace. B: potentiation of synaptic activity in an inspiratory neuron by forskolin. The whole cell membrane current was measured at holding potential of 45 mV. Time in minutes is indicated near each trace (0 min corresponds to the control).

Adenosine and intracellular cAMP

As mentioned in the INTRODUCTION, adenosine actions can be mediated by intracellular cAMP. It was found that all modulations of intracellular cAMP concentration altered the respiratory rhythm (Fig. 6A). [cAMP]_i increase induced by forskolin (0.5-2 µM), an activator of adenylyl cyclase, enhanced the amplitude of inspiratory XII bursts, their frequency, and duration. Similar effects were observed after application of IBMX (30-90 µM), an inhibitor of cAMP-phosphodiesterase, and with application of the membrane-permeable analog, 8-Br-cAMP (200-500 µM). NaF (100-800 µM), which activates G_i-proteins at submillimolar concentrations (Blackmore et al. 1985), decreased the frequency and amplitude of inspiratory bursts. Under control conditions, the mean interval between inspiratory bursts was 5.0 ± 0.4 s, and in the presence of cAMP-elevating agents, it decreased to 2.2 ± 0.2 s after forskolin (N = 26), to 2.1 ± 0.2 s after IBMX (N = 17), to 2.4 ± 0.3 s after 8-Br-cAMP (N = 7), or increased to 8.8 ± 0.4 s after NaF (N = 14). In the presence of forskolin, synaptic drives and sPSCs were potentiated (Fig. 6B).

Next, we examined whether K_ATP channels could be manipulated by changing [cAMP]_i levels. Forskolin decreased the channel activity (Fig. 7). In the presence of 1 µM forskolin, p_open decreased from 0.29 ± 0.11 to 0.09 ± 0.05 (n = 11). The channels were further inhibited and finally blocked by glibenclamide (n = 5; Fig. 7). NaF activated K_ATP channels (Fig. 8). Five minutes after NaF was applied to the bath, action potential discharges disappeared and the activity of K_ATP channels increased. These effects of NaF developed slowly and did not show an apparent dependence on drug concentrations that varied from 100 to 800 µM. Neither channel conductance nor its voltage-dependence was affected. NaF increased p_open from 0.07 ± 0.04 to 0.31 ± 0.11 (n = 11).

View larger version (50K): [in this window] [in a new window]   FIG. 7. Sequential inhibition of KATP channels by forskolin (1 µM) and glibenclamide (70 µM). Recordings were made in a cell-attached patch at a potential of 0 mV. Horizontal dotted lines indicate the closed (top) and dashed lines indicate open channel levels. Time (min) after the drug addition is indicated near each trace (0 min corresponds to the control). Glibenclamide was added in the presence of forskolin after the forskolin response had reached a steady-state.

View larger version (30K): [in this window] [in a new window]   FIG. 8. Channel recordings obtained in a cell-attached patch show the effects of NaF. The channel activity was measured at different holding potentials (near each trace) in control (A) and 7 min after addition of 250 µM NaF (B) to the bath. A: recordings were obtained during the inspiratory phase as indicated by the presence of action potentials (vertical lines crossing the baseline). The channel conductance was 74 pS. Note the disappearance of spikes after NaF has been applied.

cAMP elevation reversed the effects caused by activation of A₁ receptors. After addition of 1 µM CCPA, the frequency of hypoglossal nerve discharge decreased (Fig. 9A) and the activity of K_ATP channels increased (Fig. 9B), but application of permeable cAMP analogue, 8-Br-cAMP (250 µM), reversed these effects (N = 4). Figure 9B also illustrates the inhibition of K_ATP channels by ATP in an inside-out patch excised in the end of experiment.

In this study only the data obtained for inspiratory neurons are presented, but similar effects were observed in expiratory neurons (n = 18). For these cells the analysis was difficult to perform because on-going discharge activity considerably masked the channel activity and distorted the current baseline.

Adenosine and hypoxic response

The response of the respiratory center to hypoxia was biphasic. It consisted of augmentation that started ~1 min after the beginning of the hypoxic episode. An increase in frequency and amplitude of inspiratory bursts was transient and was followed by depression when the rhythm disappeared (Fig. 10). Enhancement of the rate of inspiratory bursts during early hypoxia was superimposed on a slowly developing DC-signal in integrated XII nerve discharge (Fig. 10) caused by enhanced tonic activity of hypoglossal motoneurons (Haddad and Donnelly 1990; Mironov and Richter 1998; Mironov et al. 1998).

View larger version (44K): [in this window] [in a new window]   FIG. 10. Modulation of the hypoxic respiratory response by A1 receptors and intracellular cAMP. Responses to hypoxia, which was applied at the beginning of each trace. Italic lettering: depicts schematically 2 phases of hypoxia. After 4-5 min of each hypoxic episode, the slice recovered 10 min in the presence of oxygen until the respiratory rhythm was restored and remained stable for 5 min. A: recordings were made during control conditions, 5 min after addition of 200 nM CCPA and 5 min after addition 2 µM DPCPX in the presence of CCPA as indicated. B: recordings were made for another slice during control conditions, 5 min after addition of 1 µM forskolin and 5 min after addition 20 µM CCPA in the presence of forskolin as indicated.

The hypoxic reaction was differentially changed by agonists and antagonists of A₁ receptors (Fig. 10A). In the presence of CCPA, augmentation had a shorter duration and respiratory depression developed earlier ( Fig. 10A, second trace). The mean durations of augmentation were 122 ± 21 s during control conditions and 61 ± 32 s in the presence of 200 nM CCPA (N = 6). In addition, the transient DC signal recorded from XII nerve was also diminished, possibly reflecting an A₁ receptor-mediated inhibition of hypoglossal motoneurons. Addition of DPCPX restored and slightly potentated the initial hypoxic response (Fig. 10A, third trace; N = 6). However, 1 µM DPCPX alone was not able to prevent depression, although in its presence the augmentation was prolonged (191 ± 31 s vs. 118 ± 24 s, N = 4).

The effects of forskolin on the respiratory response to hypoxia (Fig. 10B) were opposite to those observed after activation of A₁ receptors. In the presence of forskolin, the augmentation started earlier and lasted longer, resulting in delayed respiratory depression. CCPA restored the initial form of the hypoxic respiratory response. It should be noted that, at submicromolar concentrations (see Fig. 2), CCPA was no longer effective, and to change both the respiratory rhythm and its response to hypoxia in the presence of forskolin, the dose of CCPA had to be increased by about 70-fold (Fig. 10B, third trace; N = 7). Similar results were obtained with 10-50 µM IBMX (N = 12, data not shown).

In the presence of drugs, specific to K_ATP channels, both the XII nerve activity and its hypoxic reaction were modified. Diazoxide (20 µM) increased the interval between inspiratory bursts from 5.4 ± 0.3 s to 8.0 ± 0.4 s, and the duration of augmentation shortened from 112 ± 18 s to 71 ± 12 s (N = 4). In the presence of 70 µM glibenclamide, the interval between inspiratory bursts decreased from 5.3 ± 0.4 s to 4.0 ± 0.4 s, and the duration of augmentation increased from 102 ± 18 s to 174 ± 32 s (N = 5). The inability to glibenclamide to abolish respiratory depression could be explained by loss of sensitivity of K_ATP channels to glibenclamide during hypoxia (Mukai et al. 1998; Venkatesh et al. 1991) and/or contribution of other factor(s) (Hochachka et al. 1996) to respiratory depression.

	DISCUSSION

Abstract Introduction Methods Results Discussion References

Adenosine is a purinergic metabolite and a potent neuromodulator. Its various actions within the CNS were widely reviewed (Dunwiddie 1985; Fredholm 1995; Greene and Haas 1989; Meghji 1991; Olsson and Pearson 1990; von Lubitz et al. 1995). Adenosine receptors are subdivided into four types, which differ by their affinity to adenosine, pharmacology, and coupling to intracellular signaling pathways (Fredholm 1995; Olsson and Pearson 1990). All of them activate G-proteins, one target of which is the adenylyl cyclase, which is suppressed by A₁ receptors but is stimulated by A₂ and A₃ receptors. Resulting changes in [cAMP]_i can modify the function of various ion channels and receptors (Anholt 1994; Levitan 1988).

Role of A₁ receptors and cAMP in rhythmogenesis

It was shown (Schmidt et al. 1995) that A₁ receptors are functional in ventral respiratory neurons of the in vivo cat. The present study extends these findings in demonstrating that in neonatal mice A₁ receptors modulate the respiratory rhythm and its reaction to hypoxia. Elevation of [cAMP]_i had opposite effects, and the respiratory activity was changed by agents that alter intracellular cAMP levels either directly (8-Br-cAMP) or indirectly (forskolin, IBMX, and NaF). The effects of all substances tested in the present study are consistent with their presumed action on cAMP-signaling pathway. Similar effects of [cAMP]_i on respiratory rhythm generation were observed in brain stem-spinal cord preparation from newborn rats (Arata et al. 1993). These findings also corroborate the results obtained in the cat in vivo by Lalley et al. (1997), who demostrated modulation of discharge patterns of respiratory neurons by protein kinase A.

A₁ receptors and [cAMP]_i modulate K_ATP channels

A₁ receptors activated K_ATP channels and the elevation of intracellular cAMP diminished their activity. The properties of K_ATP channels in inspiratory neurons were described previously (Mironov et al. 1998). These neurons have conductance of 75 pS, p_open = 0.1-0.2 and t_op = 1-2 ms; therefore they can be attributed to a class of large conductance K_ATP channels (Quayle et al. 1997). In cell-attached patches the channels were activated by hypoxia (Fig. 4) and diazoxide (Fig. 5) and were inhibited by glibenclamide (Figs. 5 and 7). In inside-out patches the channels were inhibited by ATP (Fig. 9B). Interestingly in the cell line INS-1, K_ATP channels (conductance of 50-70 pS) were blocked by tolbutamide and inhibited by forskolin, which led to enhanced insulin secretion (Ullrich et al. 1996).

The changes in the respiratory rhythm and the activity of K_ATP channels induced by A₁ receptors and [cAMP]_i had similar kinetics and dependence on agonist concentration. Potentiation of K_ATP currents by adenosine likely involves a G_i-protein-mediated suppression of intracellular cAMP production. Recently it was shown that inspiratory motor discharges are suppressed by activation of G-protein-coupled receptors (Johnson et al. 1996). We suggest that a pathway that includes the activation of A₁ receptors, concomitant [cAMP]_i decrease leading to activation of K_ATP channels, has particular physiological significance in inspiratory neurons, whereby respiratory rhythm and its hypoxic response is modified.

Role of A₁ receptors and cAMP in hypoxia

Extracellular adenosine originates from two sources (Meghji 1991). ATP, co-released from presynaptic terminals together with other neurotransmitters, is metabolized to adenosine by ectonucleotidases. Potentiation of the rhythmic activity by DPCPX (Fig. 2) indicates that the basal levels of adenosine are high enough to produce a tonic activation of A₁ receptors in neurons that participate in the control of respiratory rhythm. Another source of adenosine predominates during hypoxic conditions or intense neuronal activity and involves the production of adenosine from the metabolic breakdown of ATP. It rapidly crosses the membrane via a bidirectional transporter and acts extracellularly. K_ATP channels are necessary for respiratory depression (Mironov et al. 1998). In inspiratory neurons, K_ATP channels were activated by A₁ receptors, but hypoxia was able to further potentiate them. Although A₁ receptors contribute to the development of respiratory depression during hypoxia, their activation is not sufficient. For example, in the presence of DPCPX the depression was only delayed but not prevented.

	ACKNOWLEDGEMENTS

  The authors thank H. Timmermann for technical assistance.

  The study was supported by Sonderforschungbereich 406 (TP C7).

	FOOTNOTES

  Address for reprint requests: S. L. Mironov, Dept. of Physiology, University of Göttingen, Humboldtallee 23, 37073 Göttingen, Germany.

  Received 19 February 1998; accepted in final form 16 September 1998.

	REFERENCES

Abstract Introduction Methods Results Discussion References

• ANHOLT, R. R. Signal integration in the nervous system: adenylate cyclases as molecular coincidence detectors. Trends Neurosci. 17: 37-41, 1994.[Medline]
• ARATA, A., ONIMARU, H., HOMMA, I. Effects of cAMP on respiratory rhythm generation in brainstem-spinal cord preparation from newborn rat. Brain Res. 605: 193-199, 1993.[Medline]
• BLACKMORE, P. F., BOCCINO, S. B., WAYNICK, L. E., EXTON, J. H. Role of guanine nucleotide-binding regulatory protein in the hydrolysis of hepatocyte phoshatidylinositol 4,5-biphosphate by calcium-mobilizing hormones and the control of cell calcium: studies utilizing aliminium fluoride. J. Biol. Chem. 260: 14477-14483, 1985.[Abstract/Free Full Text]
• BLANTON, M. C., LO TURCO, J. J., KRIEGSTEIN, A. J. Whole-cell recordings from neurons in slices of reptilian and mammalian cerebral cortex. J. Neurosci. Methods 30: 203-210, 1989.[Medline]
• DOLPHIN, A. C., FORDA, S. R., SCOTT, R. H. Calcium-dependent currents in cultured rat dorsal root ganglion neurones are inhibited by an adenosine analogue. J. Physiol. (Lond.) 373: 47-61, 1986.[Abstract/Free Full Text]
• DUNWIDDIE, T. V. The physiological role of adenosine in the central nervous system. Int. Rev. Neurobiol. 27: 63-139, 1985.[Medline]
• ELDRIDGE, F. L., MILLHORN, D. E., KILEY, J. P. Antagonism by theophylline of respiratory inhibition induced by adenosine. J. Appl. Physiol. 59: 1428-1433, 1985.[Abstract/Free Full Text]
• FREDHOLM, B. B. Adenosine receptors in the central nervous system. News Physiol. Sci. 10: 261-266, 1995.
• GERBER, U., GÄHWILER, B. GABA_B and adenosine receptor mediate enhancement of the K⁺ current, I_AHP, by reducing adenylyl cyclase activity in rat CA3 hippocampal neurons. J. Neurophysiol. 72: 2360-2367, 1994.[Abstract/Free Full Text]
• GILMAN, A. G proteins: transducers of receptor-generated signals. Annu. Rev. Biochem. 56: 615-649, 1987.[Medline]
• GINSBORG, B. L., HIRST, G.D.S. The effects of adenosine on the release of the transmitter from the phrenic nerve of the rat. J. Physiol. (Lond.) 224: 629-645, 1972.[Abstract/Free Full Text]
• GREENE, R. W., HAAS, H. L. The electrophysiology of adenosine in the mammalian central nervous system. Prog. Neurobiol. 36: 329-341, 1989.
• HADDAD, G. G., DONNELLY, D. F. O₂ deprivation induces a major depolarization in brain stem neurons in the adult but not in the neonatal rat. J. Physiol. (Lond.) 429: 411-428, 1990.[Abstract/Free Full Text]
• HOCHACHKA, P. W., BUCK, L. T., DOLL, C. J., LAND, S. C. Unifying theory of hypoxia tolerance: molecular/metabolic defense and rescue mechanisms for surviving oxygen lack. Proc. Natl. Acad. Sci. USA 93: 9493-9498, 1996.[Abstract/Free Full Text]
• JOHNSON, S. M., SMITH, J. C., FELDMAN, J. L. Modulation of respiratory rhythm in vitro: role of G_i/o protein-mediated mechanisms. J. Appl. Physiol. 80: 2120-2133, 1996.[Abstract/Free Full Text]
• KURODA, Y., SAITO, M., KOBAYASHI, K. Concomitant changes in cyclic AMP level and postsynaptic potentials of olfactory cortex slices induced by adenosine derivatives. Brain Res. 109: 196-201, 1976.[Medline]
• LALLEY, P. M., PIERREFICHE, O., BISCHOF, A. M., RICHTER, D. W. cAMP-dependent proteinkinase modulates expiratory neurons in vivo. J. Neurophysiol. 77: 1119-1131, 1997.[Abstract/Free Full Text]
• LEVITAN, I. B. Modulation of ion channels in neurons and other cells. Annu. Rev. Neurosci. 11: 119-136, 1988.[Medline]
• VON LUBITZ, D.K.J.E., CARTER, M. F., BEENHAKKER, M., LIN, R.C.S., JACOBSON, K. A. Adenosine, a proterapeutic concept in neurodegeneration. Ann. NY Acad. Sci. 675: 163-178, 1995.
• LUTZ, P. L. Mechanisms for hypoxic survival in the vertebrate brain. Annu. Rev. Physiol. 54: 601-618, 1992.[Medline]
• MEGHJI, P. Adenosine production and transport. In: Adenosine in the Nervous System, edited by and T. W. Stone San Diego, CA: Academic, 1991, p. 281-303.
• MIRONOV, S. L., RICHTER, D. W. L-type Ca²⁺ channels in inspiratory neurones and their modulation by hypoxia. J. Physiol. (Lond.). 512: 75-87, 1998.[Abstract/Free Full Text]
• MIRONOV, S. L., LANGOHR, K., HALLER, M., RICHTER, D. W. Hypoxia activates ATP-dependent potassium channels in inspiratory neurones of neonatal mice. J. Physiol. (Lond.) 509: 755-766, 1998.[Abstract/Free Full Text]
• MUKAI, E., ISHIDA, H., KATO, S., TSUURA, Y., FUJIMOTO, S., ISHIDA-TAKAHASHI, A., HORIE, M., TSUDA, K., SEINO, Y. Metabolic inhibition impairs ATP-sensitive K+ channel block by sulfonylurea in pancreatic beta-cells. Am. J. Physiol. 274: E38-E44, 1998.[Abstract/Free Full Text]
• NAGEL, H., SCHMIDT-GARCON, R., RICHTER, D. W. Extracellular adenosine levels within the respiratory network during sustained hypoxia. XXXII Congress IUPS 141: 21P, 1993.
• OLSSON, R. A., PEARSON, J. D. Cardiovascular purinoreceptors. Physiol. Rev. 70: 761-845, 1990.[Free Full Text]
• QUAYLE, J. M., NELSON, M., STANDEN, N. B. ATP-sensitive and inwardly rectifying potassium currents in smooth muscle. Physiol. Rev. 77: 1165-1216, 1997.[Abstract/Free Full Text]
• RAMIREZ, J. M., QUELLMALZ, U.J.A., RICHTER, D. W. Postnatal changes in the mammalian respiratory network as revealed by the transverse brain stem slice of mice. J. Physiol. (Lond.) 491: 799-812, 1996.[Medline]
• REPPERT, S. M., WEAVER, D. R., STEHLE, J. H., RIVKEES, S. A. Molecular cloning and characterization of a rat A₁-adenosine receptor that is widely expressed in brain and spinal cord. Mol. Endocrinol. 5: 1037-1048, 1991.[Abstract]
• RICHTER, D. W., BISCHOFF, A. M., ANDERS, K., BELLINGHAM, M., WINDHORST, U. Response of the medullary respiratory network of the cat to hypoxia. J. Physiol. (Lond.) 443: 231-256, 1991.[Abstract/Free Full Text]
• SCHMIDT, C., BELLINGHAM, M. C., RICHTER, D. W. Adenosinergic modulation of respiratory neurones and hypoxic responses in the anaesthezised cat. J. Physiol. (Lond.) 483: 769-781, 1995.[Medline]
• SCHOLZ, K. P., MILLER, R. J. Analysis of adenosine actions on Ca²⁺-currents and synaptic transmission in cultured rat hippocampal pyramidal neurones. J. Physiol. (Lond.) 435: 373-393, 1991.[Abstract/Free Full Text]
• SMITH, J. C., ELLENBERGER, H. H., BALLANYI, K., RICHTER, D. W., FELDMAN, J. L. Pre-Bötzinger complex: a brainstem region that may generate respiratory rhythm in mammals. Science 254: 726-729, 1991.[Abstract/Free Full Text]
• THOMPSON, S. M., CAPOGNA, M., SCANZIANI, M. Presynaptic mechanisms in the hippocampus. Trends Neurosci. 16: 222-227, 1993.[Medline]
• TRUSSELL, L. O., JACKSON, M. B. Dependence of an adenosine-activated potassium current on a GTP-binding protein in mammalian central neurons. J. Neurosci. 7: 3306-3316, 1987.[Abstract]
• ULLRICH, S., ABEL, K. B., LEHR, S., GREGER, R. Effects of glucose, forskolin and tolbutamide on membrane potential and insulin secretion in the insulin-secreting cell line INS-1. Pflügers Arch. 432: 630-636, 1996.[Medline]
• VENKATESH, N., LAMP, S. T., WEISS, J. N. Sulfonylureas, ATP-sensitive K⁺ channels, and cellular K⁺ loss during hypoxia, ischemia, and metabolic inhibition in mammalian ventricle. Circ. Res. 69: 623-637, 1991.[Abstract/Free Full Text]

0022-3077/99 $5.00 Copyright ©1999 The American Physiological Society