Postnatal development of GABAB-receptor-mediated modulation of potassium currents in brainstem respiratory network of mouse

Respiratory Physiology & Neurobiology 158 (2007) 22–29

Postnatal development of GABAB-receptor-mediated modulation of potassium currents in brainstem respiratory network of mouse A. Pfeiffer a , W. Zhang a,b,∗ a

Center of Physiology and Pathophysiology, University of G¨ottingen, Humboldtallee 23, 37073 G¨ottingen, Germany b DFG Research Center for the Molecular Physiology of the Brain (CMPB), Germany Accepted 3 March 2007

Abstract The GABAB -receptor is known to activate a potassium conductance that is inwardly-rectifying, Ba2+ -sensitive and mediated by G-proteincoupled mechanism. The network that generates respiratory rhythm is located in the brainstem and is modulated by GABAB -receptors. The present study investigated the mechanisms by which GABAB -receptor activation modulates respiratory rhythm and how these effects change during the first 2 weeks of postnatal development (P0–P15). Whole-cell patch clamp recordings were obtained from inspiratory neurons in the ventral respiratory column of acute brain stem slice of mouse. In presence of TTX and cadmium, application of baclofen, a GABAB -receptor agonist, activated an inwardly-rectifying potassium current. The reversal potential of the current was around −78 mV, which was close to the calculated equilibrium potential of potassium. The action of baclofen was dose-dependent and could be partially blocked (>85%) by a selective GABAB -receptor antagonist CGP 55845A. The current density of the baclofen-activated potassium currents increased over the first 2 postnatal weeks. At the cellular level, baclofen-activated potassium currents hyperpolarized inspiratory neurons in a concentration- and age-dependent manner. At the network level, the frequency of the respiratory rhythm decreased or was abolished depending on the concentration of baclofen applied. Our results indicate that the endogenous modulation of respiratory rhythm by GABAB -receptors that we have demonstrated previously is mediated at least in part through activation of an inwardly rectifying K+ conductance and that this effect increases postnatally. © 2007 Elsevier B.V. All rights reserved. Keywords: Baclofen; CGP 55845A; Inwardly-rectifying potassium current

1. Introduction In the vertebrate central nervous system, GABA is the major inhibitory neurotransmitter that activates two postsynaptic conductances (Nicoll et al., 1990). One is a chloride conductance activated by GABAA,C -receptors. The second is a potassium conductance activated by GABAB -receptors (Misgeld et al., 1995). The GABAB -receptor-activated potassium conductance is inwardly-rectifying and mediated by G-protein-coupled mechanisms (Gahwiler and Brown, 1985; Newberry and Nicoll, 1985; Thalmann, 1988). In addition, activation of the GABAB receptors evokes a membrane hyperpolarization that inhibits evoked and spontaneous synaptic activity (Dutar and Nicoll, 1988). ∗ Corresponding author at: Centre of Physiology and Pathophysiology, University of G¨ottingen, Humboldtallee 23, 37073 G¨ottingen, Germany. Tel.: +49 551 393767; fax: +49 551 394178. E-mail address: [email protected] (W. Zhang).

1569-9048/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.resp.2007.03.002

Neuronal circuits that generate rhythmic motor patterns such as respiration and locomotion depend on interactions between excitatory and inhibitory synaptic processes (Grillner et al., 1991, 1995; Richter, 1996). The pre-B¨otzinger complex (preB¨otC) in the brainstem contains the kernel of the respiratory rhythm-generating network (Smith et al., 1991). In neonatal animals, GABAB -receptor-mediated modulation of respiratory rhythm is evident in respiratory neurons from the 1st postnatal day (Johnson et al., 1996; Brockhaus and Ballanyi, 1998; Zhang et al., 2002). Activation of GABAB -receptors enhances the low voltage activated Ca2+ -currents (LVA) in early neonates, whereas it decreases the LVA in older animals (Zhang et al., 1999). In addition activation of GABAB -receptors hyperpolarizes the membrane potential of respiratory neurons in neonatal rat (Johnson et al., 1996; Brockhaus and Ballanyi, 1998), whereas application of GABAB -receptor antagonist CGP 55845 significantly modulates the respiratory rhythm (Zhang et al., 2002). Never the less, the postnatal ontogeny of the GABAB receptor-activated potassium conductance in neurons of the

A. Pfeiffer, W. Zhang / Respiratory Physiology & Neurobiology 158 (2007) 22–29

respiratory network during the postnatal development has not been characterized in detail. In the present study, we have investigated developmentally related changes in the GABAB -receptor-activated potassium conductance in inspiratory neurons, and analyzed how the activation of GABAB -receptors modulates the respiratory burst activity in brain stem slice preparation of mice during postnatal development (P0–P15). We found that activation of GABAB -receptor evoked potassium currents already from the 1st postnatal day on. The density of these currents increases with age. Our investigation has also revealed age-dependent effects of GABAB activation on cellular properties of rhythmic neurons.

23

The brain and upper cervical spinal cord were isolated in icecold artificial cerebrospinal fluid (ACSF) which was bubbled with carbogen (95% O2 and 5% CO2 ). The cerebellum and forebrain were removed to expose the brain stem. The brain stem was sectioned serially from rostral to caudal until the nucleus ambiguus and inferior olive were seen. Then a single slice (700 ␮m) containing the ventral respiratory column was cut. It was immediately transferred into the recording chamber and submerged under ACSF (28–30 ◦ C; flow rate 10 ml/min). After a stabilization period of 15 min, the concentration of potassium in ACSF was raised from 3 to 8 mM over a period of 20 min for initiating spontaneous rhythmic activity (Smith et al., 1991).

2. Methods 2.2. Recording and data analysis 2.1. Preparation and general procedures Dissection of the brain stem slice followed the procedures described previously (Zhang et al., 1999). Briefly, male or female mice (NMRI mice: postnatal days P0–P15) were deeply anaesthetized with ether and decapitated at the C3–C4 spinal level.

Rhythmic burst activity was recorded from the hypoglossal nerve rootlets (NXII in Figs. 1 and 2) with a suction electrode, amplified and filtered (high pass, 1.5 kHz; low pass, 250 Hz). The activity was rectified, low-pass filtered  and integrated (Paynter filter, time constant τ = 20–30 ms; NXII in Fig. 2). Whole-

Fig. 1. Effects of GABAB -receptor activation on respiratory rhythm at different postnatal ages. (A and B) Example traces of age-dependent effects of baclofen (1, 5 ␮M) on respiratory burst discharge of hypoglossal nerve rootlets in P2 (A) and P9 (B) slice preparations. (C) Instant frequency of respiratory burst discharge in response to bath application of baclofen (0.5 ␮M) and CGP 55845A (0.1 ␮M) is plotted against time in one individual slice (P8). Note that the effect of baclofen became obvious within less than 30 s. (D) Dose–response curve summarizing the concentration-dependent effects of baclofen in P0–P3 (open circles) and P7–P15 (filled circles) slice preparations. The threshold concentration of baclofen was 250 nM at P0–P3 and 100 nM in P7–P15 preparations. The IC50 was 1 ␮M at P0–P3 and 0.5 ␮M at P7–P15. The numbers of the mice tested is indicated. (E) Bar graph demonstrated the effect of application of 0.5 ␮M baclofen at P0–P3 (n = 15; open circles) and at P7–P15 (n = 15; filled circles) slice preparations. Application of CGP 55845A restored the respiratory frequency to more than 85% of control in all mice.

24

A. Pfeiffer, W. Zhang / Respiratory Physiology & Neurobiology 158 (2007) 22–29

Fig. 2. Effects of GABAB -receptor activation on the membrane potential and discharge properties of inspiratory neurons at different postnatal ages. (A and B) Sample recordings in current-clamp-mode (PBC) illustrating the effects of different baclofen concentrations on membrane potential and properties of rhythmic discharge of inspiratory   neurons in P0 (A) and P7 (B) slice preparations. Lower traces show the corresponding integrals of rhythmic discharges from NXII nerve  rootlets NXII . (C) Recordings of single respiratory burst discharge of inspiratory neurons (upper panels, PBC) and of NXII nerve rootlets (lower panels, NXII) at higher temporal resolution. It shows baclofen-dependent hyperpolarization of inspiratory neurons, which could be partly reversed by adding CGP 55845A (500 nM). (D and E) Summary of the concentration-dependency of baclofen-induced membrane hyperpolarization and decrease of spike discharge of inspiratory neurons in P0–P3 (filled circles, n = 6) and P7–P15 (open circles, n = 5) slice preparations. The changes could be partly reversed by CGP 55845A (500 nM, indicates by CGP).

cell recordings were obtained from the somata of inspiratory or unidentified neurons located in the preB¨otC or B¨otC when the rostral surface of the slice were facing up. Recording electrodes had resistance of 4–6 M and were prepared by pulling borosilicate glass micropipettes (GC150-10F, Clark Electromedical Instruments, England) on a multistage puller (Sutter Instrument Co., P87, tip size ∼2 ␮m). Neurons were approached under visual control using a microscope (Axioscope, Zeiss, Germany) equipped with an infrared contrast enhancement system (C2400, Hamamatsu Photonics, Enfield, Middlesex, UK). After the electrodes contacted neurons, seal resistances of up to 5 G were typically formed. Once a giga-seal was established, the wholecell configuration was obtained by gentle suction and the holding potential was set to −70 mV, unless otherwise stated. At least 80% of serial resistance was compensated. Patch clamp electrodes were connected to an Axopatch 200 amplifier (Axon Instrument Inc., USA). The membrane currents were filtered by a 4-pole Bessel filter set at a cutoff frequency of

2 kHz and digitized at a sampling rate of 5 kHz using a DigiData 1200 interface (Axon Instrument Inc., USA). Commercial software was used for voltage protocols as well as acquisition and analysis of data (pClamp 6.0 and AxoGraph 3.5, Axon Instrument Inc., USA). Neurons showing rhythmic burst activity that was synchronized with the discharge of the NXII rootlet were referred to as inspiratory neurons (cf. Fig. 2), while neurons in the vicinity of the inspiratory neurons that showed no synchronized rhythmic activity are referred to as unidentified neurons. The rhythmic activity of inspiratory neurons was recorded in current-clamp mode (Fig. 2). As cell size change during early postnatal development, we estimated such changes by determining the membrane capacitance of each recorded neuron before subsequent compensation. The membrane capacitance was calculated from the integral of the current transient induced by 10 mV hyperpolarizing voltage commands from a holding potential of −70 mV immediately after rupture of the cell membrane (Gillis, 1995). Based on

A. Pfeiffer, W. Zhang / Respiratory Physiology & Neurobiology 158 (2007) 22–29

the capacitance measurements, current densities (pA/pF) were estimated and compared at different ages. Previous data demonstrated that both GABAB -receptormediated modulation of Ca2+ -currents (Zhang et al., 1999) and

25

synaptic transmission (Zhang et al., 2002) change during the first 2 postnatal weeks, the most dramatic changes occurring at first 4 days (P0–P3) and at 2nd week (P7–P15, cf. Zhang et al., 2002). In the present study we therefore compared changes

Fig. 3. GABAB -receptor-activated inwardly-rectifying potassium currents at different postnatal ages. (A) Sample traces of voltage-activated potassium currents under control (a), in presence of baclofen (b) and CGP (c). The lower trace was obtained by the subtraction of above traces: IBac (d), IBac in the presence of CGP (e) and Ba2+ (f). The currents were evoked by a 200 ms step from a holding potential of −60 mV to various test potentials (see inset). (B) Bar graph shows averaged current density of constitutive currents at different ages as indicated (n = 5–8 for each age group). (C) Sample traces of baclofen-evoked potassium currents (IBac ), obtained by subtracting the constitutive current from the voltage-activated responses in presence of baclofen from neurons of different postnatal ages. (D) Diagram shows I–V-relationships of baclofen-activated currents at different postnatal ages as indicated. (E and F) Bar graphs show averaged current amplitudes (E) and current densities (F) of IBac (20 ␮M baclofen) at different ages as indicated (n = 5–8 for each age group). Note the corresponding grey bars indicate the remaining currents in the presence of CGP 55845A (500 nM). (G) The concentration-response curves summarizing the relative IBac current density in P0–P3 (open circles, n = 8), P4–P6 (open squares, n = 5) and P7–P15 (filled circles, n = 7) slice preparations at different baclofen concentrations as compared to the corresponding constitutive currents. (H) Plot of IBac against membrane potential for the same neuron measured in 5 mM (filled circles), 15 mM (open circles) and 30 mM (open triangles) of extracellular potassium concentrations ([K+ ]o ). (I) The reversal potential of the IBac currents strongly depends on [K+ ]o . (J) Plot of IBac against membrane potential for the same neuron measured in ASCF (filled circles) and in the presence of 500 ␮M Ba2+ .

26

A. Pfeiffer, W. Zhang / Respiratory Physiology & Neurobiology 158 (2007) 22–29

between inspiratory neurons obtained from mice at the postnatal ages of P0–P3, P4–P6 and P7–P15. As the oldest age group (P7–P15) ranges 8 days, the precise composition of the group varied between the experimental series. But, as a basic rule for each experimental series, the oldest age group must contain one or more animals from the subgroups P7–P9, P10–P12, and P13–P15. The voltage-dependent potassium currents were recorded in the presence of 0.5 ␮M TTX and 300 ␮M Cd2+ . The holding potential of −60 mV was stepped for 200 ms to various test potentials (cf. inset in Fig. 3A). The current obtained under control conditions was defined as constitutive potassium current. The baclofen-activated currents (IBac ) were assessed by subtracting the constitutive current under control condition from the voltage-activated K+ currents in presence of baclofen (cf. Fig. 3A). 2.3. Solution and drugs Experiments were carried out in ACSF containing (mM): 118 NaCl, 3 KCl, 1.5 CaCl2 , 1 MgCl2 , 25 NaHCO3 , 1 NaH2 PO4 , 10 glucose equilibrated with carbogen (95% O2 and 5% CO2 ) at 28–30 ◦ C (pH 7.4). Different concentrations of KCl (3–8 mM) were used for different experiments as indicated in text. Patch electrodes were filled with a solution containing (mM): 135 Kgluconate, 1 CaCl2 , 10 EGTA, 2 MgCl2 , 4 Na3 ATP, 0.5 Na3 GTP, 10 HEPES (adjusted to pH 7.3 by KOH). (±)-baclofen (RBI, UK) was prepared as stock solution and added in bath solution as desired. Tetrodotoxin (TTX) was obtained from Alomone Labs (Israel). All other chemicals were obtained from Sigma–Aldrich (USA). 2.4. Statistics Statistical data are expressed as means ± standard error of mean (S.E.M.). Statistical significance of differences between means was assessed with one-way-ANOVA follow by Tukey–Kramer multiple comparison tests as indicated in the text (InStat, GraphPad Software Inc., USA). The level of significance was set at P < 0.05. The n values in the text represent the number of experiments performed. 3. Results 3.1. Activation of GABAB -receptor slows respiratory burst activity in early postnatal stages To investigate the effects of GABAB -receptor-activation in respiratory rhythm modulation, we first measured NXII inspiratory burst activity after application of the GABAB receptor agonist baclofen in mouse brainstem of different postnatal ages (Fig. 1A and B). Various concentrations of baclofen (25 nM to 5 ␮M) were tested. The measurements of steady-state were made after 3 min of incubation for each concentration although the effect became obvious within shorter time (Fig. 1C). After the application of 5 ␮M baclofen, the selective GABAB -receptor antagonist CGP 55845A

(100 nM) was used to antagonize the baclofen-induced effects (Fig. 1E). The presence of baclofen decreased the respiratory burst frequency at all ages tested (Fig. 1, P0–P15). In P0–P3 preparations, the threshold concentration was 250 nM and IC50 was 1 ␮M baclofen (Fig. 1A and D). Burst frequency was reduced by 15 ± 6% at 250 nM, 33 ± 5% at 500 nM, 45 ± 6% at 750 nM and 58 ± 7% at 1 ␮M baclofen (n = 15, Fig. 1D). In the presence of 750 nM and 1 ␮M baclofen the burst frequency ceased in 3 and 7 of 15 animals, respectively, while 5 ␮M baclofen completely abolished the burst activity in all tested animals (n = 15, Fig. 1A). Blockade of GABAB -receptors by 100 nM CGP 55845A in the presence of 5 ␮M baclofen (10 min) consistently restored the burst frequency to 88 ± 10% of control (n = 15, open symbols in Fig. 1E). In P7–P15 neonates, the threshold concentration was 100 nM and IC50 was 500 nM baclofen (Fig. 1B and D). Burst frequency was subsequently reduced by 14 ± 7% at 100 nM, 26 ± 4% at 250 nM, 51 ± 4% at 500 nM and 63 ± 5% at 750 nM baclofen (n = 15, Fig. 1D). In the presence of 750 nM baclofen the burst frequency ceased in 10 of 15 animals, while 1 ␮M or higher completely abolished the burst activity in all tested animals (n = 15, Fig. 1B). In all cases, blockade of GABAB -receptor by adding 100 nM CGP in the presence of 5 ␮M baclofen for 10 min restored the burst frequency to 89 ± 12% values (n = 15, closed symbols in Fig. 1E). In all experiments, partial recovery of the drug effect was obtained within 15 min of washout (n = 29, Fig. 1E). Thus, these experiments indicate that activation of GABAB -receptors modulates respiratory rhythm and the sensitivity to the GABAB receptor agonist baclofen increases with postnatal ages. 3.2. Membrane hyperpolarization contributes to GABAB -receptor-mediated inhibition of respiratory rhythm in early postnatal life The effect of GABAB -receptor-activated potassium current (IBac ) on respiratory rhythm and the effect of GABAB -receptor activation on membrane potential and discharge properties of inspiratory neurons were tested in whole-cell patch-clamp experiments (in current-clamp mode). Under control conditions, the resting membrane potential of rhythmically active inspiratory neurons ranged in between −48 and −60 mV (Fig. 2A, B and D). There was a slight but not significant difference between the membrane potentials of inspiratory neurons in different postnatal ages (n = 10, cf. also to Brockhaus and Ballanyi, 1998; Ritter and Zhang, 2000). In neurons of P0–P3 neonates (n = 5), 100 nM baclofen had only slight effects on membrane potential (−1.0 ± 0.5 mV), whereas 500 and 750 nM baclofen significantly hyperpolarized the membrane potential by 2.5 ± 0.5 and 4.5 ± 0.9 mV, respectively, and decreased the number of action potentials per respiratory cycle (Fig. 2A, D and E, filled circles). Neurons of P7–P15 exhibited greater sensitivity to baclofen. Bath application of 100 nM baclofen hyperpolarized the membrane potential by 2.2 ± 0.43 mV, while 500 and 750 nM hyperpolarized the membrane potential by 4.5 ± 0.45 and 7.2 ± 0.6 mV, respec-

A. Pfeiffer, W. Zhang / Respiratory Physiology & Neurobiology 158 (2007) 22–29

tively, and decreased the spike discharge per cycle (Fig. 2B–E, open circles). In all animals tested, 1 ␮M or higher concentrations of baclofen completely abolished the burst discharge (cf. also Fig. 1A, B and E). In addition, application of the selective GABAB -receptor antagonist CGP 55845A (100 nM) partially reversed the baclofen-induced membrane hyperpolarization and restored the burst discharges (n = 10, CGP in Fig. 2C–E). Thus, the present findings suggest that at least part of the GABAB receptor-mediated inhibition on respiratory frequency could be caused by the direct hyperpolarization on inspiratory neurons, and/or by the inhibition of excitatory drive to the respiratory network (cf. also Brockhaus and Ballanyi, 1998; Zhang et al., 2002). 3.3. The density of constitutive voltage-activated K+ -currents increases within the first 2 postnatal weeks Before investigating the mechanisms underlying the GABAB -mediated membrane hyperpolarization and decrease in frequency, changes in the constitutive voltage-activated potassium currents were quantified under control conditions in inspiratory and unidentified brainstem neurons at postnatal ages P0–P15 in the presence of 0.5 ␮M TTX and 300 ␮M Cd2+ . The averaged membrane capacitance increased within the 1st postnatal week, whereas there was little change over the 2nd week (Table 1). Voltage-activated inwardly-rectifying potassium currents were evoked by 200 ms step from a holding potential of −60 mV to various test potential (inset in Fig. 3A). Both the current amplitude and the current density of the constitutive K+ -currents significantly increased within the 1st postnatal week, whereas there was little change over the 2nd postnatal week (Table 1). There was no significant difference between inspiratory and unidentified neurons (Fig. 3B; Table 1). 3.4. GABAB -receptor-activated potassium currents increase during the first 2 postnatal weeks In the present study, the GABAB -receptor agonist baclofeninduced currents (IBac ) were defined by subtracting the consecutive current under pre-treatment conditions from the cur-

27

rents evoked in the presence of baclofen (inset d in Fig. 3A(d), cf. Section 2). Fig. 3C shows typical baclofen-induced IBac currents of different postnatal ages with inwardly-rectifying characteristics (Fig. 3D). In inspiratory neurons, the amplitude of IBac were 59.5 ± 8.3 pA at P0–P3 (n = 5), 92.5 ± 10.6 pA at P4–P6 (n = 4) and 186.5 ± 17.4 pA at P7–P15 (n = 5, close bars in Fig. 3E), the results for unidentified neurons being similar (open bars in Fig. 3E). The corresponding current densities were 1.57 ± 0.24 pA/pF at P1–P3 (n = 5), 1.89 ± 0.21 pA/pF at P4–P6 (n = 4) and 3.76 ± 0.4 pA/pF at P7–P15 (n = 5, close bars in Fig. 3F). Unidentified neurons (open bars in Fig. 3F) did not differ from inspiratory neurons. The current density of IBac , thus, significantly increased after the 6th postnatal day, whereas there was only a slight increase within the 1st week. In all tested neurons, IBac depended on the concentration of baclofen (Fig. 3G) and it could be partially blocked by the selective GABAB antagonist CGP 55845A (100 nM, Fig. 3E and F, related grayed bars), verifying that the effect was indeed mediated by GABAB -receptors. It is interesting to note that, in all neurons tested, around 15–20% of the total IBac was insensitive to the GABAB -receptor antagonist CGP 55845A. This was independent of the age of the animals (grayed bars in Fig. 3E and F). To further test whether the baclofen-activated IBac was carried by potassium ion, the dependency of IBac on the extracellular potassium concentration [K+ ]o was elucidated. Elevation of [K+ ]o from 5 mM to 15 or 30 mM increased the current amplitude and shifted the current/voltage relationship of IBac currents to the positive direction (Fig. 3H). The corresponding reversal potentials of IBac were also dependent on [K+ ]o , being −77 ± 1.5 mV at 5 mM [K+ ]o , −55 ± 1.4 mV at 15 mM [K+ ]o and −41 ± 1.7 mV at 30 mM [K+ ]o (Fig. 3I), which closely resembles the calculated equilibrium potentials of −83, −57 and −39 mV at 5, 15 and 30 mM [K+ ]o , respectively. Furthermore, the IBac could be blocked by BaCl2 (0.5 mM, n = 10, Fig. 3A(f) and J), but was not sensitive to the blocker of I(h)current, ZD 7288 (100 ␮M, n = 5, data not shown). Thus, our data indicate that, under our experimental conditions, application of baclofen activated an inwardly-rectifying potassium current.

Table 1 GABAB -receptor activated potassium currents contribute to respiratory rhythm modulation during postnatal development of mouse Ages

Neurones

Membrane capacitance (pF)

n

Constitutive KIR -currents (pA)

Constitutive KIR -current density (pA/pF)

P0–P3

Inspiratory Unidentified

38.8 ± 1.9 37.6 ± 2.6

17 50

203.6 ± 48.3 208.5 ± 53.5

4.9 ± 0.5 5.3 ± 0.4

P4–P6

Inspiratory Unidentified

47.3 ± 2.7** 46.9 ± 2.3**

16 77

418.3 ± 86.1* 428.7 ± 95.2*

9.0 ± 1.1** 9.4 ± 1.3***

P7–P15

Inspiratory Unidentified

49.2 ± 3.2 48.0 ± 2.6

15 53

443.3 ± 68.5 462.3 ± 88.5

9.3 ± 1.8 10.0 ± 1.7

n 5 14 4 8 5 7

Summary of membrane capacitance, current and current density of constitutive potassium conductances determined in inspiratory and unidentified neurons at different postnatal ages. n = numbers of mice tested. Note the level of significance is indicated by asterisks. * P < 0.05. ** P < 0.01. *** P < 0.005.

28

A. Pfeiffer, W. Zhang / Respiratory Physiology & Neurobiology 158 (2007) 22–29

4. Discussion The present investigation has revealed age-related changes in GABAB -receptor-mediated modulation of respiratory rhythm generation during postnatal maturation of mice. GABAB receptors were present and functional in respiratory network of neonatal mice. Activation of GABAB -receptor evoked a voltagedependent K+ -current (IBac ) that showed inward-rectifying characteristics. The current density of IBac increased with postnatal age. Together with our previous data (Zhang et al., 1999, 2002), the present data suggest that the GABAB -receptormediated modulation of both K+ - and Ca2+ -currents contributes to the GABAB -receptor-mediated modulation of respiratory rhythm generation from the 1st postnatal day on. 4.1. Discrepancy between GABAB -receptor-mediated effect on the cellular and network-level In the respiratory network, spontaneous respiratory rhythm can be blocked at baclofen concentrations of 1 ␮M (Figs. 1 and 3, cf. also Johnson et al., 1996; Brockhaus and Ballanyi, 1998; Zhang et al., 2002), whereas higher concentrations of baclofen are necessary to maximally affect voltage-dependent potassium (20 ␮M, Fig. 3) and HVA-Ca2+ -currents (30 ␮M, Zhang et al., 1999) in inspiratory neurons of neonatal mice. Thus, compared to Ca2+ - and K+ -currents (up to 30 ␮M), rhythmic network activity seems to be much more sensitive to baclofen (up to 1 ␮M). There are at least two possibilities that might account for this obvious discrepancy: (i) GABAB -receptors are localized at extrasynaptic sites (Fritschy et al., 1999; Ritter et al., 2004). Bath application of 1 ␮M baclofen will activate both populations of GABAB -receptors leading to an increase of potassium conductance and decrease of Ca2+ -currents. The greater sensitivity of rhythm to GABAB -receptor activation may then simply reflect that the effect associated with opening of K+ channels disrupts rhythm before the effects of K+ and Ca2+ channels on membrane current or potential are detectable (Fig. 3, cf. also Johnson et al., 1996; Brockhaus and Ballanyi, 1998; Zhang et al., 2002), but will shunt all postsynaptic currents in the network and lead to cessation of rhythmic network activity. (ii) GABAB -receptors can also modulate presynaptic Ca2+ -currents in respiratory network of brain stem (Zhang et al., 1999, 2002), as well as in most other brain region (Reviews Misgeld et al., 1995). The greater sensitivity of rhythm to GABAB -receptors may therefore also reflect presynaptic mechanisms. Taken together, it is perhaps not surprising that GABAB -receptor agonists can block rhythm at lower concentrations than required to maximally activate K+ channels or inhibit Ca2+ channels. 4.2. Involvement of GABAB -receptor-mediated inhibition in rhythm modulation during early postnatal stage The present study demonstrated that activation of GABAB receptors by application of its agonist baclofen reduced respiratory frequency in brainstem slice preparation of early postnatal mouse (Figs. 1 and 2), which confirm previously data in rat (Johnson et al., 1996; Brockhaus and Ballanyi, 1998).

On the other hand, application of GABAB -receptor antagonist CGP 55845A alone increased the respiratory rhythm during the same postnatal period in mouse brainstem (Zhang et al., 2002), suggesting that GABAB -receptors are activated by endogenous GABA during on-going rhythmic activity. Together these data suggest that GABAB -receptors are functionally involved in respiratory rhythm modulation during early postnatal stage. In the present study, the GABAB -receptor-activated currents (IBac ) was assessed by subtracting the constitutive current under control condition from the voltage-activated K+ currents in presence of baclofen (inset d in Fig. 3A(d), cf. Section 2). As GABAB -receptors are spontaneously active under control conditions (Zhang et al., 2002), part of the constitutive current shown in Fig. 3B could be the GABAB -receptor-evoked current, which is not included in IBac in the present study. If so, we will rather have underestimated the magnitude of the GABAB -receptorinduced current in the present study. The present study also demonstrated that the sensitivity of the GABAB -receptor to baclofen increases with postnatal age. In P0–P3 slices the threshold concentration for baclofen was 250 nM, while IC50 was 1 ␮M for baclofen-induced modulation of respiratory rhythms. In P7–P15 slices, the threshold concentration for baclofen was 100 nM, while IC50 was 500 nM (cf. Fig. 1). This is consistent with our previous finding that the sensitivity of baclofen-induced reduction of HVA-Ca2+ currents is age-dependent (Zhang et al., 1999). One possibility that might account for this increased sensitivity is that the expression of GABAB -receptor subtypes changes during postnatal period. Fritschy et al. (1999) have shown that expression of the splice variant GABAB 1a is predominant at birth, while the splice variant GABAB 1b is predominant in the adult brain. It is therefore possible that different GABAB -receptor populations are expressed in the respiratory network during postnatal development and that they show different agonist binding affinities (Malitschek et al., 1998). 4.3. Inhibition of GABAB -receptor-activated K+ -currents by early postnatal high intracellular Cl− concentration The main finding of the present study is that the responses to GABAB -receptor activation increased with postnatal age (P0–P15, Fig. 3). Baclofen, in the presence of TTX and Cd2+ , activated a voltage-dependent, inwardly rectified potassium current from the 1st postnatal day (IBac , Fig. 3D and H). Its current density significantly increased from the 1st to 2nd postnatal week (Fig. 3E). This could explain why application of less than 1 ␮M baclofen hyperpolarized the membrane potential and abolished rhythmic burst activity of inspiratory neurons during ongoing respiratory activity (cf. Fig. 2). We attribute these findings, at least partly, to the increasing functional presence of postsynaptic GABAB -receptor-mediated activation of inwardlyrectified potassium currents in the respiratory network. In neonatal mice (P0–P3), the equilibrium potential of chloride in neurons of different brain regions is more positive than the resting membrane potential, which is also true for respiratory network in neonatal mice (Ritter and Zhang, 2000), but not for neonatal rat (Shao and Feldman, 1997; Brockhaus and

A. Pfeiffer, W. Zhang / Respiratory Physiology & Neurobiology 158 (2007) 22–29

Ballanyi, 1998). During the same postnatal age (P0–P15), the equilibrium potential of chloride decreases significantly from −10 mV at P0 to −75 mV at P7–P15 in the respiratory network of mice, which corresponds with an intracellular chloride concentration of more than 100 mM at P0 (Ritter and Zhang, 2000). Such high concentrations of intracellular chloride strongly attenuate GABAB -receptor-activated potassium currents as has been shown in hippocampal CA1 pyramidal cells (Lenz et al., 1997). Thus, changes in equilibrium potential of chloride during postnatal stage might affect GABAB -receptor-mediated modulation of potassium currents. As suggested by Lenz et al. (1997) possible mechanisms for such chloride dependency of action of GABAB -receptors might be direct interactions of Cl− ions with either the GABAB -receptor, the associated G-protein-mediated pathway or finally the K+ channel itself. Acknowledgements We are grateful to Drs. Hess and M¨uller for their valuable comments on the manuscript, to A. Herdlitschke, C. Batje and C. H¨uhne for skilful technical assistance and to Novartis Ltd. for the generous supply of CGP 55845A. The present project was funded by Deutsche Forschungsgemeinschaft through the DFG-Research Center for Molecular Physiology of the Brain. References Brockhaus, J., Ballanyi, K., 1998. Synaptic inhibition in the isolated respiratory network of neonatal rats. Eur. J. Neurosci. 10, 3823–3839. Dutar, P., Nicoll, R.A., 1988. A physiological role for GABAB receptors in the central nervous system. Nature 332, 156–158. Fritschy, J.M., Meskenaite, V., Weinmann, O., Honer, M., Benke, D., Mohler, H., 1999. GABAB-receptor splice variants GB1a and GB1b in rat brain: developmental regulation, cellular distribution and extrasynaptic localization. Eur. J. Neurosci. 11, 761–768. Gahwiler, B.H., Brown, D.A., 1985. GABAB-receptor-activated K+ current in voltage-clamped CA3 pyramidal cells in hippocampal cultures. Proc. Natl. Acad. Sci. U.S.A. 82, 1558–1562. Gillis, K.D., 1995. Techniques for membrane capacitance measurements. In: Sakmann, B., Neher, E. (Eds.), Single-Channel Recording. Plenum Press, New York, pp. 155–198. Grillner, S., Deliagina, T., Ekeberg, O., el Manira, A., Hill, R.H., Lansner, A., Orlovsky, G.N., Wallen, P., 1995. Neural networks that co-ordinate locomotion and body orientation in lamprey. Trends Neurosci. 18, 270–279.

29

Grillner, S., Wall´en, P., Brodin, L., Lasner, A., 1991. Neuronal network generating locomotor behavior in lamprey: circuitry, transmitters, membrane properties, and stimulation. Annu. Rev. Neurosci. 14, 169–199. Johnson, S.M., Smith, J.C., Feldman, J.L., 1996. Modulation of respiratory rhythm in vitro: role of Gi/o protein-mediated mechanisms. J. Appl. Physiol. 80, 2120–2133. Lenz, R.A., Pitler, T.A., Alger, B.E., 1997. High intracellular Cl-concentrations depress G-protein-modulated ionic conductances. J. Neurosci. 17, 6133–6141. Malitschek, B., Ruegg, D., Heid, J., Kaupmann, K., Bittiger, H., Frostl, W., Bettler, B., Kuhn, R., 1998. Developmental changes of agonist affinity at GABABR1 receptor variants in rat brain. Mol. Cell. Neurosci. 12, 56– 64. Misgeld, U., Bijak, M., Jarolimek, W., 1995. A physiological role for GABAB receptors and the effects of baclofen in the mammalian central nervous system. Prog. Neurobiol. 46, 423–462. Newberry, N.R., Nicoll, R.A., 1985. Comparison of the action of baclofen with gamma-aminobutyric acid on rat hippocampal pyramidal cells in vitro. J. Physiol. (Lond.) 360, 161–185. Nicoll, R.A., Malenka, R.C., Kauer, J.A., 1990. Functional comparison of neurotransmitter receptor subtypes in mammalian central nervous system. Physiol. Rev. 70, 513–565. Richter, D.W., 1996. Neural Regulation of Respiration: Rhythmogenesis and Afferent Control. Springer-Verlag, Berlin Heidelberg, pp. 2079–2095. Ritter, B., Zhang, W., 2000. Early postnatal maturation of GABAA-mediated inhibition in the brainstem respiratory rhythm-generating network of the mouse. Eur. J. Neurosci. 12, 2975–2984. Ritter, B., Ochojski, M., K¨uhn, T., Schwarzacher, S.W., Zhang, W., 2004. Subcellular vesicular aggregations of GABAB R1a and R1b receptors increase with age in neurons of the developing mouse brain. Cell Tissue Res. 319, 181–189. Shao, X.M., Feldman, J.L., 1997. Respiratory rhythm generation and synaptic inhibition of expiratory neurons in pre-Botzinger complex: differential roles of glycinergic and GABAergic neural transmission. J. Neurophysiol. 77, 1853–1860. Smith, J.C., Ellenberger, H.H., Ballanyi, K., Richter, D.W., Feldman, J.L., 1991. Pre-Botzinger complex: a brainstem region that may generate respiratory rhythm in mammals. Science 254, 726–729. Thalmann, R.H., 1988. Evidence that guanosine triphosphate (GTP)-binding proteins control a synaptic response in brain: effect of pertussis toxin and GTP gamma S on the late inhibitory postsynaptic potential of hippocampal CA3 neurons. J. Neurosci. 8, 4589–4602. Zhang, W., Barnbrock, A., Gajic, S., Pfeiffer, A., Ritter, B., 2002. Differential ontogeny of GABA(B)-receptor-mediated pre- and postsynaptic modulation of GABA and glycine transmission in respiratory rhythm-generating network in mouse. J. Physiol. 540, 435–446. Zhang, W., Elsen, F., Barnbrock, A., Richter, D.W., 1999. Postnatal development of GABAB receptor-mediated modulation of voltage-activated Ca2+ currents in mouse brain-stem neurons. Eur. J. Neurosci. 11, 2332–2342.