Skip to main content
PMC home page
The Journal of Physiology logoLink to The Journal of Physiology
. 1998 Mar 15;507(Pt 3):783–794. doi: 10.1111/j.1469-7793.1998.783bs.x

Glycine-activated currents are changed by coincident membrane depolarization in developing rat auditory brainstem neurones

Kurt H Backus *, Joachim W Deitmer *, Eckhard Friauf †
PMCID: PMC2230818  PMID: 9508839

Abstract

  1. During early ontogeny, glycine receptors (GlyRs) exert depolarizing responses which may be of developmental relevance. We have used the gramicidin-perforated patch technique to elucidate the mechanism of glycine-activated currents in developing neurones of the rat lateral superior olive (LSO).

  2. When the holding potential was set to −60 mV, perforated-patch recordings revealed glycine-induced inward currents in 59%, outward currents in 5% and biphasic currents in 34% of the LSO neurones tested (n = 44). The biphasic currents were characterized by a transient outward phase which was followed by an inward phase.

  3. Ion substitution experiments showed that both Cl− and HCO3− contributed to the glycine- induced biphasic current responses.

  4. In the biphasic responses, the reversal potential of the glycine-induced current (Egly) depended on the response phase. A strong shift of Egly from a mean of −72 mV during the outward phase of the glycine response to a mean of −51 mV during the inward phase was observed, suggesting a shift of an ion gradient.

  5. When the membrane potential was depolarized, ‘tail’ currents were induced in the presence of glycine. An increased duration or amplitude of the evoked depolarizations resulted in a proportional enlargement of these tail currents, indicating that they were produced by a shift of an ion gradient. Since changes of the HCO3− gradient are negligible, because of the carbonic anhydrase activity, we suggest that these tail currents were caused by a shift of the Cl− gradient.

  6. We conclude that Cl− accumulates intracellularly during the activation of GlyRs and, consequently, Egly moves towards more positive values.

  7. Coincident depolarizing stimuli enhanced intracellular Cl− accumulation and the shift of Egly, thereby switching hyperpolarizing to depolarizing action. This change could assist in an activity-dependent strengthening and refinement of glycinergic synapses during the maturation of inhibitory connectivity.

At present, the mechanisms underlying the maturation of precisely organized inhibitory connections in the central nervous system are still unsolved. This is in contrast to the development of topographically arranged excitatory connections, for which several mechanisms have been proposed (for review see Constantine-Paton, Cline & Debski, 1990; Goodman & Shatz, 1993). Interestingly, glycine and γ-aminobutyric acid (GABA), the two major inhibitory neurotransmitters in vertebrates, evoke membrane depolarizations during fetal and early postnatal life in many areas of the nervous system (e.g. spinal cord: Wu, Ziskind-Conhaim & Sweet, 1992; Reichling, Kyrozis, Wang & MacDermott, 1994; cortex: Luhmann & Prince, 1991; Owens, Boyce, Davis & Kriegstein, 1996; brainstem: Kandler & Friauf, 1995; hippocampus: Ben-Ari, Cherubini, Corradetti & Gaiarsa, 1989). This depolarizing action can in effect be excitatory (Nishimaru, Iizuka, Ozaki & Kudo, 1996) or induce a Ca2+ influx (Reichling et al. 1994; Owens et al. 1996). In essence, the transiently depolarizing action of inhibitory transmitters has been suggested to be crucially involved during early maturational processes of inhibitory synapse formation.

The lateral superior olive (LSO) is an auditory brainstem nucleus with specific anatomical features: excitatory and inhibitory inputs, mediated via anatomically distinct pathways from the ipsilateral and the contralateral ear, respectively, converge within the LSO. Both types of input are highly topographically organized (Caird & Klinke, 1983) and act via glutamate receptors and glycine receptors (GlyRs) (Wenthold, 1991). Thus, the LSO provides a favourable system for experimental investigation of the maturation of inhibitory and excitatory synapses (Sanes & Rubel, 1988; Kandler & Friauf, 1995). In order to explore the ionic basis of glycinergic action during development, we applied the whole-cell and the gramicidin-perforated patch technique and recorded from neurones in the LSO of rats between postnatal day (P) 2 and P13, which corresponds to the period of synapse maturation (Kandler & Friauf, 1993; Kil, Kageyama, Semple & Kitzes, 1995). Our results show that glycine-evoked currents in neurones are carried by Cl− and HCO3− anions. Furthermore, the reversal potential of the glycine-induced current (Egly) becomes acutely shifted towards more positive values and this effect is significantly enhanced by coincident depolarizing pulses. Such a context-dependent change from hyperpolarizing to depolarizing activity may play a role in the maturation of glycinergic synapses. Parts of the results were communicated in abstract form (Backus & Friauf, 1996).

METHODS

Preparation of brainstem slices

Brainstem slices were prepared as described elsewhere (Edwards, Konnerth, Sakmann & Takahashi, 1989). In brief, rat pups (P2-P13) were decapitated and their brains rapidly removed and incubated in a chilled, Na+-reduced saline (ca 1–2°C; composition given below; adapted from Barnes-Davis & Forsythe, 1995). This solution was continuously bubbled with 95% O2-5% CO2 to supply the tissue with oxygen and to maintain the pH at 7.4. Frontal slices (250–300 μm thickness) of the brainstem were cut on a vibratome (Campden Instruments, Loughborough, UK). The slices were preincubated at 37°C for 60 min and transferred to the HCO3−-buffered standard saline (listed below) in which they were stored at room temperature (ca 22°C) and bubbled with 95% O2-5% CO2. Electrophysiological experiments were started 1 h after the preparation of the slices.

Solutions and drug application

The preparation of the slices and the preincubation was done in a salt solution containing (mm): KCl, 2.5; NaHCO3, 26; glucose, 260; CaCl2, 2.0; MgCl2, 1.0; sodium pyruvate, 2.0; myo-inositol, 3.0; kynurenic acid, 1.0. The pH was adjusted to 7.2 with NaOH and maintained by bubbling with 95% O2-5% CO2. During the experiments, the slices were continuously superfused with a standard HCO3−-buffered saline containing (mm): NaCl, 125; KCl, 2.5; NaHCO3, 26; NaH2PO4, 1.25; glucose, 25; CaCl2, 2.0; MgCl2, 1.0; pH adjusted to 7.4 with NaOH and maintained by bubbling with 95% O2-5% CO2. For the determination of the ionic basis of glycine-induced currents, several external solutions were used: (i) a Hepes-buffered solution containing (mm): NaCl, 125; KCl, 2.5; sodium acetate, 26; NaH2PO4, 1.25; glucose, 25; CaCl2, 2.0; MgCl2, 1.0; Hepes, 10; pH adjusted to 7.4 with NaOH; oxygen supply was maintained by bubbling with 100% O2, (ii) a nominally Cl−-free HCO3−-buffered solution containing (mm): sodium acetate, 125; potassium acetate, 2.5; NaHCO3, 26; NaH2PO4, 1.25; glucose, 25; caesium acetate, 2.0; magnesium acetate, 1.0; Hepes, 10; pH adjusted to 7.4 with NaOH and maintained by bubbling with 95% O2-5% CO2, and (iii) a nominally Cl−-free, Hepes-buffered solution containing (mm): sodium acetate, 151; potassium acetate, 2.5; glucose, 25; calcium acetate, 2.0; magnesium acetate, 1.0; Hepes, 10; the pH was adjusted to 7.4 with NaOH and oxygen supply was maintained by bubbling with 100% O2. Gramicidin (5 mg ml−1; Sigma) was dissolved in dimethylsulphoxide (DMSO), vortexed for 1 min, sonicated for 20 s, and then added to the pipette solution to give a final concentration of 5–30 μg ml−1. The pipette solution could be used the entire day and contained (mm): KCl, 140; NaCl, 5; CaCl2, 0.5; EGTA, 5; Hepes, 10; pH adjusted to 7.2 with KOH. The pipette solution for intracellular Cl−-free recordings contained (mm): potassium acetate, 140; sodium acetate, 5; calcium acetate, 0.5; magnesium acetate, 1; EGTA, 5; Hepes, 10; pH adjusted to 7.2 with KOH. A stock solution of glycine was prepared (1 m in distilled water) and glycine was added to the saline shortly before the experiment. An application tool, positioned at the slice surface at a distance of about 100 μm from the cell body, was used for a rapid focal drug delivery (for details of the application method see Knoflach et al. 1992).

Experimental set-up

The recording chamber was mounted on an upright microscope (Axioskop FS, Zeiss, Germany) that was equipped with Nomarski optics and an infrared video camera system (Hamamatsu, Japan). Brainstem slices were fixed with a nylon grid in a recording chamber that was continuously superfused with saline. Patch pipettes were made from borosilicate glass and had resistances between 3 and 5 MΩ when filled with the pipette solution. Gramicidin-perforated patch recordings were performed following the method of Kyrozis & Reichling (1995). Briefly, the electrode's tip was filled (for 1–180 s) with gramicidin-free pipette solution to avoid problems with seal formation. Then the electrode was backfilled with the gramicidin-containing pipette solution. In most experiments, an AgCl-coated silver wire was used as the bath electrode. In the set of experiments in which Cl−-free solutions (compositions see above) were used, the bath was earthed with a salt bridge to keep junction potentials minimal. To do so, agar-agar (Serva, Heidelberg, Germany; 1%; dissolved in the above standard extracellular saline) was filled into the tip of a glass tube (diameter, 3 mm) and backfilled with the Cl− free pipette solution (composition see above) into which an AgCl-coated silver wire was inserted. As no reversal potentials were determined in this set of experiments, we have not compensated for junction potentials. The reference potential for all measurements was the zero current potential of the pipette in the bath before establishment of the gigaseal. After the formation of a gigaseal (Hamill, Marty, Neher, Sakmann & Sigworth, 1981), short rectangular voltage commands (−2 mV; 5 ms) were continuously applied to monitor the gradual decrease in series resistance. Drug application was not started until the series resistance had fallen below 30 MΩ (usually after 20–30 min). Series resistance was compensated for at least 80%. If not stated otherwise, the holding potential (Vh) was set to −60 mV. In ten neurones, the conventional whole-cell configuration was established after recording in the gramicidin-perforated patch configuration. This resulted in a dialysis of the cell interior with the pipette solution and in a shift of the chloride equilibrium potential (ECl) towards 0 mV within the next 3 min. In the conventional whole-cell configuration, glycine induced unidirectional inward currents whose amplitudes were one or two orders of magnitude larger than those of currents recorded in the gramicidin-perforated patch configuration (see Fig. 1). Therefore, it was possible to reject the recordings from cells in which a spontaneous rupture of the cell membrane had unintentionally produced a conventional whole-cell recording. Current signals were amplified by an Axopatch-1D (Axon Instruments) patch-clamp amplifier. Before digitization (sampling rate, 10 kHz), the signals were filtered at 5 kHz with a 3-pole low-pass Bessel filter. Data were stored and evaluated with the aid of the pCLAMP hardware and software package (Axon Instruments) for a personal computer.

Figure 1. Glycine-activated responses in LSO neurones.

Figure 1

Open in a new tab

Glycine (1 mm) was applied to the bath as indicated by the horizontal bars. A, gramicidin-perforated patch recording in the current-clamp mode showed biphasic membrane potential changes in 50% of neurones. B, when recording in the voltage-clamp mode of the perforated-patch configuration, glycine evoked an outward current in 5% of the neurones (left), but an inward current in 59% (right). C, in 34% of the neurones, glycine induced biphasic current responses in the perforated-patch configuration (left). After establishment of the conventional whole-cell configuration (WCC), glycine action was characterized by a strong inward current (right). Both traces in C were obtained from the same LSO neurone. Note the clear difference in current amplitude. Pipette solution contained 146 mm Cl−, and [Cl−]o was 133.5 mm in all experiments shown. Vh = −60 mV in B and C.

Determination of current-voltage relation, data evaluation, and statistics

Glycine was applied to the bath solution only after the holding current was stable for at least 3 min. In order to analyse the current-voltage (I-V) relation of glycine-evoked currents, voltage ramps (from 20 or 0 mV to −100 mV; duration, 100 ms) were applied both in the absence and presence of glycine. The digitized response to a ramp prior to glycine application was subtracted from that in the presence of glycine. This was done in order to eliminate the contamination by leakage and voltage-activated currents and to obtain the ‘pure’ glycine-induced I-V relation. Data are given as means ± s.e.m. The significance of the difference between mean values of two samples was determined using Student's two-tailed t test.

RESULTS

Glycine-activated currents in LSO neurones

Glycine-evoked responses were investigated in LSO neurones (P2-P13) using the perforated patch-clamp technique with gramicidin as the membrane-perforating agent. Gramicidin forms pores of negligible Cl− permeability (Myers & Haydon, 1972; Kyrozis & Reichling, 1995) and, thus, the intracellular Cl− concentration ([Cl−]i) should remain unaffected. In recordings performed in the current-clamp mode, glycine (1 mm; 12–60 s) evoked biphasic changes of the membrane potential in four cells, characterized by an initial hyperpolarization which was followed by a depolarization (Fig. 1A). Glycine-evoked depolarizations were observed in two neurones and hyperpolarizations in another two (not shown). When glycine was applied while recordings were done in the voltage-clamp mode, three patterns of current responses were observed (Fig. 1). Five percent of the neurones (2 of 44) responded with an outward current (Fig. 1B), 59% with an inward current (Fig. 1B), and 34% of the cells showed a biphasic response, which was characterized by an initial, transient outward component, followed by a sustained inward component (Fig. 1C); in one cell, glycine failed to evoke any response. The two neurones which responded with an outward current were from P2 and P9 rats. The mean age of cells responding with an inward current was 6.1 ± 0.6 days (range, P3-P13; n = 26) and that of cells showing a biphasic response 6.4 ± 1.1 days (range, P2-P10; n = 15). These mean values did not significantly differ between each other. In addition, we divided the LSO neurones into classes with respect to age (bin width, 2 days), and found that the relative frequency of cells responding with biphasic currents (35 ± 4%) versus those responding with inward currents (61 ± 5%) was relatively constant throughout the time period under investigation (from P2 to P13). We therefore conclude that there is no correlation between the response type (biphasic or inward) and the age of the cells.

After recording in the gramicidin-perforated patch configuration, we established the conventional whole-cell configuration in ten neurones by rupturing the cell membrane under the tip of the patch pipette. Within 3 min, the response behaviour to glycine changed: when kept at a holding potential (Vh) of −60 mV, the neurones now displayed a large inward current (Fig. 1C) as expected after the cell interior has become dialysed with a pipette solution containing 146 mm Cl− (extracellular solution, 133.5 mm Cl−). In all thirty-five neurones which were recorded exclusively in the conventional whole-cell configuration (pipette and extracellular solutions as above), glycine induced large inward currents. These results indicate a substantial contribution of Cl− ions to the current through GlyR channels in LSO neurones.

In order to exclude the possibility that glycine-evoked biphasic responses, in particular the sustained inward component, were mediated via indirect synaptic effects, we blocked synaptic transmission and neuronal activity. In these experiments, glycine was applied in a nominally Ca2+-free saline containing 10 mm Mg2+ and 300 nm tetrodotoxin. This protocol had hardly any effect on amplitudes and kinetics of glycine-induced currents, neither in the perforated patch (n = 5; Fig. 2A) nor in the conventional whole-cell configuration (n = 5; Fig. 2B), demonstrating that the observed responses were in fact directly elicited by glycine on the neurones from which the recordings were obtained.

Figure 2. Effect of glycine on LSO neurones while synaptic transmission and spike activity were blocked.

Figure 2

Open in a new tab

Glycine (1 mm) was applied to the bath as indicated by the bars. A, responses obtained in the perforated-patch mode (GPP); B, responses recorded in the conventional whole-cell configuration (WCC). The current responses were almost unchanged in a nominally Ca2+-free saline which contained 10 mm Mg2+ and 300 nm tetrodotoxin to block synaptic transmission and neuronal spike activity, indicating that glycine acted directly on the neurone under investigation.

Ionic contribution of HCO3− to glycine-activated currents

GlyRs form ion channels which are not only permeable to Cl−, but also to HCO3− (Bormann, Hamill & Sakmann, 1987). Assuming an intracellular pH of 7.2 (Chesler, 1990), the internal HCO3− concentration would be near 16 mm, resulting in an HCO3− equilibrium potential (EHCO3) of ca−12 mV. Therefore, part of the depolarizing glycinergic action may be caused by an efflux of HCO3−. To address this issue, the ionic mechanism underlying the glycine-evoked currents was investigated in ion substitution experiments. In five of five neurones recorded in the gramicidin-perforated patch mode, withdrawal of CO2-HCO3− from the extracellular saline, by superfusing a Hepes-buffered saline, greatly reduced the inward component of the glycine-induced biphasic response and increased the outward component (Fig. 3A). Readdition of the CO2-HCO3−-buffered saline reestablished the inward component, indicating that it was probably mediated by an efflux of HCO3−. In the conventional whole-cell configuration using a Cl−-free pipette solution (ECl << Vh), glycine evoked a biphasic current response (Fig. 3B, left trace). When the extracellular solution was then substituted by a nominally Cl−-free solution, the glycine-induced outward current component disappeared (n = 8; Fig. 3B, middle trace), suggesting that it was carried by Cl− ions entering the neurone. When glycine was finally applied in a Cl−-free, CO2-HCO3−-free saline, only a very small, if any, residual current could be observed (n = 11; Fig. 3B, right trace). These data indicate that the glycine-induced transient outward current component is primarily carried by an influx of Cl−, while the sustained inward current component is largely carried by an efflux of HCO3−.

Figure 3. Contribution of HCO3− and Cl− to glycine-induced currents in LSO neurones.

Figure 3

Open in a new tab

A, gramicidin-perforated patch configuration. At a Vh of −60 mV, bath-applied glycine (1 mm as indicated by the bars) evoked an inward current in CO2-HCO3−-buffered saline ([HCO3−]o, 26 mm). A switch from HCO3− to a HCO3−-free, Hepes-O2-buffered saline (indicated by the filled bar) greatly reduced the inward current and gave rise to a transient outward component, suggesting that the inward current was largely mediated by an efflux of HCO3−; [Cl−]o, 133.5 mm. B, conventional whole-cell clamp configuration. At a Vh of −60 mV, bath-applied glycine (1 mm; see bars) evoked a biphasic current response shortly after the establishment of the whole-cell configuration (left trace). Withdrawal of Cl− from the bath solution resulted in the disappearance of the outward component and in an increase of the inward component (middle trace). In a solution containing neither Cl− nor HCO3− almost no current response was observed (right trace). Pipette solution in B was nominally Cl− free.

The reversal potential of glycine-activated currents

Based on our finding that the biphasic current responses were directly mediated by GlyRs, we analysed the current-voltage (I-V) relation of the glycine-induced responses and determined their reversal potential (Egly). Voltage ramp protocols were applied before, during, and after glycine-induced current responses (Fig. 4A). The control response was subtracted from that seen in the presence of glycine to determine the ‘pure’ glycine-induced I-V relations, isolated from leakage and/or voltage-activated currents. Based on these I-V relations, Egly was determined at different times during the glycine action, because the value of Egly depended on the response phase (Fig. 4B). A strong shift of the current reversal from −72.1 ± 7.2 mV (n = 4) ca 12 s after the beginning of the glycine response to −50.6 ± 4.6 mV (n = 6) after ca 48 s became apparent during the course of the glycine response (Fig. 4B and C), indicating some change in the distribution of the ions involved in this response.

Figure 4. Reversal potential of glycine-induced currents in LSO neurones.

Figure 4

Open in a new tab

A, in order to determine the I-V relation of the glycine-induced current, voltage ramp protocols (Vh, −60 mV; range, −100 to 0 mV; duration, 100 ms) were applied before (con) and during the current response activated by 1 mm glycine (a-d). Note, current responses to voltage ramps are truncated due to the slow time resolution of the pen writer. B, I-V relations of the current responses corresponding to the voltage ramp responses a and c in A were created by subtracting the current response before glycine from the response obtained during the glycine action. The reversal potential during the outward current component (a - con) was −88 mV and during the inward current component (c - con) −52 mV, indicating that Egly was shifted about +36 mV during one minute. Traces are illustrated for the potential range between −100 and −40 mV. C, left panel, plot of Egly as a function of time during a glycine-evoked current response from a single neurone; □, cell membrane potential before glycine (−73 mV); ▪, Egly in the presence of glycine. Right panel, plot of mean Egly as a function of time obtained from 10 neurones. ^, mean membrane potential before glycine (−64 mV, s.d. 15 mV; n = 10); •, mean Egly in the presence of glycine; bars indicate s.d.; n = 2–10 for each data point; data were binned into classes of 6 s width. Dashed lines indicate the resting membrane potential. Note that Egly shifts on average by about 30 mV.

Glycine-evoked tail currents in LSO neurones

Since the gramicidin-perforated patch configuration allows for response-mediated changes in the intracellular ion concentrations, a shift in the Cl− gradient could have caused the reversal of the current response evoked by glycine. We therefore used another technique to attempt to shift ECl to more positive values. Application of depolarizing voltage pulses during the activation of GlyRs should force additional Cl− into the cell, thereby driving ECl, and hence Egly, towards more positive values. The membrane potential was stepped from a Vh of −80 mV to +60 mV for 50–200 ms. In the presence of glycine, inwardly directed tail currents followed these depolarizing pulses; they lasted several seconds and displayed a slow relaxation (Fig. 5). The net glycine-evoked tail currents at different potentials, ranging between −10 and −100 mV (Fig. 5C and top panel of Fig. 5, respectively), were isolated by subtracting the control current trace recorded before glycine application (Fig. 5A) from that obtained in the presence of glycine (Fig. 5B). The reversal potential of the glycine-evoked tail currents (Fig. 5D) was strongly dependent on the response phase and shifted towards more positive values with time. The reversal potential of the glycine-dependent tail current, determined at different phases of the response in five neurones, shifted from −73 to −35 mV, depending on whether the protocol was applied during the outward or the inward component, respectively.

Figure 5. Glycine-induced tail currents in LSO neurones.

Figure 5

Open in a new tab

In the presence of glycine (1 mm), tail currents were observed after the end of depolarizing rectangular command pulses (duration, 50 ms) when the voltage was stepped back to variable potentials (from −10 to −100 mV, as indicated above; Vh, −80 mV), and they lasted several seconds. The voltage protocols were applied in the absence and in the presence of glycine, giving rise to the tail currents illustrated in A and B, respectively. C, difference ‘B - A’ shows the net effect of glycine at different membrane potentials. Note that only the tail currents, but not the current responses to the preceding depolarizing rectangular command pulses are displayed in A-C. D, I-V relation. The peak of the tail current (indicated by an arrow in C) was plotted as a function of the voltage (reversal potential, −35 mV in this cell, but see text).

Our data indicate that the tail currents, which were produced by the coincident occurrence of depolarization and activation of GlyRs, were caused by a shift of an ion gradient. This finding was confirmed by the results of experiments in which the driving force was enhanced by increasing either the duration or the amplitude of depolarizing voltage pulses. The prolongation of the depolarizing voltage step from 10 to 200 ms increased the amplitude of the tail currents as well as their duration in five of five neurones (Fig. 6A). Likewise, increasing the amplitude of the depolarizing voltage step (from −20 mV to +20 or +60 mV) resulted in an increased tail current amplitude in all neurones tested (Fig. 6B; n = 5).

Figure 6. Effect of pulse duration (A) and pulse amplitude (B) on glycine-induced tail currents.

Figure 6

Open in a new tab

A, five successive voltage steps (from Vh of −60 mV to +40 mV) with increasing duration (10, 50, 100, 150 and 200 ms) were applied. In the absence of glycine (a, control), the pulse duration had no effect on the tail current. With 1 mm glycine in the bath solution (b), a prolonged tail current component was present. The amplitude of the tail current increased with increasing pulse duration. Bottom traces show the subtracted current traces as indicated. B, three successive voltage steps (duration, 20 ms; Vh, −60 mV) to increasing levels (−20, +20 and +60 mV) were applied. In the absence of glycine (a, control), the pulse amplitude had no effect on the amplitude of the tail current. In the presence of 1 mm glycine (b), the amplitude of the tail current increased with increasing pulse amplitude. Bottom traces show the subtracted current traces. Pulse amplitudes are truncated and baselines were adjusted in A and B. Dotted line indicates zero current.

Synchronous depolarization shifts Egly

The effect of coincident depolarization on Egly was further analysed in a set of experiments in which we depolarized the neurones with pulse trains which were designed to mimic action potential burst activity (10 pulses of 2 ms duration over a period of 250 ms; from −80 mV to +40 mV). Egly was determined by applying ten voltage ramp commands at 2 Hz. The voltage commands were started during the initial outward current transient of the glycine-induced response. The superimposed current responses to the voltage ramp commands recorded in the absence and presence of glycine are displayed in Fig. 7A and B, respectively. Subtracting the ramp responses recorded in the absence of glycine from those obtained in its presence allowed the isolation of the ‘pure’ glycine-evoked I-V relation (Fig. 7C). Shifts of Egly were calculated as the difference between Egly determined with the first, and Egly determined with the last ramp command. Application of the voltage ramps alone shifted Egly by +5.4 ± 1.2 mV (n = 7; Fig. 7D, left panel). However, when each voltage ramp was preceded by the pulse train, Egly shifted about +9.9 ± 2.0 mV (n = 7; Fig. 7D, right panel). The mean shift of Egly determined with pulse trains was significantly larger than that obtained when no pulse trains were applied (P < 0.001; paired t test).

Figure 7. Coincident depolarizing activity shifts Egly in LSO neurones.

Figure 7

Open in a new tab

The coincident occurrence of spike activity and glycinergic input (left panel) was simulated by applying a train of 10 depolarizing pulses (from −80 mV to +40 mV; pulse duration 2 ms; pulse frequency 40 Hz; right panel) during the presence of currents activated by glycine (1 mm). Egly was determined using the voltage ramp method as described in the text. The time-dependent shift of Egly was determined by the successive application of 10 voltage ramps at 2 Hz. A, superposition of 10 current ramp responses in the absence of glycine (control). B, same as in A but in the presence of glycine. C, the current traces shown in A were subtracted from the corresponding traces displayed in B. The difference reflects the shift of Egly. D, part of C at higher magnification, illustrating the shift of the reversal potential.

DISCUSSION

Our study provides two major results. First, the application of the gramicidin-perforated patch configuration shows that glycine primarily induces inward currents and biphasic current responses which are mediated by Cl− and HCO3− anions in LSO neurones. Second, during the activation of GlyRs, Egly is shifted towards more positive values, which is potentiated by the coincident occurrence of depolarizing stimuli.

To take into account that changes in [Cl−]i are important when the physiology of GlyRs is studied, we have used the perforated patch configuration with gramicidin which forms pores selectively permeable to monovalent cations, but impermeable to anions and divalent cations (Myers & Haydon, 1972; Kyrozis & Reichling, 1995). In the perforated patch configuration, glycine-evoked currents were relatively small, because Egly in LSO neurones appeared to be close to Vh (−60 mV) between P2 and P13. In the whole-cell configuration, Egly moved towards the new ECl (+2.2 mV) and, therefore, glycine-induced currents were exclusively inwardly directed at a Vh of −60 mV and one to two orders of magnitude higher in amplitude than seen in the perforated patch configuration. These differences were used as an indicator to distinguish between perforated patch clamps and unintended whole-cell clamps during our experiments.

Ionic basis of glycine-activated biphasic currents

We found three response types in LSO neurones which were mediated by GlyRs. In more than half of the neurones studied (26 of 44), glycine evoked an inward current, indicating an Egly more positive than the Vh of −60 mV. Indeed, glycine-induced depolarizations were previously found in LSO neurones between embryonic day 18 and P7, while hyperpolarizations predominated after P8 (Kandler & Friauf, 1995). Glycinergic depolarizing activity has also been observed during the development of several other neuronal cell types, e.g. neonatal rat hippocampal neurones (Ito & Cherubini, 1991) and in spinal cord neurones (Wu et al. 1992; Reichling et al. 1994). The main inhibitory transmitter in the brain, GABA, also shows transient depolarizing activity during early developing stages, e.g. in the hippocampus (Ben-Ari et al. 1989), neocortex (Luhmann & Prince, 1991; Owens et al. 1996), hypothalamus, and the spinal cord (Wu et al. 1992; Reichling et al. 1994; Rohrbough & Spitzer, 1996).

The transient depolarizing activity of glycine and GABA may be due to an active intracellular Cl− accumulation in neurones during early development. This model requires the expression of some transport mechanisms responsible for the continuous maintenance of a relatively high [Cl−]i such as the Na+-dependent Cl− cotransport demonstrated in Rohon-Beard spinal neurones of Xenopus larvae (Rohrbough & Spitzer, 1996), the Na+-K+-Cl− cotransport that has been identified in rat sympathetic neurones (Ballanyi & Grafe, 1985) and the Cl−-HCO3− exchange system found in rat cerebellar Purkinje cells (Gaillard & Dupont, 1990). After the refinement of the inhibitory connections, the relevant transporters might be down-regulated, resulting in a passive Cl− distribution and, thus, in a change from a depolarizing to a hyperpolarizing response. Indeed, GABA action is depolarizing in rat cerebellar granule cells at P7, occasionally triggering single spikes, whereas by P18, the GABA reversal potential has shifted close to the resting potential, when GABA produces a shunting inhibition (Brickley, Cull-Candy & Farrant, 1997).

In 34% of the LSO neurones, we observed biphasic responses directly evoked by glycine. GlyR ion channels have been reported to be permeable mainly to Cl− and, to a lesser extent, to HCO3− (Bormann et al. 1987; Kaila, 1994). Therefore, Egly is determined by both anion gradients and the permeability ratio of HCO3− and Cl− (PHCO3/PCl). Egly can be calculated by the following equation:

graphic file with name tjp0507-0783-m1.jpg (1)

where R is the gas constant; T is the absolute temperature; F is the Faraday constant; [Cl−]o and [Cl−]i are the extra- and intracellular Cl− concentrations, respectively; [HCO3−]o and [HCO3−]i are the extra- and intracellular HCO3− concentrations, respectively; and b represents PHCO3/PCl, which was calculated to be 0.11 in cultured spinal cord neurones (Bormann et al. 1987).

In our study, ion substitution experiments revealed that the outward component of the biphasic glycine response disappeared in LSO neurones when the extracellular Cl− was withdrawn, indicating that this component was mediated by a Cl− influx. On the other hand, the withdrawal of HCO3− from the saline led to a strong decrease of the inward component, indicating that HCO3− efflux contributed significantly to this component. Therefore, both Cl− and HCO3− appear to contribute to the glycine-induced biphasic current responses in LSO neurones.

However, this finding does not explain the polarity change of the response during the activation of GlyRs. A change of the polarity during the response could be explained by a shift in the Cl− and/or HCO3− gradient. While a substantial change in the intracellular HCO3− concentration would be limited due to the CO2 shuttle and a presumed carbonic anhydrase activity, an uptake of Cl− may be more likely to cause the reversal of the current response. With Egly being more negative than the membrane potential, Cl− will initially enter the cell. This will lead to an increase in [Cl−]i and to a shift of Egly towards more positive values (cf. eqn (1)). Based on our finding that Egly is approximately −70 mV during the outward component (Fig. 4C), using eqn (1), the resting [Cl−]i is calculated to be 7 mm. According to eqn (1), an increase in [Cl−]i of about 5 mm would be sufficient to shift Egly positive to −60 mV. This would change the direction of the glycine-induced current from outward to inward. Our finding that tail currents occurred when depolarizing voltage pulses were applied in the presence of glycine supports this conclusion. While GlyRs are activated, a depolarizing voltage pulse drives Cl− into the neurone through the open GlyR channels, thereby shifting Egly into the positive direction. When the membrane potential is stepped back to Vh, Egly is more positive than Vh. As GlyR ion channels are still open, Cl− now leaves the neurone and, thereby, it produces inwardly directed tail currents. In accordance with this, the tail currents are enlarged when the amplitude or the duration of the depolarizing voltage pulses is increased (cf. Fig. 6). A significant contribution of a depolarization-induced ‘passive’ Cl− flux, as previously reported in spinal cord neurones (Bührle & Sonnhof, 1983; Forsythe & Redman, 1988), is unlikely, because the depolarizing voltage pulses that we applied did not produce any significant tail currents in the absence of glycine. Therefore, we conclude that the depolarization-induced shift of Egly is primarily depending on Cl− flux through open GlyRs and not through a passive Cl− conductance of the membrane.

To our knowledge, glycine-evoked biphasic responses have not been described previously. Biphasic responses induced by GABA, being characterized by a transient hyperpolarization and a following sustained depolarization, have already been reported (Grover, Lambert, Schwartzkroin & Teyler, 1993; Staley, Soldo & Proctor, 1995; see also Perkins & Wong, 1996). Based on the permeability of GABAA receptors to Cl− and HCO3− (Bormann et al. 1987; Kaila, 1994), Staley et al. (1995) recently published a model that describes a possible mechanism of biphasic GABA responses in CA1 neurones of adult hippocampal slices. They suggest that a differential, activity-dependent collapse of the opposing concentration gradients of Cl− and HCO3− causes the depolarizing phase of GABA-evoked biphasic membrane potential changes. The findings that GABA indeed produced a large Cl− flux in rat sympathetic neurones (Ballanyi & Grafe, 1985) and that repetitive GABA application induced shifts in EGABA in acutely dissociated hippocampal neurones (Huguenard & Alger, 1986) support this model. We postulate a similar mechanism for biphasic responses evoked by glycine. Taken together, Cl− accumulation might be a more general mechanism to shift the polarity of postsynaptic potentials, regardless of whether they are mediated by GlyRs or by GABAA receptors.

An alternative mechanism explaining polyphasic GABA-induced responses is based on the contribution of several receptor subtypes (Perkins & Wong, 1996). We cannot rule out that the glycine-evoked biphasic responses in LSO neurones are mediated by two or more GlyR subtypes which have different permeabilities for Cl− and HCO3−. Indeed, there is a diversity of GlyR subtypes (Kuhse, Betz & Kirsch, 1995), yet there is no evidence that this molecular diversity reflects differences in permeability properties.

Age dependence of glycine-evoked responses in LSO neurones

Recently, it has been reported that glycine-induced potential changes in LSO neurones reversed their polarity from depolarizations to hyperpolarizations after about the first postnatal week (Kandler & Friauf, 1995). A similar age-dependent switch of glycine- or GABA-evoked membrane responses has been found in several brain areas (Ben-Ari et al. 1989; Luhmann & Prince, 1991; Wu et al. 1992; Owens et al. 1996). This shift may be attributed to a developmentally controlled change in the Cl− distribution, i.e. to an active Cl− uptake mechanism which is expressed in fetal and early postnatal stages (Ballanyi & Grafe, 1985; Rohrbough & Spitzer, 1996) and replaced by a passive Cl− distribution or by an active Cl− extrusion mechanism (Zhang, Spigelman & Carlen, 1991). Our findings do not contrast with this hypothesis, because 93% of the LSO neurones responded with an inward current or with a biphasic current comprising an inward component. The mean age of neurones which responded with an inward current did not significantly differ from that of neurones which showed a biphasic response. In addition, there was no age-dependent change in the relative frequency of neurones responding with inward currents (61 ± 5%) versus those responding with biphasic currents (35 ± 4%). The time period during which [Cl−]i becomes redistributed may vary from cell to cell, and neurones responding with biphasic currents may display an ECl more negative than Vm at earlier times than those responding merely with an inward current. As a consequence of glycine-induced intracellular Cl− accumulation, their Egly could shift into the positive direction. Only two of forty-four LSO neurones responded exclusively with an outward current, and it is possible that the glycine application was not sufficient to cause a Cl− accumulation in these two cells. It should also be noted that [Cl−]i depends on the efficiency of intracellular Cl− regulation, e.g. on a Cl− extrusion mechanism which shows an age-dependent expression in rat hippocampal neurones (Zhang et al. 1991). However, the fact that Egly dynamically shifts during GlyR activation hinders us from a precise determination of the time point at which the glycine-induced response changes its direction.

Simultaneous occurrence of GlyR activation and excitatory activity shifts Egly

LSO neurones receive convergent glutamatergic and glycinergic input (Wenthold, 1991). To analyse possible interactions between these inputs, we simulated their synchronous occurrence by applying depolarizing pulse trains in the presence of glycine-induced currents. These pulse trains were designed in an attempt to simulate a burst of action potentials, similar to the bursts which have been reported to occur spontaneously in developing auditory brainstem neurones (Kotak & Sanes, 1995). The voltage ramps, used to determine Egly, themselves produced a small shift of Egly. However, when the ramps were preceded by a conditioning pulse train, the shift of Egly was strongly enhanced, suggesting that during the activation of GlyRs, coincident depolarization drives Cl− into the neurone and, consequently, leads to a shift of the ECl into the positive direction. In turn, this shift results in a pronounced depolarization when GlyRs are subsequently activated. Such depolarizations could remove the voltage-dependent Mg2+ block of NMDA receptors, as reported for GABA in the hippocampus (Staley et al. 1995), that would enhance glutamatergic excitation and Ca2+ influx through these receptors. In addition, glycine-evoked depolarizations could lead to the activation of voltage-dependent Ca2+ channels and to an additional Ca2+ influx as it was recently demonstrated for developing rat spinal cord neurones (Reichling et al. 1994). Since increases of [Ca2+]i play a crucial role in the establishment and refinement of synaptic connections (Malenka, Kauer, Perkel & Nicoll, 1989; Segal, 1993), the glycine-evoked depolarization might contribute to the maturation of synaptic connectivity during development. In conclusion, we suggest that depolarizations induced by glycine or GABA could be due to two different mechanisms: (1) a genetically determined mechanism, characterized by the expression of transporter molecules which mediate an active Cl− accumulation during early neuronal development; (2) a context-dependent mechanism, characterized by an acute shift in the Cl− gradient, facilitated by the contribution of HCO3− and potentiated by coincident neuronal activity which might be present throughout the whole life. Both mechanisms may coexist and be involved in neuronal maturation and synaptic plasticity.

Acknowledgments

This work was supported by grants from the Deutsche Forschungsgemeinschaft to K. H. B and J. W. D (De231/11-1) and to E. F. (SFB269-B5). We thank Professor P. Jonas for reading the manuscript and for helpful comments. We thank R. Dallwig and M. Frech for discussions and S. Bergstein for her excellent technical assistance.

References

  1. Backus KH, Friauf E. Effects of synchronous depolarization on glycine-induced currents in developing rat auditory brainstem neurons. Society for Neuroscience Abstracts. 1996;22:647. [Google Scholar]
  2. Ballanyi K, Grafe P. An intracellular analysis of γ-aminobutyric-acid-associated ion movements in rat sympathetic ganglions. Journal of Physiology. 1985;365:41–58. doi: 10.1113/jphysiol.1985.sp015758. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Barnes-Davis M, Forsythe ID. Pre- and postsynaptic glutamate receptors at a giant excitatory synapse in rat auditory brainstem slices. Journal of Physiology. 1995;488:387–406. doi: 10.1113/jphysiol.1995.sp020974. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Ben-Ari Y, Cherubini E, Corradetti R, Gaiarsa JL. Giant synaptic potentials in immature rat CA3 hippocampal neurones. Journal of Physiology. 1989;416:303–325. doi: 10.1113/jphysiol.1989.sp017762. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bormann J, Hamill OP, Sakmann B. Mechanism of anion permeation through channels gated by glycine and gamma-aminobutyric acid in mouse cultured spinal neurones. Journal of Physiology. 1987;385:243–286. doi: 10.1113/jphysiol.1987.sp016493. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Brickley SG, Cull-Candy SG, Farrant M. Development of a tonic form of synaptic inhibition in rat cerebellar granule cells resulting from persistent activation of GABAA receptors. Journal of Physiology. 1997;497:753–759. doi: 10.1113/jphysiol.1996.sp021806. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bührle CP, Sonnhof U. The ionic mechanism of the excitatory action of glutamate upon the membranes of motoneurones of the frog. Pflügers Archiv. 1983;396:154–162. doi: 10.1007/BF00615520. [DOI] [PubMed] [Google Scholar]
  8. Caird D, Klinke R. Processing of binaural stimuli by cat superior olivary complex neurons. Experimental Brain Research. 1983;52:385–399. doi: 10.1007/BF00238032. [DOI] [PubMed] [Google Scholar]
  9. Chesler M. The regulation and modulation of pH in the nervous system. Progress in Neurobiology. 1990;34:401–427. doi: 10.1016/0301-0082(90)90034-e. [DOI] [PubMed] [Google Scholar]
  10. Constantine-Paton M, Cline HT, Debski E. Patterned activity, synaptic convergence, and the NMDA receptor in developing visual pathways. Annual Reviews in Neuroscience. 1990;13:129–154. doi: 10.1146/annurev.ne.13.030190.001021. [DOI] [PubMed] [Google Scholar]
  11. Edwards FA, Konnerth A, Sakmann B, Takahashi T. A thin slice preparation for patch clamp recordings from neurones of the mammalian nervous system. Pflügers Archiv. 1989;414:600–612. doi: 10.1007/BF00580998. [DOI] [PubMed] [Google Scholar]
  12. Forsythe ID, Redman SJ. The dependence of motoneurone membrane potential on extracellular ion concentrations studied in isolated rat spinal cord. Journal of Physiology. 1988;404:83–99. doi: 10.1113/jphysiol.1988.sp017280. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Gaillard S, Dupont JL. Ionic control of intracellular pH in rat cerebellar Purkinje cells maintained in culture. Journal of Physiology. 1990;425:71–83. doi: 10.1113/jphysiol.1990.sp018093. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Goodman CS, Shatz CJ. Developmental mechanisms that generate precise patterns of neuronal connectivity. Neuron. 1993;10:77–98. doi: 10.1016/s0092-8674(05)80030-3. [DOI] [PubMed] [Google Scholar]
  15. Grover LM, Lambert NA, Schwartzkroin PA, Teyler TJ. Role of HCO3− ions depolarizing GABAA receptor-mediated responses in pyramidal cells of rat hippocampus. Journal of Neurophysiology. 1993;69:1541. doi: 10.1152/jn.1993.69.5.1541. [DOI] [PubMed] [Google Scholar]
  16. Hamill OP, Marty A, Neher E, Sakmann B, Sigworth FJ. Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflügers Archiv. 1981;391:85–100. doi: 10.1007/BF00656997. [DOI] [PubMed] [Google Scholar]
  17. Huguenard JR, Alger BE. Whole-cell voltage-clamp study of the fading of GABA-activated currents in acutely dissociated hippocampal neurons. Journal of Neurophysiology. 1986;56:1–18. doi: 10.1152/jn.1986.56.1.1. [DOI] [PubMed] [Google Scholar]
  18. Ito S, Cherubini E. Strychnine-sensitive glycine responses of neonatal rat hippocampal neurones. Journal of Physiology. 1991;440:67–83. doi: 10.1113/jphysiol.1991.sp018696. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Kaila K. Ionic basis of GABAA-receptor channel function in the nervous system. Progress in Neurobiology. 1994;42:489–537. doi: 10.1016/0301-0082(94)90049-3. [DOI] [PubMed] [Google Scholar]
  20. Kandler K, Friauf E. Pre- and postnatal development of efferent connections of the cochlear nucleus in the rat. Journal of Comparative Neurology. 1993;328:161–184. doi: 10.1002/cne.903280202. [DOI] [PubMed] [Google Scholar]
  21. Kandler K, Friauf E. Development of glycinergic and glutamatergic synaptic transmission in the auditory brainstem of perinatal rats. Journal of Neuroscience. 1995;15:6890–6904. doi: 10.1523/JNEUROSCI.15-10-06890.1995. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kil J, Kageyama GH, Semple MN, Kitzes MN. Development of ventral cochlear nucleus projections to the superior olivary complex in gerbil. Journal of Comparative Neurology. 1995;353:317–340. doi: 10.1002/cne.903530302. [DOI] [PubMed] [Google Scholar]
  23. Knoflach F, Backus KH, Giller T, Malherbe P, Pflimlin P, Möhler H, Trube G. Pharmacological and electrophysiological properties of recombinant GABAA receptors comprising the α3, β1 and γ2 subunits. European Journal of Neuroscience. 1992;4:1–9. doi: 10.1111/j.1460-9568.1992.tb00103.x. [DOI] [PubMed] [Google Scholar]
  24. Kotak VC, Sanes DH. Synaptically evoked prolonged depolarizations in the developing auditory system. Journal of Neurophysiology. 1995;74:1611–1620. doi: 10.1152/jn.1995.74.4.1611. [DOI] [PubMed] [Google Scholar]
  25. Kuhse J, Betz H, Kirsch J. The inhibitory glycine receptor: architecture, synaptic localization and molecular pathology of a postsynaptic ion-channel complex. Current Opinion in Neurobiology. 1995;5:318–323. doi: 10.1016/0959-4388(95)80044-1. [DOI] [PubMed] [Google Scholar]
  26. Kyrozis A, Reichling DB. Perforated-patch recording with gramicidin avoids artifactual changes in intracellular chloride concentration. Journal of Neuroscience Methods. 1995;57:27–35. doi: 10.1016/0165-0270(94)00116-x. [DOI] [PubMed] [Google Scholar]
  27. Luhmann HJ, Prince DA. Postnatal maturation of the GABAergic system in rat neocortex. Journal of Neurophysiology. 1991;65:247–263. doi: 10.1152/jn.1991.65.2.247. [DOI] [PubMed] [Google Scholar]
  28. Malenka RC, Kauer JA, Perkel DJ, Nicoll RA. The impact of postsynaptic calcium on synaptic transmission: Its role in long-term potentiation. Trends in Neurosciences. 1989;12:444–450. doi: 10.1016/0166-2236(89)90094-5. [DOI] [PubMed] [Google Scholar]
  29. Myers VB, Haydon DA. Ion transfer across lipid membranes in the presence of gramicidin A. II The ion selectivity. Biochimica et Biophysica Acta. 1972;274:313–322. doi: 10.1016/0005-2736(72)90179-4. [DOI] [PubMed] [Google Scholar]
  30. Nishimaru H, Iizuka M, Ozaki S, Kudo N. Spontaneous motoneuronal activity mediated by glycine and GABA in the spinal cord of rat fetuses in vitro. Journal of Physiology. 1996;497:131–143. doi: 10.1113/jphysiol.1996.sp021755. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Owens DF, Boyce LH, Davis MBE, Kriegstein AR. Excitatory GABA responses in embryonic and neonatal cortical slices demonstrated by gramicidin perforated-patch recordings and calcium imaging. Journal of Neuroscience. 1996;16:6414–6423. doi: 10.1523/JNEUROSCI.16-20-06414.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Perkins KL, Wong RKS. Ionic basis of the postsynaptic depolarizing GABA response in hippocampal pyramidal cells. Journal of Neurophysiology. 1996;76:3886–3894. doi: 10.1152/jn.1996.76.6.3886. [DOI] [PubMed] [Google Scholar]
  33. Reichling DB, Kyrozis A, Wang J, MacDermott AB. Mechanisms of GABA and glycine depolarization-induced calcium transients in rat dorsal horn neurons. Journal of Physiology. 1994;476:411–421. doi: 10.1113/jphysiol.1994.sp020142. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Rohrbough J, Spitzer NC. Regulation of intracellular Cl− levels by Na+-dependent Cl− cotransport distinguishes depolarizing from hyperpolarizing GABAA receptor-mediated responses in spinal neurons. Journal of Neuroscience. 1996;16:82–91. doi: 10.1523/JNEUROSCI.16-01-00082.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Sanes DH, Rubel EW. The ontogeny of inhibition and excitation in the gerbil lateral superior olive. Journal of Neuroscience. 1988;8:682–700. doi: 10.1523/JNEUROSCI.08-02-00682.1988. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Segal M. Calcium and neuronal plasticity. Israel Journal of Medical Sciences. 1993;29:543–548. [PubMed] [Google Scholar]
  37. Staley KJ, Soldo BL, Proctor WR. Ionic mechanisms of neuronal excitation by inhibitory GABAA receptors. Science. 1995;269:977–981. doi: 10.1126/science.7638623. [DOI] [PubMed] [Google Scholar]
  38. Wenthold RJ. Neurotransmitters of brainstem auditory nuclei. In: Altschuler RA, Bobbin RP, Clopton BM, Hoffman DW, editors. Neurobiology of Hearing: The Central Auditory System. New York: Raven Press; 1991. pp. 121–139. [Google Scholar]
  39. Wu W, Ziskind-Conhaim L, Sweet MA. Early development of glycine- and GABA-mediated synapses in rat spinal cord. Journal of Neuroscience. 1992;12:3935–3945. doi: 10.1523/JNEUROSCI.12-10-03935.1992. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Zhang L, Spigelman I, Carlen PL. Development of GABA-mediated, chloride-dependent inhibition in CA1 pyramidal neurones of immature rat hippocampal slices. Journal of Physiology. 1991;444:25–49. doi: 10.1113/jphysiol.1991.sp018864. [DOI] [PMC free article] [PubMed] [Google Scholar]

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

RESOURCES