| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Neurophysiology, Tohoku University School of Medicine and Core Research for the Evolutional Science and Technology Program, Sendai 980-8575, Japan
Submitted 9 May 2003; accepted in final form 11 July 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
; Legendre 2001). Recent molecular studies have shown that glycine receptors are expressed more extensively than previously thought, particularly in the higher brain of mammals, including the cerebral cortex, hippocampus, midbrain, and cerebellum (Chattipakorn and McMahon 2002; Elster et al. 1998; Kilb et al. 2002; Mangin et al. 2002; Martin and Siggins 2002; Sergeeva and Haas 2001). Characteristics of glycine receptors in these areas are regulated distinctly in terms of subunit compositions and their amounts as well as in terms of functional plasticity during or after development (Mangin et al. 2002). At an early stage of development, glycine and its agonists cause excitation rather than inhibition of receptive neurons because the concentration of Cl- in immature neurons is higher than that in developed neurons (Ehrlich et al. 1999; Kilb et al. 2002; Laube et al. 2002; Rivera et al. 1999). Such excitatory action of glycine mediated by subsynaptic or extrasynaptic glycine receptors may play an essential role in cellular and functional maturation of neurons (Flint and Kriegstein 1998; Legendre 2001). In the cerebellum, cellular immunoreactivity for glycine and the presence of transcripts for the neural glycine transporter (namely, GlyT2 subtype) have been described (Chen and Hillman 1993; Zafra et al. 1995). Both of them are considered essential for glycinergic transmission. Rampon et al. (1996) found that the immunoreactivity for glycine was positive both in neurons of deep cerebellar nuclei and in cerebellar Lugaro cells, one type of cortical inhibitory interneurons. Also, the mRNA transcripts for GlyT2 were detected in these cells in the cerebellum (Luque et al. 1995). Furthermore, electrical stimulation of Lugaro cells was found to evoke glycinergic synaptic currents in postsynaptic Golgi cells, the observation of which has confirmed functional glycinergic transmission in the cerebellum (Dumoulin et al. 2001).
The deep cerebellar nuclei of rodents can be divided into three subdivisions according to their location: the nucleus medialis, nucleus interpositus, and nucleus lateralis (Sastry et al. 1997; Sultan et al. 2002). Neurons in these deep cerebellar nuclei (DCN) receive input from three major extra-nuclear sources: excitatory mossy and climbing fibers and inhibitory Purkinje cell axons (Ito 1984; Sastry et al. 1997; Telgkamp and Raman 2002). The neurons in the DCN consist of three types: large glutamatergic projection neurons, smaller GABAergic projection neurons, and local interneurons (Aizenman et al. 2003; Anchisi et al. 2001; Czubayko et al. 2001; Pedroarena and Schwarz 2003). The interneurons are thought to colocalize both GABA and glycine as neurotransmitters, but little is known about their input and output connections (Baurle and Grusser-Cornehls 1997; Chen and Hillman 1993). Neither the synaptic currents in DCN neurons mediated by glycine nor the presence or properties of ionotropic glycine receptors on DCN neurons have been investigated in detail. Rather, spontaneous or evoked inhibitory synaptic currents (IPSCs) recorded in DCN neurons have been found to be completely abolished by the addition of bicuculline (Aizenman et al.1998; Ouardouz and Sastry 2000; Sastry et al. 1997), indicating that they were exclusively GABAergic and that glycinergic synapses, if any, were difficult to detect in the DCN.
In this study, the possible existence and functional properties of glycine receptors in DCN neurons were re-examined using the slice-patch technique and a rapid drug application system, the "Y tube" system, which had been modified for use in brain slice preparations (Kawa 2002a). To avoid possible regional differences among the DCN neurons, slices containing the nucleus interpositus were used, and neurons in the DCN with larger diameters were selected for whole cell recordings. The aims of this study were to determine whether functional ionotropic glycine receptors are expressed on DCN neurons and whether glycinergic postsynaptic currents can be evoked by high K+ or electrical stimulation of presynaptic elements. The results of this study confirmed the above-stated speculation to be true in the rats studied (postnatal age, day 2–14) and revealed relevant properties of glycine receptors and glycinergic transmission in the DCN. A preliminary report has appeared elsewhere (Kawa 2002b).
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
All experiments were carried out in accordance with the Guiding Principles of the Physiological Society of Japan. The procedures for preparing and preserving thin slices from a rat brain and for cleaning cells in the slices for patch-clamp recordings have been described in detail elsewhere (Kawa 2002a). Briefly, newborn male and female rats (Wister strain) on postnatal days 2–14 (P2–P14) were killed by decapitation after ether anesthesia, and the cerebellum was quickly dissected out of each rat and immersed for a few minutes in ice-cold bicarbonate-buffered saline. The tissue was cut sagittally into thin slices of 200 µm thickness with a vibrating slicer (DSK-1000, Dosaka, Kyoto, Japan). Each slice was transferred to a storage chamber containing oxygenated (95% O2-5% CO2) normal saline at 30°C. The cerebellar slices were then transferred to a recording chamber placed on the stage of a Zeiss Axioskop upright microscope. This chamber (volume, 
1 ml) was continuously perfused with oxygenated saline solution kept at room temperature (23–25°C). DCN in the cerebellum were easily identified with a low-magnification video camera (KP-140, Hitachi, Tokyo, Japan) under transillumination observation. Using a video-capture board (GV-VCP2M, I.O. Data, Kanazawa, Japan), images of the cerebellar nuclei and mesh lines covering the cerebellar slice were saved as photo files and referred to for further identification of the neurons (Fig. 1A). When observed with Nomarski optics (Zeiss, Obercochen, Germany) using a long-working-distance 40x water-immersion objective, the location of neurons for recording with a whole cell patch clamp could be identified clearly (Fig. 1D). Whole cell recordings were made from large DCN neurons (>20 µm diam) located in the nucleus interpositus. Hence, according to previously reported anatomical data (Anchisi et al. 2001; Pedroarena and Schwarz 2003), most of the recorded neurons were considered to be glutamatergic projection neurons. After each experiment, an image of the configuration of the recording system was taken using a low-magnification objective (5x) and saved as a photo file (Fig. 1C).
|
Electrical recordings
Whole cell current recordings were made from neurons in the DCN using patch pipettes with open resistances of 1.5–2 M
and an EPC-7 patch clamp amplifier with a 3-kHz internal filter (List Electronics, Darmstadt, Germany). The procedure used has been described in detail previously (Kawa 2002a). Briefly, currents and membrane potentials were monitored with an oscilloscope and were recorded on videotape using a PCM/VCR recording device (PCM-501/ES, Sony). Data were digitalized at 10 kHz using a Digidata 1320A interface (Axon Instruments) and were also displayed on a pen recorder (R62N3, Rikadenki, Tokyo, Japan). Patch pipettes were pulled from a 1.5-mm capillary glass in two stages using a vertical pipette puller (PC-10, Narishige, Tokyo, Japan) and filled with intracellular solution. During whole cell recordings, the access resistance of the electrode was occasionally monitored by applying rectangular voltage pulses (2 mV, 100 ms) and by measuring the capacitative current amplitude. The access resistance usually remained in the range of 2–5 M; the error in holding membrane potential caused by this was <1 mV. If recording conditions such as cell input resistance or access resistance changed or became unstable, the experiment was stopped. The liquid junction potentials (9.6 mV; inside negative) were determined and corrected by previously described procedures (Kawa 2002a). In some neurons, the I-V relationship of the glycine-induced currents was obtained using a saw-tooth voltage clamp (-60 ± 30 mV in 1 s), the voltage wave for which was generated by a function generator (FG-12113, NF Circuit Designing Block Co., Yokohama, Japan). In some experiments, glycinergic postsynaptic currents were evoked by electrical stimulation (voltage pulses of duration of 200 µs, intensity of 2–10 V, at 0.25 Hz) delivered to sites within the DCN through a glass micropipette filled with standard external saline (tip resistance, 20–50 M; Fig. 1C). To confirm the effectiveness of stimulation, double-pulse stimulation with a train interval of 50 ms was also used.
Solutions and drugs
The external standard saline used for slicing and incubation contained (in mM) 125 NaCl, 5 KCl, 2 CaCl2, 1 MgCl2, 1.25 NaH2PO4, 25 NaHCO3, and 25 glucose (pH 7.4 when bubbled with 95% O2-5% CO2). The "internal" (pipette) solution (referred to as Cs-methanesulphonate internal saline) consisted of (in mM) 134 Cs-methanesulphonate (CH3O3SCs), 6 CsCl, 10 EGTA, 2 MgCl2, 10 HEPES, and 2 Na2-ATP (adjusted to pH 7.3 with CsOH). In the high-potassium external solution (containing 20 mM KCl), NaCl was replaced with equimolar KCl. The estimated equilibrium potential for Cl- ions under the present experimental conditions is -67 mV. Unless otherwise stated, membrane currents and synaptic currents in DCN neurons were recorded at the holding potential of -40 mV, because excitatory and inhibitory synaptic currents at this potential appeared distinctly as inward and outward currents, respectively. Drugs were applied using a rapid application technique, the "Y tube" method, usually for 10–20 s with an interval of 5 min (Kawa 2002a). A programable pulse generator (SEN-3201, Nihon-Koden) controlled the magnetic valve of the "Y tube" system.
To block GABAA receptors, bicuculline methiodide (BCC, 10 µM; Sigma) was added to the external solution, while to block glutamatergic amino-3-hydroxy-5-methyl-isoxazol-propionate (APMA) receptors and glutamatergic N-methyl-D-aspartate (NMDA) receptors, 6-cyano-7-nitroquinoziline-2, 3-dione (CNQX, 5 µM; RBI) and D-(-)-2-amino-5-phosphonovaleric acid (APV, 50 µM, Tocris), respectively, were added to the external solution. In some experiments, the following agents were also added to the external perfusion solution: glycine (10–2,000 µM), L-alanine (1 mM), L-serine (1 mM), taurine (1 mM), and GABA (10 µM). Strychnine (1 µM) and TTX (1 µM) were also added when necessary to the external saline for blocking ionotropic glycine receptors and voltage-dependent Na+ channels, respectively. These drugs were obtained from RBI or Sigma. The drugs were dissolved in solution and kept frozen in aliquots, and they were thawed just before each experiment. CNQX was dissolved in dimethylsulfoxide (Wako) at a concentration of 5 mM and added to the perfusing saline at a final concentration of 5 µM in the dark.
Data analysis
The following software programs were used for analyses of amplitude distribution, time course of the synaptic currents, dose-response relations for the agonist, and electric charges carried by synaptic currents: Mini Analysis (Synaptosoft), Origin version 6 (Microcal Software), Excel 2000 (Microsoft), and KaleidaGraph (Synergy Software). All data are expressed as mean ± SD unless otherwise indicated. An unpaired Student's t-test and Welch's test were used to evaluate the statistical significance of differences between two groups being compared. For significant changes in the ratios, a test for the F distribution was used.
Histological study
After whole cell recordings, each slice was transferred to a small chamber and was fixed overnight with 10% formalin in a phosphate-buffered saline (PBS) and then gently rinsed four to five times with distilled water for several hours. During these processes, the chamber was shaken gently and continuously on a horizontal shaker (R-Imini, Taitec Co.). The preparation was transferred to a plastic dish containing distilled water (Falcon, Primaria 35-mm-diameter tissue culture dish, Becton Dickinson) and stained with Giemsa's solution for 2–3 min (final dilution, 30- to 40-fold). Each slice was rinsed in a plastic dish containing distilled water and transferred to a glass plate. After having been half-dried, each slice was embedded in synthetic mounting medium (Aquatex, Merk, Germany). Each slice was viewed under an upright microscope (AxioVision system, Carl Zeiss; objective 63x), and photographs of the DCN neurons and the sites of recording were taken using an attached digital camera (Fig. 1B).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
The DCN of the rat can be divided into three major subdivisions, nucleus medialis, nucleus interpositus, and nucleus lateralis, according to their locations. In parasagittal sections of the cerebellum under transparent illumination, deep cerebellar nuclei were clearly recognized as a pale mass (at P2–14, Fig. 1). At higher magnification under Nomarski optics (40x), the location of the recorded neurons in the DCN was readily confirmed by referring to grid lines covering the slice. In the present experiments, most of the whole cell recordings were made from large neurons in the nucleus interpositus (Fig. 1D). Boundaries between the DCN subnuclei were not always clear in slices from rats at P7 or younger. Recordings from neurons in the nucleus interpositus were attempted by carefully selecting an appropriate parasagittal slice in the incubation chamber. The location of the neurons was also checked by staining the slice with Giemsa's solution after recording (Fig. 1B). Most of the neurons recorded were presumably projection neurons in the nucleus, having a glutamatergic nature as judged from their morphology (namely, large oval soma with diameter of more than 20 µm and multiple dendritic shafts). More detailed analyses of these neurons and other smaller neurons using intracellular staining are needed to determine the intricate makeup of the neural circuit during the early postnatal period.
Glycine-induced membrane currents in DCN neurons
When glycine (100 µM) was applied from a "Y tube" to a neuron in the DCN, large outward currents were induced in the cell at the holding potential of -40 mV (Fig. 2A). Glycine was applied from a "Y tube" for 10–20 s with an interval of 5 min unless otherwise stated. During the application of glycine, the currents decayed slowly to a steady level, and the time courses of decay of currents in neurons at different postnatal ages were similar (e.g., P4 and P11 in Fig. 2). All of the 49 DCN neurons studied (P2–P13) responded to glycine (100 µM) by generating outward currents. The glycine-induced currents could be repeatedly observed in the same neuron without noticeable deterioration under the present experimental conditions (i.e., application of 100 µM glycine for 10–20 s with an interval of 5 min). Peak amplitudes of the observed currents in the 49 neurons ranged from 54 to 1,110 pA (460 ± 273 pA) and showed no obvious relation to postnatal age. In Fig. 2A (right), the neurons have been divided into four groups according to age, and whole cell membrane capacity of the neuron (top) and density of the glycine-induced current (as pA/pF; bottom) are shown. The latter was obtained by normalizing the peak amplitude of the current induced by glycine (100 µM) to the whole cell membrane capacity of the cell. There were no significant differences among the four groups of neurons in both parts (P < 0.05).
|
Figure 2B shows the dose-response relation of glycine-induced membrane currents in DCN neurons. The external medium used was the standard external saline containing TTX (1 µM) but containing none of the antagonists for the neuronal receptors. When glycine at a concentration of 1 mM or higher was applied to the neuron, the interval of the application was doubled to 10 min to ensure recovery from desensitization. The representative traces shown on the left were obtained from a DCN neuron at P6 under these conditions (holding potential, -40 mV). The dose-response relationship obtained from four neurons (at P5–P9) after normalization of the peak amplitude of glycine-induced currents in each cell to the control value (obtained by 1 mM glycine as shown by double circles; average value, 1,240 pA; n = 4) is shown on the right. The dotted line overlapping the data are an estimated sigmoidal relation, having an apparent dissociation constant of 170 µM and a Hill coefficient of 1.6.
I-V relationship of glycine-induced membrane currents in DCN neurons
To further characterize the glycine-induced currents in DCN neurons, the I-V relationship and the reversal potential of the currents were determined. Figure 3A shows the I-V relationship of glycine-induced currents obtained using a saw-tooth voltage clamp (voltage wave, -60 ± 30 mV over a period of 1 s). In this experiment, glycine at a low concentration (50 µM) was used to avoid desensitization of the glycine receptor. Furthermore, TTX was added to the external saline to block generation of voltage-dependent Na+ currents. The representative I-V relationship showed mild outward rectification, which could be approximated by a constant field equation (dotted line on the I-V relationship) and thus may be due to the intrinsic property of Cl- ions permeating through the open channel of the receptor. The reversal potential obtained using a saw-tooth voltage clamp was -62 ± 2 mV (n = 4). The reversal potential showed good agreement with the estimated Cl- equilibrium potential (-67 mV) if we take into consideration the fact that the internal concentration of Cl- (10 mM) tended to increase due to influx of the external saline (containing 136 mM Cl-) into the patch electrode during establishment of the whole cell configuration. In four other neurons (at P4–P9), the I-V relationship was obtained by measuring peaks of glycine-induced currents at various holding potentials (ranging from -50 to -70 mV). As the holding potential shifted to more negative values from -60 mV, the peak of glycine-induced currents gradually decreased and became inward at holding potentials below -64 mV. The I-V relationship of the currents showed mild outward rectification like in Fig. 3A. The average of the reversal potential was -62 ± 3 mV (n = 4), when measured after interpolation to the voltage axis. These results strongly indicate that currents induced by glycine in DCN neurons are carried predominantly by Cl-.
|
Pharmacological properties of glycine receptors in DCN neurons
Glycine receptors are also activated by substances other than glycine, such as taurine, L-alanine, and L-serine. To examine the potencies of these substances on the glycine receptors of DCN neurons, currents induced by 1 mM of each substance were recorded as a primary measure. Each agonist was applied only for 12–15 s with an interval of 5 min or more (Fig. 3B, top left). 
-Alanine can also activate glycine receptors, but it was not included in this study due to presumed simultaneous activation of GABAA receptors on the same neuron. The relative amplitude of agonist-induced currents to that of glycine (100 µM) in each cell was obtained and averaged (n = 4). The results presented in Fig. 3B (top right) show that glycine (100 µM) has the highest potency among the agonists examined followed by taurine (1 mM, relative potency of 97 ± 20%), L-alanine (1 mM, relative potency of 46 ± 11%), and L-serine (1 mM, relative potency of 23 ± 8%).
Strychnine is thought to be one of the specific blockers for ionotropic glycine receptors. In DCN neurons, the blocking effects of strychnine on whole cell currents induced by glycine or its agonists were examined (Fig. 3B, bottom). Representative traces in Fig. 3B (bottom left) show that strychnine at 1 µM suppressed the whole cell currents completely when induced by glycine (100 µM), taurine (1 mM), L-alanine (1 mM), or L-serine (1 mM). When the concentration of glycine was increased by 10-fold, definite, but still small, outward currents were detected in the presence of 1 µM strychnine (bottommost trace, glycine at 1 mM). It is likely that a high concentration of glycine can attenuate the blocking effect of strychnine since it is a competitive antagonist. Current amplitudes thus obtained are plotted in Fig. 3B (bottom right) after normalization to the control amplitude (which was obtained by 100 µM glycine). Interestingly, the currents induced by L-alanine (1 mM) or L-serine (1 mM) were very small but seemed inward. One possible explanation is that transporters capable of uptaking these amino acids may be electrogenic and induce inward currents under the experimental conditions. Alternatively, these amino acids might bind to some sites of ionotropic or metabotropic receptors on the neuron and activate inward currents or inhibit steady outward currents of the cell. Due to the small amplitudes (<10 pA), further analyses are needed.
Specificities of strychnine and bicuculline on glycine and GABAA receptors in DCN neurons
To determine the specificities of strychnine (a glycine receptor antagonist) and BCC (a GABAA receptor antagonist), their effects on the currents induced by glycine or by GABA were examined in the same DCN neuron. The technique used was two-step application with a "Y tube." In the experiment for which traces are shown on the left in Fig. 4Aa, glycine (100 µM) was applied to a DCN neuron from a "Y tube." Between 7 and 8 s after the start of application, the first-application saline containing glycine (100 µM) was changed to the second-application saline that contained either 0 or 1 µM strychnine in addition to glycine (100 µM). The thick trace in the figure was obtained with 0 µM strychnine, whereas the thin trace was obtained with 1 µM strychnine. The abrupt reduction in current traces at the time of switching of the external solution (i.e., between 7 and 8 s after the start of application) was due to change in flow of the solutions. As shown by control glycine- or GABA-induced current traces (thick traces in Fig. 4A), the current amplitude after the switching recovered rapidly to the original level. Thus the effect of change in flow of the external solution on the following analyses, if any, was thought to have been small. Blocking effects of strychnine were evaluated at 22 s after the start of the first application (i.e., at the right end of the trace displayed). For BCC, measurements were made using second-application saline containing either 0 or 10 µM BCC in addition to glycine (100 µM), as shown in the traces on the right of Fig. 4Aa. Results of similar experiments on GABA-induced currents are shown in Fig. 4Ab. The representative traces revealed that the currents induced by GABA (100 µM) had low sensitivity to strychnine (1 µM) but high sensitivity to BCC (10 µM). The lower graphs in Fig. 4A summarize the results obtained from four cells at P7–P11. The graphs show that the remaining amounts of glycine-induced currents after blockage by strychnine and BCC were 8 ± 4% (n = 4) and 93 ± 5% (n = 4), respectively, while the amounts of GABA-induced currents after blockade by strychnine and BCC were 85 ± 5% (n = 4) and 1 ± 2% (n = 4), respectively. These results confirmed the specificities of strychnine and BCC on the glycine and GABAA receptors, respectively, of the present preparation and thus provided a pharmacological basis for isolation of glycinergic currents in the following experiments. The specificities are consistent with those reported previously (Jonas et al. 1998; Russier et al. 2002; Sergeeva 1998; Turecek and Trussell 2002).
|
Effects of picrotoxin on glycine receptors in DCN neurons
Picrotoxin is a classical noncompetitive GABAA receptor antagonist, but it also blocks homomeric glycine receptors consisting solely of 
subunits at a low dose (Mangin et al. 2002). The effects of picrotoxin on glycine receptors of DCN neurons were studied by a two-step application technique using a "Y tube." Representative traces shown in the left of Fig. 4B are currents obtained with 0 µM picrotoxin (thick trace) and those obtained with 10 µM picrotoxin (thin trace). The time course of the blocking effects of picrotoxin is shown on the right. The difference between the two current traces was calculated and normalized to the value of the control trace (i.e., thick trace). The difference shows that the blocking effect reached a steady level at about 6 s after the start of picrotoxin application. The blocking of currents (in % measured at the steady level) was 45 ± 8% (n = 4) when measured from DCN neurons at P2–P3. This suggests that the glycine receptors expressed on DCN neurons at P2–P3 are heterogeneous in their subunit composition, and about 45% may be homomeric receptors, consisting solely of subunits and being sensitive to picrotoxin, while 55% of the receptors have presumably formed heteromers consisting of and subunits and thus show resistance to picrotoxin.
ACh- and high K+-induced synaptic currents observed in DCN neurons
Figure 5 shows that glycinergic synaptic transmission is actually present in DCN neurons. After blocking GABAergic synaptic currents by BCC (10 µM), the frequency of spontaneous synaptic currents, either inward or outward, was generally low in DCN neurons (measured over a period of 20–60 s from each neuron at P6–P9). The mean frequency of spontaneous inward synaptic currents was 0.29 ± 0.10 Hz (n = 6). To facilitate detection of the outward synaptic currents shown in Fig. 5A, the holding potential of the neuron during recording was kept at -10 mV. Most recordings show that no events or only one event occurred during a period of 20 s. Accordingly, the mean frequency of the spontaneous outward synaptic currents was only 0.03 ± 0.02 Hz (n = 6 at P6–P9). When 100 µM ACh was applied around the neuron from a "Y tube," the frequency of the spontaneous outward synaptic currents transiently increased (Fig. 5Aa; a part is expanded in bottom of figure). In a total of 10 DCN neurons studied (3 neurons at P5, 3 neurons at P7, 3 neurons at P8, and 1 neuron at P11), 6 neurons showed such facilitation of synaptic currents by 100 µM ACh (3 neurons at P5, 1 neuron at P7, 1 neuron at P8, and 1 neuron at P11). The ACh-induced synaptic currents were sensitive to strychnine (1 µM; Fig. 5Ab), thus indicating a glycinergic nature (n = 3). The facilitatory effect of ACh is thought to be due to the presence of nicotinic receptors in presynaptic glycinergic neurons (cf. Kawa 2002a). Thus it is likely that glycinergic neurons or fibers are actually present in cerebellar nuclei of the rat and that they make synapses on the DCN neurons.
|
Next, as shown in Fig. 5B, whole cell membrane currents were recorded in the presence of 5 µM CNQX, 50 µM APV, and 10 µM BCC, which blocked glutamatergic APMA receptors, glutamatergic NMDA receptors, and GABAA receptors, respectively. When external saline containing 20 mM K+ was applied around the DCN, whole cell currents of the neuron showed slow but large inward deflection (Fig. 5B). Many spike-like outward currents overlapping the slow inward currents were observed. The spike-like currents seemed to consist of summated synaptic currents as judged from their amplitudes and time courses. These spike-like currents disappeared in the presence of strychnine (1 µM). It is thus likely that the spike-like outward currents are glycinergic postsynaptic currents in the DCN neuron, evoked by presynaptic depolarization by 20 mM K+. Forty DCN neurons (P2–P14) were challenged with high K+ application, and definite spike-like currents were detected from 14 neurons. No remarkable differences between rates of detection were found in three groups of DCN neurons at different postnatal ages (P2–P5, P6–P9, and P10–P14; Fig. 5B, bottom). For more qualitative determination, the time integral of the spike-like postsynaptic currents (i.e., electric charges carried by these synaptic currents) was obtained using the software programs Mini Analysis and Excel 2000. From 5 to 15 s after the start of K+ application, synaptic currents evoked in each neuron having a peak amplitude of more than 5 pA (above the noise level during K+ application) and half-width of 6–30 ms were detected. The total number of synaptic events thus measured showed considerable intercell variation even among these 14 neurons, with a range of 9–108 events per 10 s (37 ± 30 events per 10 s). Average electric charges were calculated by using the software programs after cells had been divided into three groups of different postnatal ages (Fig. 5B, bottom right). From P2–P5 to P6–P9, the average value increased significantly (P < 0.05). This significant increase may reflect development in the pre- or postsynaptic elements during this period or it may reflect an increased sensitivity of the presynaptic terminals to 20 mM K+, possibly due to a hyperpolarizing shift of the resting potential during development.
Glycinergic synaptic currents evoked in DCN neurons
In the presence of three specific antagonists for blocking AMPA, NMDA, and GABAA receptors (5 µM CNQX, 50 µM APV and 10 µM BCC, respectively), electrical stimulation could still evoke postsynaptic currents in the DCN neurons. As can be seen in a representative current trace shown in Fig. 6A, the synaptic currents rose to a peak in a period of 1.5–2.5 ms and then decayed to the resting level in about 80 ms (at -40 mV, at 23–25°C), the time course of which can be approximated by a single exponential curve. When the neuron was hyperpolarized, the decay time course became faster, while it became slower when the neuron was depolarized. The dependence of half-decay time of the synaptic currents on holding potential is shown on the left of Fig. 6A. The slope of the regression line (dotted line) is 0.16 s/V. Similar voltage dependence has been observed in glycine receptors of rat auditory brain stem neurons (Kungel and Friauf 1997). In subsequent experiments, double-pulse electrical stimulation was usually applied (duration of 200 µs, intensity of 2–10 V, pulse interval of 50 ms, train interval of 4 s). This is because more synaptic currents could be observed during a given time, thus enabling faster estimation of the threshold of stimulus intensity. In the present study, only the first responses evoked by double pulse-stimulation were analyzed. Small-sized neurons (with diameters of 12 µm or less) at a distance of <200 µm from the recorded neuron were primarily chosen for sites of stimulation. The intensity of stimulation used was a twofold value of the threshold intensity for evoking minimal glycinergic synaptic currents, and it was kept constant during the recording. If the threshold intensity exceeded 5 V, a search for a new site for stimulation in the DCN was made. To characterize the glycinergic synaptic currents thus evoked, the I-V relationship was obtained by measuring the synaptic currents at various levels of holding potential (+20 to -90 mV). The I-V relationship showed mild outward rectification (Fig. 6A, right) like that of the glycine-induced whole cell currents. By an interpolation to the voltage axis, the reversal potential of the synaptic currents was determined to be -62 ± 2 mV (n = 4), which is consistent with those of glycine-induced membrane currents. Furthermore, the synaptic currents evoked by electrical stimuli were completely blocked by strychnine (1 µM) but recovered slowly after washing in external saline containing 5 µM CNQX, 50 µM APV, and 10 µM BCC (Fig. 6B, left). In the presence of TTX (1 µM), these synaptic currents were also reversibly blocked (Fig. 6B, right), indicating that the glycinergic transmission depended on generation of action potentials in the slice. The slow time course of recovery from the blocking by strychnine, compared with that of TTX, may reflect the low rate of unbinding of strychnine at the glycine receptor. In the present study, reversal of strychnine blockage was examined, when necessary, after washing with control saline for more than 40 min. Thus it seems reasonable to conclude that the synaptic currents observed in the DCN neurons are mediated by activation of ionotropic glycine receptors. In this series of experiments, 7 of the 15 DCN neurons at P7–P10 studied showed glycinergic IPSCs evoked by electrical stimuli. Electrical stimuli (duration of 200 µs; maximum intensity, 5 V) for each neuron were applied to four or less surrounding sites. The rate of detection of glycinergic inhibitory postsynaptic currents (IPSCs) might have been higher if a more rigorous search for stimulation sites had been carried out. Detailed morphology of these neurons and changes in detection rate during postnatal development remain to be studied.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
; Czubayko et al. 2001), but they seem to be distinguishable due to the smaller sizes of GABAergic projection neurons than those of glutamatergic projection neurons (Aizenman et al. 2003; Pedroarena and Schwarz 2003). Involvement of small local interneurons (
15 µm diam; cf. Aizenman et al. 2003) in this study was unlikely; this was also judged from the value of whole cell membrane capacity of the cells (
30 pF; Fig. 2A). Thus it is likely that glycinergic transmission is physiologically functional in principal DCN neurons (i.e., projection neurons) of the rat. Properties of glycinergic transmission in DCN
The rate of detection of glycinergic synaptic currents in DCN neurons by using high K+ saline was only about 35% (Fig. 5B). This low detection rate may explain the failure in previous studies to detect glycinergic synaptic currents in DCN neurons (e.g., Pedroarena et al. 2001). The low detection rate presumably reflects lower prevalence of glycinergic synapses than GABAergic synapses on DCN neurons (Aizenman et al. 1998; Anchisi et al. 2001; Momiyama and Takahashi 1994; Ouardouz and Sastry 2000). The magnitude and prevalence of glycinergic synaptic currents may be influenced by their developmental stages (Gao et al. 2001; Russier et al. 2002) and also by the amount of specific glycine transporters expressed (Luque et al. 1995; Zafra et al. 1995). In this regard, there is an interesting possibility that glycine might be co-released with GABA from existent "GABAergic" terminals such as those of GABAergic interneurons and generate synaptic currents with differed time courses (Gao et al. 2001; Russier et al. 2002).
Molecular subtypes of glycine receptors
Previous studies have shown that mRNAs or subunit molecules of glycine receptors are present in the DCN of the rat. For example, 1, 2, and subunit mRNAs were detected in all of the three subnuclei of DCN of the neonatal rat at P7 (Sato et al. 1992). In another study, 1 and subunits of the glycine receptor mRNA were detected in the lateral nucleus of DCN of adult rats (Malosio et al. 1991). Intense labeling with a subunit probe was seen in the whole cerebellum after embryonic day 19, but 2 hybridization signals were seen only in the DCN neurons during the period from P0–P15 (Malosio et al. 1991). Morphological studies on the development of DCN neurons in the rat have suggested that maturation of the deep cerebellar nuclei appear to be at an advanced stage by the time of birth (Altman and Bayer 1978; Ito 1984; Sastry et al. 1997). These results, however, should be interpreted carefully because it has been shown that massive GABAergic synapses are formed on DCN neurons during the postnatal period from P1 to P15 (Garin and Escher 2001) and that the metabolic state of DCN neurons also undergoes dramatic changes after birth (Console-Bram et al. 1996).
In the developing spinal cord, a switch occurs of receptor subunit expression from mostly 2 homomers at birth to 1 heteromers at around 20 days after birth (Laube et al. 2002). A similar switch in glycine receptors was also found in brain stem motoneurons and in the substantia nigra (Mangin et al. 2002; Singer et al. 1998). In contrast, it has been reported that there is no evidence of a molecular switch in midbrain neurons (Garcia-Alcocer et al. 2001). It is notable that subtypes of glycine receptors show different sensitivities to picrotoxin, a blocker of agonist-gated Cl- channels (Legendre 2001; Mangin et al. 2002). For example, adult-type 1 heteromeric glycine receptors are resistant to picrotoxin, whereas juvenile-type homomeric 2 receptors are sensitive to picrotoxin at a low concentration (10 µM). In the present preparation, picrotoxin-resistant glycine receptors, presumably the adult type, were already expressed on DCN neurons at P2–P3 (Fig. 4B), which is consistent with the results of an in situ hybridization study (Sato et al. 1992). How this fraction of glycine receptors changes during maturation is a new concern in DCN and the striatum (Sergeeva and Haas 2001).
Localization of glycine receptors
The results of this study indicate that the majority of glycine receptors exist in the extrasynaptic region of DCN neurons, as judged from a comparison of the amplitudes of glycine-induced whole cell currents (range, 300–600 pA at 100 µM glycine; cf., Figs. 2A and 4A) and those of synaptically evoked glycinergic currents (range, 10–40 pA; Figs. 5B and 6B). Glycine receptors at extrasynaptic sites may also act as taurine receptors, which modulate the resting membrane currents, or as trophic receptors during development (Flint et al. 1998; Furuya et al. 2000; Mangin et al. 2002; Mori et al. 2002; Tapia et al. 2000). It has been shown that DCN neurons in adult rat cerebella expressed 1 and 2 subunit mRNAs both in somata and dendrites, while neurons of the cerebellar cortex expressed these mRNAs only in somata (Racca et al. 1998). It has also been revealed that anchoring of glycine receptors at synaptic sites is under refined metabolic control and regulated by specific interaction between molecules of the receptor and the cytoskeleton (Kneussel and Betz 2000; Legendre 2001). These observations suggest that there are intricate mechanisms for controlling glycine receptor expression and localization during and after development.
Origin of glycinergic innervation and its functional roles
Based on results of histochemical observation, some types of neurons in the DCN, particularly local interneurons, are believed to contain GABA, glycine, or both as neurotransmitters (Baurle and Grusser-Cornehls 1997; Rampon et al. 1996). Glycine transporters of neuron-specific type (GlyT2), a marker for glycinergic terminals, have also been detected on neurons of the DCN (Luque et al. 1995; Zafra et al. 1995). The presence of glycine in local interneurons has been confirmed by electron microscopy (Chen and Hillman 1993). Thus in this study, the most probable origin of presynaptic elements containing and releasing glycine is neurons in the DCN, particularly local interneurons. This speculation is consistent with the results obtained using electrical stimuli (delivered to a small neuron in the DCN; Fig. 6). The possibility of glycinergic Purkinje axons seems unlikely from results of previous studies (Aizenman et al. 1998; Ito 1984; Ouardouz and Sastry 2000; Telgkamp and Raman 2002). Therefore the most probable source of glycine as a neurotransmitter on DCN neurons is interneurons of the DCN, which may release pure glycine or glycine and GABA simultaneously. It is relevant in this regard to note the phenomenon of co-localization of GABA and glycine in the same terminal and changes in the neurotransmitter from GABA to glycine or vice versa at certain synapses during development or under some diseased conditions can occur (Gao et al. 2001; Garcia-Alcocer et al. 2001; Singer and Berger 2000; Sultan et al. 2002; Turecek and Trussell 2002; Zhou et al. 2002). Additionally, in the cerebellar cortex of the rat, pure and mixed glycinergic and GABAergic transmission has been proved elecrophysiologically in Golgi cells (Dumoulin et al. 2001).
The roles of these glycine receptors and glycinergic transmission in the brain seem crucial for proper neural function and for regulated development of the brain (Dutschmann and Paton 2002; Legendre 2001; Zhou et al. 2002). In many immature neurons, glycine causes excitation instead of inhibition, which may trigger several Ca2+-dependent phenomena essential for the developmental process, such as neuronal proliferation, migration, and synaptic maturation (Flint et al. 1998; Kirsch and Betz 1998; Legendre 2001). The molecular mechanism underlying the shift from an excitatory nature to an inhibitory nature is thought to be the development of a K+/Cl- co-transporter that extrudes intracellular Cl- and promotes fast hyperpolarizing postsynaptic inhibition in the brain (Eilers et al. 2001; Mikawa et al. 2002; Rivera et al. 1999). When and how such a shift occurs in DCN neurons, including those at the embryonic stage, are issues for a future study. The results of this study should contribute to elucidation of the functional roles of glycine, which has so far been thought to be a minor neurotransmitter in the cerebellum but may be essential under physiological and pathological conditions in developing as well as in matured nervous systems.
|
DISCLOSURES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: K. Kawa, Dept. of Neurophysiology, Tohoku Univ. Graduate School of Medicine, 2-1, Seiryo-machi, Aoba-ku, Sendai 980-8575, Japan (E-mail: kawa-k{at}mail.cc.tohoku.ac.jp).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
Aizenman CD, Manis PB, and Linden DJ. Polarity of long-term synaptic gain change is related to postsynaptic spike firing at a cerebellar inhibitory synapse. Neuron 21: 827-835, 1998.[ISI][Medline]
Altman J and Bayer SA. Prenatal development of the cerebellar system in the rat. I. Cytogenesis and histogenesis of the deep nuclei and the cortex of the cerebellum. J Comp Neurol 179: 23-48, 1978.[ISI][Medline]
Anchisi D, Scelfo B, and Tempia F. Postsynaptic currents in deep cerebellar nuclei. J Neurophysiol 85: 323-331, 2001.
Baurle J and Grusser-Cornehls U. Differential number of glycine- and GABA-immunopositive neurons and terminals in the deep cerebellar nuclei of normal and Purkinje cell degeneration mutant mice. J Comp Neurol 382: 443-458, 1997.[ISI][Medline]
Chattipakorn SC and McMahon LL. Pharmacological characterization of glycine-gated chloride currents recorded in rat hippocampal slices. J Neurophysiol 87: 1515-1525, 2002.
Chen S and Hillman DE. Colocalization of neurotransmitters in the deep cerebellar nuclei. J Neurocytol 22: 81-91, 1993.[ISI][Medline]
Console-Bram LM, Fitzpatrick-McElligott SG, and McElligott JG. Distribution of GAP-43 mRNA in the immature and adult cerebellum: a role for GAP-43 in cerebellar development and neuroplasticity. Brain Res Dev Brain Res 95: 97-106, 1996.[Medline]
Czubayko U, Sultan F, Thier P, and Schwarz C. Two types of neurons in the rat cerebellar nuclei as distinguished by membrane potentials and intracellular fillings. J Neurophysiol 85: 2017-2029, 2001.
Dumoulin A, Triller A, and Dieudonne S. IPSC kinetics at identified GABAergic and mixed GABAergic and glycinergic synapses onto cerebellar Golgi cells. J Neurosci 21: 6045-6057, 2001.
Dutschmann M and Paton JF. Glycinergic inhibition is essential for coordinating cranial and spinal respiratory motor outputs in the neonatal rat. J Physiol 543: 643-653, 2002.
Ehrlich I, Lohrke S, and Friauf E. Shift from depolarizing to hyperpolarizing glycine action in rat auditory neurones is due to age-dependent Cl- regulation. J Physiol 520: 121-137, 1999.
Eilers J, Plant TD, Marandi N, and Konnerth A. GABA-mediated Ca2+ signalling in developing rat cerebellar Purkinje neurones. J Physiol 536: 429-437, 2001.
Elster L, Banke T, Kristiansen U, Schousboe A, and Wahl P. Functional properties of glycine receptors expressed in primary cultures of mouse cerebellar granule cells. Neuroscience 84: 519-528, 1998.[ISI][Medline]
Flint AC, Liu X, and Kriegstein AR. Nonsynaptic glycine receptor activation during early neocortical development. Neuron 20: 43-53, 1998.[ISI][Medline]
Furuya S, Tabata T, Mitoma J, Yamada K, Yamasaki M, Makino A, Yamamoto T, Watanabe M, Kano M, and Hirabayashi Y. L-serine and glycine serve as major astroglia-derived trophic factors for cerebellar Purkinje neurons. Proc Natl Acad Sci USA 97: 11528-11533, 2000.
Gao BX, Stricker C, and Ziskind-Conhaim L. Transition from GABAergic to glycinergic synaptic transmission in newly formed spinal networks. J Neurophysiol 86: 492-502, 2001.
Garcia-Alcocer G, Garcia-Colunga J, Martinez-Torres A, and Miledi R. Characteristics of glycine receptors expressed by embryonic rat brain mRNAs. Proc Natl Acad Sci USA 98: 2781-2785, 2001.
Garin N and Escher G. The development of inhibitory synaptic specializations in the mouse deep cerebellar nuclei. Neuroscience 105: 431-441, 2001.[ISI][Medline]
Ito M. The Cerebellum and Neural Control. Raven Press. New York, 1984.
Jonas P, Bischofberger J, and Sandkuhler J. Corelease of two fast neurotransmitters at a central synapse. Science 281: 419-424, 1998.
Kawa K. Acute synaptic modulation by nicotinic agonists in developing cerebellar Purkinje cells of the rat. J Physiol 538: 87-102, 2002a.
Kawa K. Glycine-induced responses from neurones in cerebellar nuclei of the neonatal rat: a slice-patch study. Neurosci Res 26: S18, 2002b.
Kilb W, Ikeda M, Uchida K, Okabe A, Fukuda A, and Luhmann HJ. Depolarizing glycine responses in Cajal-Retzius cells of neonatal rat cerebral cortex. Neuroscience 112: 299-307, 2002.[ISI][Medline]
Kirsch J and Betz H. Glycine-receptor activation is required for receptor clustering in spinal neurons. Nature 392: 717-720, 1998.[Medline]
Kneussel M and Betz H. Clustering of inhibitory neurotransmitter receptors at developing postsynaptic sites: the membrane activation model. Trends Neurosci 23: 429-435, 2000.[ISI]</
