| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Journal of Physiology (2002), 545.1, pp. 169-181
© Copyright 2002 The Physiological Society
DOI: 10.1113/jphysiol.2002.026534
subunits |
ABSTRACT |
|---|
 TopAbstract IntroductionMethodsResultsDiscussionReferences |
|---|
Each GABAA receptor consists of two
(Resubmitted12 June 2002; accepted after revision 12 September 2002; first published online 4 November 2002)
Corresponding author A. B. Brussaard: Department of Experimental Neurophysiology, Research Institute Neuroscience, Faculty of Earth and Life Sciences, Centre for Neurogenomics and Cognitive Research, Vrije Universiteit Amsterdam, De Boelelaan 1085, 1081 HV Amsterdam, The Netherlands. Email: brssrd{at}cncr.vu.nl
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Neuronal network activity is shaped by the combined action of inhibitory and excitatory synapses. Most fast inhibitory synapses in the central nervous system express the GABAA receptor, which occurs in a variety of structurally and functionally different subtypes (Hevers & Lüddens, 1998). Detailed knowledge of the spatial and temporal distribution of the individual GABAA receptor subtypes is required for a proper understanding of the impact of inhibitory neurotransmission on brain functioning.
The native GABAA receptor is a heteropentameric protein, most often consisting of two , two 
and a fifth subunit, which can be either a third , a 
, a 
or an 
subunit (Farrar et al. 1999; Baumann et al. 2001; Klausberger et al. 2001). Different isoforms of the , and subunits have been described (Hevers & Lüddens, 1998). The GABAA receptor subunit expression differs not only between various brain areas (Wisden et al. 1992; Pirker et al. 2000), but also changes during development (Laurie et al. 1992; Tia et al. 1996). In addition, it can change under pathological conditions, like epilepsy (Brooks-Kayal et al. 1998; Banerjee et al. 1999). It has been proposed that specific subunits are preferentially associated with the induction of certain types of behaviour, including anxiety (Rudolph et al. 1999; McKernan et al. 2000; Gulinello et al. 2001).
Earlier studies have indicated clear changes in GABAA receptor subunit expression during early postnatal development of the rat visual cortex (Laurie et al. 1992; Fritschy et al. 1994). If one considers the alterations in GABAA receptor subunit mRNA expression in layers I-IV of the rat visual cortex in relation to the moment of eye opening at postnatal day 13 (around pn13), it is remarkable that the alterations in subunit expression appear to be more pronounced amongst the subunits than amongst the other GABAA receptor subunits. The clearest changes are a developmental decrease in 3 and an increase in 1 and, to a lesser extent, in 4 subunit mRNA expression (Laurie et al. 1992).
Here we tested the extent to which such alterations in mRNA expression are reflected in changes in the functioning of GABAA receptors in the postsynaptic membrane. To this end, whole-cell voltage-clamp recordings in acutely prepared visual cortex slices were combined with a pharmacological screening for the function and contribution of the GABAA receptor subunits at the postsynaptic membranes. Since the developmental changes in GABAA receptor subunit expression were especially prominent amongst the subunits, we have focused here on subunit expression.
First, we assessed the relative contribution of 1, 3 and 4 subunits to synaptic responses before (pn6) and at the end (pn21) of neonatal development, using subunit-specific pharmacology. Then we analysed the kinetic contribution of each subunit in its native setting. We confirmed the results obtained with the pharmacological experiments for the 1 subunit by two independent genetic methods: 1 knockout mice and 1 antisense oligonucleotide deletion. Our data are in line with a general concept that in particular subunit-expression is causally related to the decay kinetics of sIPSCs (Vicini et al. 2001).
|
METHODS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Animals and preparations
Non-anaesthetized Wistar rats were decapitated, and their brains quickly removed and placed in ice-cold artificial cerebrospinal fluid (ACSF, mM: 125 NaCl, 25 NaHCO3, 3 KCl, 1.2 NaH2PO4, 2.4 CaCl2, 1.3 MgSO4, 10 D(+)-glucose (carboxygenated with 5 % CO2-95 % O2, 304 mosmol l-1, pH 7.4)). Coronal sections (400 µm thick) of the visual cortex were cut using a Leica VT1000S vibratome slicer. Slices were stored for up to 8 h in continuously carboxygenated ACSF at room temperature.
In some experiments, tissue cultures were used. These were made from the visual cortex of pn6/7 Wistar rat pups as previously described (Baker & Van Pelt, 1997). Two pieces of cortex were incubated, with their ventral sites touching, for 2 weeks in a chemically defined medium. During the second week of culturing, antisense oligonucleotides directed against the GABAA receptor 1 subunit were present in some cultures. The oligonucleotide sequence used as antisense probe was 5'-GCT GGT TGC TGT AGG-3'. As a control, we used a random sequence (5'-GTC GGG GTC TCT CTG-3'). The oligonucleotides were present at a concentration of 10 µM in the culture medium and refreshed every 48 h (see also Brussaard et al. 1997).
Mice lacking the 1 subunit of the GABAA receptor were of a mixed 50 % C57BL6-50 % 129SvEv background, as previously described (Sur et al. 2001) and measured during adulthood (> 3 months). Both wild-type and knockout mice were sedated by ketamine (I.P., ~0.1 mg (g body weight)-1) 20 min prior to decapitation, as required during transport from the stable to the lab, by the Animal Welfare Committee of the Vrije Universiteit Amsterdam, in accordance with the Dutch law. All experimental methods were approved by the Animal Welfare Committee of our university.
Cellular recordings
In situ whole-cell voltage-clamp recordings were made of randomly selected neurons of layer II-III of the visual cortex using an Axopatch 200A amplifier (Axon Instruments, Union City, CA, USA) and borosilicate glass (Harvard Apparatus Ltd, UK) electrodes with tip resistances of 2-5 M
. All experiments were performed at 33 °C using ACSF with 20 µM DNQX and 20 µM APV (both from Sigma) to block the ionotropic glutamate receptors. The pipettes were filled with (mM): 135 CsCl, 1 CaCl2, 10 EGTA, 10 Hepes, 2 MgATP, 296 mosmol l-1, pH 7.2 (with CsOH). All spontaneous IPSCs (sIPSCs) could be blocked by the specific GABAA receptor antagonist bicuculline methiodide (10 µM, Sigma). For zolpidem (Tocris, UK), bretazenil (a kind gift from Roche Nederland, NL) and flunitrazepam (Bufa B.V., NL) we used stock solutions in DMSO. The final DMSO concentration was maximally 0.l %. Furosemide (Sigma) was freshly dissolved each day using equimolar NaOH. There was no tetrodotoxin (TTX) added to the extracellular solution, implying that the sIPSCs measured were a mixture of action potential-evoked and miniature IPSCs.
Data analysis
Experimental data were stored on digital tapes and analysed later using the Computer Disk Recorder v1.3 and the Whole Cell Program v2.3 (both kindly provided by Dr J. Dempster, Strathclyde University, UK). Sampling rate was 5 kHz, with 1 kHz low-pass filtering. Events were detected using current and time thresholds. All automatically detected events were individually checked. Only single events with a sharp rising phase starting from a stable base line were accepted. Previous work by others on the rise times of sIPSCs in melanotropes of Xenopus laevis (Borst et al. 1994) showed that rise times of more than 1 ms may occur, even in the absence of dendritic filtering. However, variation in rise time of sIPSCs may also arise from dendritic filtering. To exclude dendritically filtered sIPSCs from further analysis, within each experiment, a correlation diagram was made of the 10-90 % rise time versus peak current of individual sIPSCs. The maximal 10-90 % rise time, at which no correlation between rise time and peak current was observed, was determined per experiment. At pn6, 94 % of all sIPSCs were accepted, having a maximal rise time of 3.16 ± 0.46 ms. At pn21, 96 % of all sIPSCs were accepted, having a maximal rise time of 2.78 ± 0.68 ms. Moreover, in our recordings there was no difference in the compensated RC-time constants between pn6 (RC 
= 223 ± 138 µs) and pn21 (210 ± 139 µs; P > 0.4; Mann-Whitney test). The resistance 
capacitance (RC) time constants were calculated by multiplication of the compensated series resistance (pn6: 9.4 ± 5.4 M; pn21: 9.6 ± 4.3 M) and the membrane capicitance (pn6: 24.4 ± 8.8 pF; pn21: 25.3 ± 17.5 pF). Both parameters were read off the amplifier. The average input resistance was 752 ± 244 M (pn6) and 543 ± 278 M (pn21) (all values are means ± S.D.).
Therefore, it is unlikely that the variations in recording conditions and/or subsequent analysis introduced bias towards one of the two ages under investigation. One might argue that setting the rise time criterion for analysis per individual experiment, may have biased selection of sIPSCs depending on the developemental stage being recorded from. Instead we could have used one objective strict rise time criterion (e.g. < 0.6 ms, see Nusser et al. 2001). Application of this criterion to our data would have rejected most of the events and would have introduced a bias towards neurons having a high frequency of sIPSCs.
Inter-event times (sIPSC interval times) were measured between all accepted sIPSCs, irrespective of their rise times. From all sIPSCs that were not dendritically filtered, we measured the peak current and the synaptic current decay time constant (decay). The decay was calculated from mono-exponential fits to the decay phase of each individual sIPSC. We established that the experimental decay data were best described by mono-exponential fits (Aikake's Information Criterion, see also Bozdogan (1987)). Possibly, the relatively small amplitudes of the sIPSCs, and therefore the less optimal signal to noise ratio of our recordings, made discrimination between fast desensitization and deactivation of GABAA receptors impossible in this experimental setup.
For each neuron, histograms were made of the peak currents, decay values and interval times of all the accepted sIPSCs. Neurons which had less than 50 accepted control sIPSCs were rejected from further analysis. Typically, we analysed approximately 400 sIPSCs per condition per neuron. Peak current and decay histograms were not normally distributed (Kolmogorov-Smirnov (K-S) test). Instead, they were skewed towards larger values. These distributions could be well fitted with lognormal curves. In contrast, the interval time histograms were best fitted with mono-exponential functions, as described previously (Brussaard et al. 1996).
Whether a specific neuron was sensitive for a certain ligand was tested by comparing the appropriate sIPSC parameter (interval time, peak current or decay) of all sIPSCs before and after ligand application with a Kolmogorov-Smirnov test. Ratios of sensitive neurons at pn6 and pn21 were tested for significant difference using Fisher's exact test. In order to do this, a neuron was termed 'sensitive' if indicated as such by the K-S test. The amplitudes of the pharmacological effects were compared between pn6 and pn21 using Student's t test. Throughout this paper, the level of statistical significance used was 5 %, unless otherwise stated.
|
RESULTS |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Postnatal changes in peak current and decay of IPSCs
The GABAA receptor subunit expression pattern is highly plastic during neonatal development of the visual cortex, as has been demonstrated by mRNA expression studies (Laurie et al. 1992) and protein measurements using antibodies (Fritschy et al. 1994). The changes in subunit expression, which take place between pn6 and pn21, correlate well in time with alterations in the peak currents and decay time constants (decay) of spontaneous IPSCs (sIPSCs) mediated by GABAA receptors (Laurie et al. 1992; Fritschy et al. 1994). However, we have only just begun to comprehend the causal relation between subunit transcription and postsynaptic activity.
As an overall starting point towards this goal, we investigated whether GABAA receptors with different kinetics are specifically targeted to synapses with low or high receptor numbers. We plotted histograms of peak currents versus decay values (Fig. 1) of all sIPSCs from all experiments carried out at different developmental stages. Immature neurons (at pn6) displayed a heterogeneous distribution of all combinations of peak current and decay: all combinations of peak currents and decay values occurred equally often (Fig. 1A). Then during the neonatal development, large slow-decaying sIPSCs disappeared first, whereas simultaneously large fast-decaying sIPSCs appeared (pn11, Fig. 1B). Between pn14 and pn21, the large fast-decaying sIPSCs disappeared gradually (Fig. 1C and D) and eventually also the small slow-decaying sIPSCs vanished (pn35, Fig. 1E), leaving the small fast-decaying sIPSCs. Thus, the large heterogeneity at pn6 changed into a rather clustered population of sIPSCs at pn35.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 1. Development of postsynaptic GABAergic currents Colour coded histograms of peak current vs. | ||
The average 10-90 % rise time of the sIPSCs decreased during development (1.58 ± 0.83 ms (pn6; n = 60 experiments; average ± S.D. of all sIPSCs pooled) vs. 1.16 ± 0.83 ms (pn21; n = 119 experiments; P < 0.05; Mann-Whitney test). This may indicate that the passive conductance properties of the neurons changed during development, also affecting the decay of sIPSCs. However, since the RC-time constants of individual recordings did not differ between the distinct developmental stages (see Methods) and since within each indivual experiment we selected sIPSCs that were not obviously under the influence of dendritic filtering, we conclude that the shift in decay of sIPSCs is largely due to developmental changes in GABAA receptor properties, including subunit switching. In order to analyse the specific contribution of subunit switching to the decay properties of GABAA receptors during development, we subsequently analysed the functional contribution of some of the subunits, both by means of pharmacological and genetic manipulations.
Pharmacological indications for subunit switching of GABAA receptors
For the subsequent pharmacological characterization of the synaptic GABAA receptors, we took pn6 as a starting point for neonatal development of the rat visual cortex. At that time, cortical layer formation has been completed and a period of massive synaptogenesis starts (Blue & Parnavelas, 1983). GABAergic synapses in pn21 rats were considered matured, since we did not find significant differences in either peak currents or decay values of sIPSCs between pn21 and pn35 (K-S test). Since rats open their eyes around pn13, we measured 1 week before and 1 week after eye opening. The pharmacological screening was used to study the functional, postsynaptic contribution of the different subunits present in the visual cortex. For 1, we used 100 nM zolpidem (Ruano et al. 1992; Renard et al. 1999; Vicini et al. 2001) and for 4, 100 µM furosemide (Korpi & Lüddens, 1997; Thompson et al. 1999). Ligands available for 3 are less specific. We used the best one available - 1 µM bretazenil (Puia et al. 1992). We realise that the subunit selectivity of bretazenil is far from optimal. Therefore, the interpretation of the results obtained may have to be considered with caution.
First, we tested whether the sIPSCs at pn6 and pn21 were affected by the application of 100 nM zolpidem (Fig 2A and C). We compared the decay of the sIPSCs before and after zolpidem application by means of a K-S test. Since 100 nM zolpidem preferentially elongates the decay of sIPSCs mediated by the 1 subunit containing GABAA receptors (Ruano et al. 1992; Renard et al. 1999), we could assess the 1 contribution at pn6 and at pn21. However, aspecific reactions of zolpidem with non-1 subunit-containing GABAA receptors at this concentration do occur and may lead to up to 25 % decay-elongation (Vicini et al. 2001). Therefore, we introduced a specificity threshold for the zolpidem effect, which was set at a relative effect of 30 %. Accordingly, we present three categories of neurons in Fig. 2D: neurons in the first category showed no significant elongation of decay upon zolpidem application (and thus probably have no or few synaptic 1 subunits). The second category comprises neurons that did show a significant zolpidem effect, but below the 30 % threshold. Hence in these neurons, the effects cannot with certainty be attributed to only 1 subunits. Finally, in the third category are the neurons that showed large (> 30 %), significant zolpidem effects, which we interpret as having a large portion of postsynaptic 1 receptors. When comparing pn6 and pn21, we observed not only an increase in the fraction of cells that were zolpidem sensitive (Fig. 2D), but also an increase in the relative effect of zolpidem per neuron (Fig. 2E). This implies that at pn21 there are not only more neurons expressing 1 at their synapses than at pn6, but also that the overall fraction of 1 subunits per synapse increases with neonatal development.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 2. The contribution of the zolpidem-sensitive, postsynaptic GABAA receptors increases during neonatal development A, example sIPSCs aligned to their rising phases. B, the same as A, but in the presence of 100 nM zolpidem. C, the effect of zolpidem increases with development. | ||
In a similar way, we analysed the sensitivity to the other two ligands. The second agent tested was bretazenil at 1 µM, which elongates the decay of all GABAA receptors but has been reported to have the largest effect on 3-containing GABAA receptors (Puia et al. 1992). As shown in Fig. 3A-C, bretazenil clearly affected the decay values of sIPSCs in immature tissue, but had hardly any effect in mature neurons. All neurons tested at pn6 showed clear potentiation by bretazenil, whereas only half the number of neurons at pn21 showed small effects (Fig. 3D). Furthermore, the effects in bretazenil-sensitive neurons decreased significantly with development (Fig. 3E). We conclude that the immature GABAA receptors are particularly affected by 1 µM bretazenil, in contrast to the mature receptor types at pn21. Since 3 is the only subunit that is abundant at pn6, but is downregulated at pn21 (Laurie et al. 1992), whereas all other subunits remain constant or are even upregulated, we hypothesize that the bretazenil action observed here can be largely attributed to specific binding of bretazenil to 3-containing GABAA receptors.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 3. The contribution of the bretazenil-sensitive, postsynaptic GABAA receptors decreases during neonatal development A, example sIPSCs aligned to their rising phases. B, the same as A, but in the presence of 1 µM bretazenil. C, the effect of bretazenil increases with development. Histograms of equal numbers of sIPSCs pooled from all experiments before ( | ||
Finally, we used the GABAA receptor antagonist furosemide at a relatively low concentration of 100 µM to test for the contribution of 4-containing GABAA receptors to the synaptic activity (Korpi & Lüddens, 1997; Thompson et al. 1999). Only one out of five immature neurons tested showed an effect of furosemide (both on sIPSC amplitude and sIPSC frequency). In contrast, most mature neurons displayed furosemide effects (Fig. 4A-G). Thus, the 4 subunit is almost absent in immature cells, but plays a significant role in a subset of mature neurons.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 4. The contribution of furosemide-sensitive, postsynaptic GABAA receptor increases during neonatal development A, example sIPSCs aligned to their rising phases. B, the same as A, but in the presence of 100 µM furosemide. C, the effect of furosemide increases with development. Peak current histograms of equal numbers of sIPSCs pooled from all experiments before ( | ||
From the pharmacological screening, we conclude that GABAA receptors at pn6 are especially sensitive to bretazenil, in contrast to the GABAA receptors at pn21, which are particularly affected by zolpidem.
Pharmacological classification of sIPSC kinetics
In order to test whether one can predict the average decay of a neuron on the basis of the GABAA receptor pharmacology of a particular neuron, we studied the putative correlation between the response to the different reagents used and the average decay in individual neurons under control conditions (i.e. before application). Figure 5Aa-c show scatter plots of the average decay and the average pharmacological effect of all neurons tested. To this end, we have superimposed the immature and the mature neurons and analysed putative correlations in the data sets thus obtained.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 5. The contribution of the different, pharmacologically classified GABAA receptor types to the overall sIPSC kinetics of neurons Aa-Ac, the average (± S.D.) | ||
The zolpidem effect was negatively correlated with the average control decay. This implies that neurons with, on average, faster sIPSCs show larger zolpidem effects than neurons with GABAA receptors that mediate, on average, slower sIPSCs (Fig. 5Aa). The opposite is true for bretazenil, indicating that bretazenil-sensitive GABAA receptors are preferentially located in neurons with, on average, slower sIPSCs (Fig. 5Ab). Finally, furosemide showed a (non-significant) trend towards a negative correlation, pointing at a possible, modest contribution of furosemide-sensitive GABAA receptors to the fast sIPSCs (Fig. 5Ac).
We next addressed the question of whether the correlations between GABAA receptor pharmacology and decay, as described above on the cellular level, could also be found at the level of individual synapses. Therefore, we compared the decay values of equal numbers of sIPSCs coming from all experiments before and after ligand application (Fig. 5Ba-Cc).
Histograms were made of the decay values of all these sIPSCs. The difference in distributions before and after application gives information about the typical decay kinetics of the sIPSCs mediated by GABAA receptors that were sensitive to the compound being tested. In the case of zolpidem, for instance, sIPSCs mediated by 1-containing GABAA receptors had an increased decay after application. This means that the fast sIPSCs that were present before, but not after, zolpidem application were affected by zolpidem and thus may be mediated by 1-containing GABAA receptors. The same reasoning holds true for bretazenil, which works in a similar way to zolpidem. For the antagonist furosemide, we tested for the sIPSCs that had disappeared as a consequence of furosemide application.
Analysis of 1 subunit deletion
The above pharmacological analysis indicates that shortening of sIPSC kinetics with development is caused by changes in subunit expression of GABAA receptors. To further substantiate the causal relation between subunit expression and sIPSC decay kinetics, we chose to corroborate our pharmacological analysis for the 1 subunit by means of genetic manipulation, where the subunit contribution was altered, while putative development of the passive cable properties of the dendrites could still take place. To this end, we made use of two independent genetic manipulation methods. We measured acute slices from adult GABAA receptor 1 knockout mice (Sur et al. 2001) and organotypical slice cultures from rats (Baker & Van Pelt, 1997) treated with antisense oligonucleotides against the 1 subunit. In both preparations we investigated the effects of 1 deletion on the decay kinetics of the sIPSCs.
As shown in Fig. 6A-D, sIPSCs from the 1 knockout mice had longer decay values than the sIPSCs from the wild-type mice. Apparently, the postsynaptic GABAA receptor pool is different after deletion of the 1 subunit (in line with the findings of Vicini et al. (2001)). Since these measurements were made in constitutive knockouts, compensation by, for instance, other subunits is quite possible. Therefore, we also used another genetic manipulation. In organotypical slice cultures, made from pn6 rats, we applied antisense oligonucleotides against the 1 subunit (Fig. 6E-H). We argue that the elegance of this method is that 1 deletion now only occurs during the week before measurement, leaving less opportunity for compensatory mechanisms. Also in these experiments we found an elongation of the average decay (Fig. 6G). With respect to possible compensatory mechanisms, it should be noted that the peak currents of the sIPSCs in the antisense-treated cultures were severely reduced (Fig. 6E), whereas those in the knockouts did not differ from the wild-types (Fig. 6A). This could point towards compensatory upregulation of other subunits in the knockouts, but not in the antisense-treated cultures.
 |
View larger version [in this window] [in a new window] |
|
|
Figure 6. Deletion of the GABAA receptor A, example sIPSCs aligned to their rising phases from adult wild-type mice (left) and adult | ||
The fact that deletion of the 1 subunit, either in knockout animals or in antisense-treated slice cultures, led to a lengthening of the decay of the sIPSCs showed that subunit switching can be held responsible for the developmental shift in decay of the sIPSCs. Thus the role of putative changes in passive electrical properties occurring with development can be largely excluded. The 10-90 % rise time of accepted sIPSCs in wild-type (WT) animals was 0.96 ± 0.42 ms, whereas in 1 -/- animals it was 1.09 ± 0.52 ms (obtained from ~7000 sIPSCs pooled from n = 17 for WT and n = 21 for 1 -/- experiments, respectively; P < 0.05; Student's t test), whereas their RC values were not significantly different (i.e. 131 ± 62 µs versus 124.4 ± 50 µs, respectively; P > 0.7; Student's t test). In 1 -/- mice the morphological development of visual cortex neurons is unalterated in Golgi staining (data not shown). Moreover, the fact that RC times of individual recordings were not altered indicates that the alterations in passive properties cannot be held responsible for the small but significant shift in rise time of sIPSCs coming from WT versus 1 -/- mice. Alternatively, we cannot exclude that some of the variation in rise times of sIPSCs may also depend on the subunit composition of the postsynaptic GABAA receptors being active (see also Haas & Macdonald (1999)). The finding that average rise time in 1 -/- mice was longer than in 1 +/+ mice corroborates this view.
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
We have studied the functional contribution of distinct pharmacological subtypes of GABAA receptor to the postsynaptic receptor activity in the rat visual cortex before and after neonatal development. We found that the fraction of postsynaptic GABAA receptors that was sensitive to bretazenil decreased between pn6 and pn21, whereas the effects of zolpidem and furosemide increased. Thus, in neonatal neurons there is probably a predominant functional synaptic contribution of the 3 subunit, although the contribution of 1 cannot be excluded. In matured neurons, the 1 subunit is probably the most prevalent, followed by 4 and to a lesser extent also 3 (Fig. 7). These conclusions based on the analysis the experiments using subunit-specific pharmacology are in line with the quantitative mRNA transcript measurements performed previously (Laurie et al. 1992).
 |
View larger version [in this window] [in a new window] |
|
|
Figure 7. Putative subunit composition of GABAA receptors before and after neonatal development of the visual cortex In immature synapses, the most abundant | ||
Furthermore, we demonstrate that the presence of synaptic GABAA receptor subunits is very important for determining the decay kinetics of GABAergic sIPSCs. Both our pharmacological experiments and the genetic manipulations show that the 1 subunit is responsible for fast decay. The upregulation of 4-containing GABAA receptors is characterized by even faster sIPSCs, whereas 3-containing receptors in both juvenile and mature neurons appear to mediate slow decaying sIPSCs.
The 2, 5 and 6 subunits were not investigated. There is no good pharamacological tool available for the 2 subunit. However, in earlier experiments performed in our lab, it has been shown that 2 antisense deletion in visual cortex tissue cultures accelerates the decay kinetics, indicating that 2 contributes to GABAA receptors involved in slow decaying IPSCs (Brussaard et al. 1997). In acutely dissected tissue, the 5 subunit is present at low levels at pn6 (13 % of all subunit expression), but virtually absent from mature neurons (data not shown). In cultured cortical neurons, 5 plays a larger role than in acutely sliced tissue (Dunning et al. 1999; Hutcheon et al. 2000). Unfortunately, we had no good ligand to test for 5 contribution. Since its modest role, however, this does not seem to be a major problem. The 6 subunit is not expressed in the visual cortex (Wisden et al. 1992).
Specificity of the pharmacology used
The specificity of the GABAA receptor subunit-selective ligands currently available is under dispute. In particular, the specificity of bretazenil for 3 has been doubted (Ebert et al. 1997; Smith et al. 2001). While bretazenil has been reported to elicit the greatest enhancement at 3-containing receptors (Puia et al. 1992), this study did not take into account the lower GABA affinity of the 32 receptor subtype relative to the other GABAA receptor subtypes used in these experiments. Consequently it would not be surprising to observe greater enhancement at the same concentration of GABA. Other studies on heterologous expression systems, for instance Smith et al. (2001), demonstrated a lack of subtype selectivity for bretazenil. However, here we show that bretazenil affected sIPSCs at pn6, but not at pn21. Since 3 is the only subunit that is highly expressed at pn6 but hardly at pn21 (Laurie et al. 1992), we propose that, at least in the postsynaptic receptors of the visual cortex, bretazenil is exhibiting significant selectivity for 3-containing GABAA receptors.
Zolpidem does not potentiate subunit-containing GABAA receptors (Hevers & Lüddens, 1998). Mature, but not immature, neurons of the visual cortex have relatively high levels of subunit expression (Laurie et al. 1992; Pirker et al. 2000). This could make some GABAA receptors zolpidem insensitive, even when they have an 1 subunit. The described developmental upregulation of 1 may therefore be underestimated, although in the cerebellum, the subunit is exclusively located outside the synapse (Nusser et al. 1998).
GABAA receptor plasticity during neonatal development
Earlier studies have shown that the expression patterns of GABAA receptor subunits are spatially (Wisden et al. 1992; Pirker et al. 2000) and developmentally (Laurie et al. 1992; Fritschy et al. 1994) restricted. Most of these studies used mRNA experiments. In the present study we validate, at least for layers II-III of the visual cortex, these expression studies by showing that the described changes in transcription are reflected in plasticity of the functional, postsynaptic receptor proteins. Our results are further in agreement with those of Dunning et al. (1999) and Hutcheon et al. (2000), who showed that 1 and 4 are upregulated during in vitro development of dissociated embryonic cortical cells, whereas 3 is downregulated.
We have used randomly selected neurons in order to get an overall impression of the subunit contribution of all neurons in layers II-III of the rat visual cortex, instead of only in a subset of neurons. Approximately 66 % of the neurons in layers II-IV are pyramidal cells (Winfield et al. 1980). The sIPSCs recorded in these neurons reflect synaptic input from all types of interneurons. In addition, interneuron-interneuron connections have been included in this study.
Kinetic contributions of the GABAA receptor subunits
Our data support the general concept that, amongst other factors, subunit expression is causally related to decay kinetics of sIPSCs (Hevers & Lüddens, 1998). This is in line with data obtained either in recombinant expression studies (in Xenopus oocytes or HEK cells) or by single-cell PCR of one or few subunits (Dunning et al. 1999; Okada et al. 2000). It has been demonstrated that recombinant GABAA receptors containing the 2 (Lavoie et al. 1997) or 3 (Verdoorn, 1994) subunits decay more slowly than 1-containing GABAA receptors (see also Hevers & Lüddens, 1998). This has been confirmed by studies on an 1 knockout mouse line, generated independently from the one we used, for the cerebellum (Vicini et al. 2001) and by 2 antisense deletion in organotypical slice cultures of the visual cortex (Brussaard et al. 1997). Furthermore, it is in line with kinetic changes in neurons of the hypothalamus that display an endogenous 1 to 2 switch (Brussaard et al. 1997). The advantage of our present study on neonatal development is that we were able to study the kinetic properties of three abundant subunits in their native settings, without having to consider the compensatory regulation of in vitro culture conditions or gene knockout mice. We have shown that the different subunits are indeed associated with different kinetics, but that there is quite a large, heterogeneous range of action, especially for 3.
Finally, we have shown that deletion of the 1 subunit, either by generating a knockout mouse or by applying antisense oligonucleotides in an organotypical rat slice culture, yields very similar results when compared to our pharmacological analysis using 100 nM zolpidem in normal rat tissue. The differences between the decay properties of 1 knockout mice and antisense-treated cultures can be explained in two ways. At first, there are also differences in the controls (Fig. 6D and H). These can be caused by differences in expression of other subunits, but also by other factors, like phosphorylation (Jones & Westbrook, 1997). Secondly, long term compensatory mechanisms in the 1 -/- mice may occur. In the antisense-treated cultures, a marked reduction of the peak currents was observed (Fig. 6E), whereas there were no differences in peak currents between 1 +/+ and 1 -/- mice (Fig. 6A). This may indicate that in the knockout mice, but not (or less) in the antisense-treated cultures, compensatory upregulation of other subunits occurred.
Coexistence of different subunits in specific neurons
All immature neurons, at pn6, show clear bretazenil-sensitivity. This implies that all these neurons probably express substantial amounts of postsynaptic 3 subunits. Therefore, the other subunits present at that age must co-occur in 3-expressing neurons. The same holds true for zolpidem sensitivity in pn21 neurons. Since zolpidem affected virtually all pn21 neurons, we postulate that 1 may occur together with 2 and 4 in these neurons. This implies, in our view, that different subunits are co-expressed in individual neurons, although one cannot exclude that a minor fraction of the neurons express only one subunit isomer.
Differences in kinetic properties of sIPSCs mediated by the same subunit between different synapses can be explained in four, not mutually exclusive, ways. The first explanation is that there is more than one type of subunit per synapse, either in the same GABAA receptor (Verdoorn, 1994; Araujo et al. 1996) or in different receptors. This could imply that synapses with only 3 mediate slow-decaying sIPSCs, whereas synapses combining 3 and 1 mediate relatively fast-decaying synaptic currents.
Also the other, and /, subunits can influence the kinetic properties of GABAA receptors. It has been described that subunit-containing GABAA receptors have faster kinetics than 2-containing ones (Haas & Macdonald, 1999). Developmental upregulation of mRNA in the visual cortex (Laurie et al. 1992) may therefore also contribute to faster sIPSCs after neonatal development, although the possibility of synaptic localization of the subunit has been rejected in other brain areas (Nusser et al. 1998). Furthermore, post-translational modification, for instance by (de)phosphorylation, is known to contribute to alterations in the kinetics in some cell types (see for instance Jones & Westbrook, 1997; McDonald et al. 1998). However, since the subunits display much larger changes during development than the other subunits, we argue that the changes in non- subunits have less impact than changes in the expression of subunits during development. Finally, also variations in the kinetics of GABA in the synaptic cleft may affect GABAA receptor kinetics (Draguhn & Heinemann, 1996; Nusser et al. 2001).
Conclusions
Developmental changes in subunit expression occur in order to mediate faster decaying GABAergic IPSCs. Both at juvenile and mature stages, a predominant contribution of particular subunits can be observed, with the 3 subunit being the most abundant juvenile subunit and the 1 the dominant mature subunit. The presence of other subunits, mainly 1 at immature synapses and 4 at mature synapses, was confirmed. In addition, co-existence of more than one type of subunit within single neurons is evident. Each individual subunit generates its own kinetic receptor profile. However, combination with other subunits may lead to intermediate receptor kinetics, thereby fine tuning the effect of transcriptional subunit switching during neonatal development.
|
REFERENCES |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
| ARAUJO, F., TAN, S., RUANO, D., SCHOEMAKER, H., BENAVIDES, J. & VITORICA, J. (1996). Molecular and pharmacological characterization of native cortical -aminobutyric acidA receptors containing both 1 and 3 subunits. Journal of Biological Chemistry 271, 27902-27911 |
[Abstract/Full Text] |
| BAKER, R. E. & VAN PELT, J. (1997). Cocultured, but not isolated, cortical explants display normal dendritic development: a long-term quantitative study. Brain Research Developmental Brain Research 98, 21-29 | [Medline] |
| BANERJEE, P. K., OLSEN, R. W. & SNEAD, O. C. III (1999). Zinc inhibition of -aminobutyric acidA receptor function is decreased in the cerebral cortex during pilocarpine-induced status epilepticus. Journal of Pharmacology and Experimental Therapeutics 291, 361-366 |
[Abstract/Full Text] |
| BAUMANN, S. W., BAUR, R. & SIGEL, E. (2001). Subunit arrangement of -aminobutyric acid type A receptors. Journal of Biological Chemistry 276, 36275-36280 |
[Abstract/Full Text] |
| BLUE, M. E. & PARNAVELAS, J. G. (1983). The formation and maturation of synapses in the visual cortex of the rat. II. Quantitative analysis. Journal of Neurocytology 12, 697-712 | [Medline] |
| BORST, J. G., LODDER, J. C. & KITS, K. S. (1994). Large amplitude variability of GABAergic IPSCs in melanotropes from Xenopus laevis: evidence that quantal size differs between synapses. Journal of Neurophysiology 71, 639-655 | [Abstract] |
| BOZDOGAN, H. (1987). Model selection and Akaike's Information Criterion (AIC): the general theory and its analytical extensions. Psychometrika 52, 345-370 | |
| BROOKS-KAYAL, A. R., SHUMATE, M. D., JIN, H., RIKHTER, T. Y. & COULTER, D. A. (1998). Selective changes in single cell GABAA receptor subunit expression and function in temporal lobe epilepsy. Nature Medicine 4, 1166-1172 | [Medline] |
| BRUSSAARD, A. B., KITS, K. S., BAKER, R. E., WILLEMS, W. P., LEYTING-VERMEULEN, J. W., VOORN, P., SMIT, A. B., BICKNELL, R. J. & HERBISON, A. E. (1997). Plasticity in fast synaptic inhibition of adult oxytocin neurons caused by switch in GABAA receptor subunit expression. Neuron 19, 1103-1114 | [Medline] |
| BRUSSAARD, A. B., KITS, K. S. & DE VLIEGER, T. A. (1996). Postsynaptic mechanism of depression of GABAergic synapses by oxytocin in the supraoptic nucleus of immature rat. Journal of Physiology 497, 495-507 | [Abstract] |
| DRAGUHN, A. & HEINEMANN, U. (1996). Different mechanisms regulate IPSC kinetics in early postnatal and juvenile hippocampal granule cells. Journal of Neurophysiology 76, 3983-3993 | [Abstract] |
| DUNNING, D. D., HOOVER, C. L., SOLTESZ, I., SMITH, M. A. & O'DOWD, D. K. (1999). GABAA receptor-mediated miniature postsynaptic currents and alpha-subunit expression in developing cortical neurons. Journal of Neurophysiology 82, 3286-3297 | [Abstract/Full Text] |
| EBERT, B., THOMPSON, S. A., SAOUNATSOU, K., MCKERNAN, R., KROGSGAARD-LARSEN, P. & WAFFORD, K. A. (1997). Differences in agonist/antagonist binding affinity and receptor transduction using recombinant human -aminobutyric acid type A receptors. Molecular Pharmacology 52, 1150-1156 |
[Abstract/Full Text] |
| FARRAR, S. J., WHITING, P. J., BONNERT, T. P. & MCKERNAN, R. M. (1999). Stoichiometry of a ligand-gated ion channel determined by fluorescence energy transfer. Journal of Biological Chemistry 274, 10100-10104 | [Abstract/Full Text] |
| FRITSCHY, J. M., PAYSAN, J., ENNA, A. & MÖHLER, H. (1994). Switch in the expression of rat GABAA-receptor subtypes during postnatal development: an immunohistochemical study. Journal of Neuroscience 14, 5302-5324 | [Abstract] |
| GUILINELLO, M., GONG, Q. H., LI, X. & SMITH, S. S. (2001). Short-term exposure to a neuroactive steroid increases 4 GABAA receptor subunit levels in association with increased anxiety in the female rat. Brain Research 910, 55-66 |
[Medline] |
| HAAS, K. F. & MACDONALD, R. L. (1999). GABAA receptor subunit 2 and subtypes confer unique kinetic properties on recombinant GABAA receptor currents in mouse fibroblasts. Journal of Physiology 514, 27-45 |
[Abstract/Full Text] |
| HEVERS, W. & LÜDDENS, H. (1998). The diversity of GABAA receptors. Pharmacological and electrophysiological properties of GABAA channel subtypes. Molecular Neurobiology 18, 35-86 | [Medline] |
| HUTCHEON, B., MORLEY, P. & POULTER, M. O. (2000). Developmental change in GABAA receptor desensitization kinetics and its role in synapse function in rat cortical neurons. Journal of Physiology 522, 3-17 | [Abstract/Full Text] |
| JONES, M. V. & WESTBROOK, G. L. (1997). Shaping of IPSCs by endogenous calcineurin activity. Journal of Neuroscience 17, 7626-7633 | [Abstract/Full Text] |
| KLAUSBERGER, T., SARTO, I., EHYA, N., FUCHS, K., FURTMULLER, R., MAYER, B., HUCK, S. & SIEGHART, W. (2001). Alternate use of distinct intersubunit contacts controls GABAA receptor assembly and stoichiometry. Journal of Neuroscience 21, 9124-9133 | [Abstract/Full Text] |
| KORPI, E. R. & LÜDDENS, H. (1997). Furosemide interactions with brain GABAA receptors. British Journal of Pharmacology 120, 741-748 | [Abstract] |
| LAURIE, D. J., WISDEN, W. & SEEBURG, P. H. (1992). The distribution of thirteen GABAA receptor subunit mRNAs in the rat brain. III. Embryonic and postnatal development. Journal of Neuroscience 12, 4151-4172 | [Abstract] |
| LAVOIE, A. M., TINGEY, J. J., HARRISON, N. L., PRITCHETT, D. B. & TWYMAN, R. E. (1997). Activation and deactivation rates of recombinant GABAA receptor channels are dependent on -subunit isoform. Biophysical Journal 73, 2518-2526 |
[Abstract] |
| MCDONALD, B. J., AMATO, A., CONNOLLY, C. N., BENKE, D., MOSS, S. J. & SMART, T. G. (1998). Adjacent phosphorylation sites on GABAA receptor subunits determine regulation by cAMP-dependent protein kinase. Nature Neuroscience 1, 23-28 |
[Medline] |
| MCKERNAN, R. M., ROSAHL, T. W., REYNOLDS, D. S., SUR, C., WAFFORD, K. A., ATACK, J. R., FARRAR, S., MYERS, J., COOK, G., FERRIS, P., GARRETT, L., BRISTOW, L., MARSHALL, G., MACAULAY, A., BROWN, N., HOWELL, O., MOORE, K. W., CARLING, R. W., STREET, L. J., CASTRO, J. L., RAGAN, C. I., DAWSON, G. R. & WHITING, P. J. (2000). Sedative but not anxiolytic properties of benzodiazepines are mediated by the GABAA receptor 1 subtype. Nature Neuroscience 3, 587-592 |
[Medline] |
| NUSSER, Z., NAYLOR, D. & MODY, I. (2001). Synapse-specific contribution of the variation of transmitter concentration to the decay of inhibitory postsynaptic currents. Biophysical Journal 80, 1251-1261 | [Abstract/Full Text] |
| NUSSER, Z., SIEGHART, W. & SOMOGYI, P. (1998). Segregation of different GABAA receptors to synaptic and extrasynaptic membranes of cerebellar granule cells. Journal of Neuroscience 18, 1693-1703 | [Abstract/Full Text] |
| OKADA, M., ONODERA, K., VAN RENTERGHEM, C., SIEGHART, W. & TAKAHASHI, T. (2000). Functional correlation of GABAA receptor subunits expression with the properties of IPSCs in the developing thalamus. Journal of Neuroscience 20, 2202-2208 |
[Abstract/Full Text] |
| PIRKER, S., SCHWARZER, C., WIESELTHALER, A., SIEGHART, W. & SPERK, G. (2000). GABAA receptors: immunocytochemical distribution of 13 subunits in the adult rat brain. Neuroscience 101, 815-850 | [Medline] |
| PUIA, G., DUCIC, I., VICINI, S. & COSTA, E. (1992). Molecular mechanisms of the partial allosteric modulatory effects of bretazenil at -aminobutyric acid type A receptor. Proceedings of the National Acadamy of Sciences of the USA 89, 3620-3624 |
|
| RENARD, S., OLIVIER, A., GRANGER, P., AVENET, P., GRAHAM, D., SEVRIN, M., GEORGE, P. & BESNARD, F. (1999). Structural elements of the -aminobutyric acid type A receptor conferring subtype selectivity for benzodiazepine site ligands. Journal of Biological Chemistry 274, 13370-13374 |
[Abstract/Full Text] |
| RUANO, D., VIZUETE, M., CANO, J., MACHADO, A. & VITORICA, J. (1992). Heterogeneity in the allosteric interaction between the -aminobutyric acid (GABA) binding site and three different benzodiazepine binding sites of the GABAA/benzodiazepine receptor complex in the rat nervous system. Journal of Neurochemistry 58, 485-493 |
[Abstract] |
| RUDOLPH, U., CRESTANI, F., BENKE, D., BRÜNIG, I., BENSON, J. A., FRITSCHY, J. M., MARTIN, J. R., BLUETHMANN, H. & MÖHLER, H. (1999). Benzodiazepine actions mediated by specific -aminobutyric acid(A) receptor subtypes. Nature 401, 796-800 |
[Medline] |
| SMITH, A. J., ALDER, L., SILK, J., ADKINS, C., FLETCHER, A. E., SCALES, T., KERBY, J., MARSHALL, G., WAFFORD, K. A., MCKERNAN, R. M. & ATACK, J. R. (2001). Effect of subunit on allosteric modulation of ion channel function in stably expressed human recombinant gamma-aminobutyric acid(A) receptors determined using (36)Cl ion flux. Molecular Pharmacology 59, 1108-1118 |
[Abstract/Full Text] |
| SUR, C., WAFFORD, K. A., REYNOLDS, D. S., HADINGHAM, K. L., BROMIDGE, F., MACAULAY, A., COLLINSON, N., O'MEARA, G., HOWELL, O., NEWMAN, R., MYERS, J., ATACK, J. R., DAWSON, G. R., MCKERNAN, R. M., WHITING, P. J. & ROSAHL, T. W. (2001). Loss of the major GABAA receptor subtype in the brain is not lethal in mice. Journal of Neuroscience 21, 3409-3418 | [Abstract/Full Text] |
| THOMPSON, S. A., ARDEN, S. A., MARSHALL, G., WINGROVE, P. B., WHITING, P. J. & WAFFORD, K. A. (1999). Residues in transmembrane domains I and II determine -aminobutyric acid type AA receptor subtype-selective antagonism by furosemide. Molecular Pharmacology 55, 993-999 |
[Abstract/Full Text] |
| TIA, S., WANG, J. F., KOTCHABHAKDI, N. & VICINI, S. (1996). Developmental changes of inhibitory synaptic currents in cerebellar granule neurons: role of GABAA receptor 6 subunit. Journal of Neuroscience 16, 3630-3640 |
[Abstract/Full Text] |
| VERDOORN, T. A. (1994). Formation of heteromeric -aminobutyric acid type A receptors containing two different subunits. Molecular Pharmacology 45, 475-480 |
[Abstract] |
| VICINI, S., FERGUSON, C., PRYBYLOWSKI, K., KRALIC, J., MORROW, A. L. & HOMANICS, G. E. (2001). GABAA receptor 1 subunit deletion prevents developmental changes of inhibitory synaptic currents in cerebellar neurons. Journal of Neuroscience 21, 3009-3016 |
[Abstract/Full Text] |
| WINFIELD, D. A., GATTER, K. C. & POWELL, T. P. (1980). An electron microscopic study of the types and proportions of neurons in the cortex of the motor and visual areas of the cat and rat. Brain 103, 245-258 | [Medline] |
| WISDEN, W., LAURIE, D. J., MONYER, H. & SEEBURG, P. H. (1992). The distribution of 13 GABAA receptor subunit mRNAs in the rat brain. I. Telencephalon, diencephalon, mesencephalon. Journal of Neuroscience 12, 1040-1062 |
Acknowledgements
This study was supported by ALW-NWO (no. 805.33.370) from the Netherlands to A.B.B. The authors like to thank Ms K. Heinen and Drs J. J. M. Bedaux, N. Burnashev and K. A. Wafford for helpful discussion and comments on previous versions of this manuscript. The authors are grateful to Mr J. C. Lodder and Mrs T. E. Busé-Pot for their excellent technical support.
This article has been cited by other articles:
|
|
 |
|
  A. B. Ali and A. M. Thomson Synaptic {alpha}5 Subunit-Containing GABAA Receptors Mediate IPSPs Elicited by Dendrite-Preferring Cells in Rat Neocortex Cereb Cortex, June 1, 2008; 18(6): 1260 - 1271. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Epsztein, Y. Ben-Ari, A. Represa, and V. Crepel Late-Onset Epileptogenesis and Seizure Genesis: Lessons From Models of Cerebral Ischemia Neuroscientist, February 1, 2008; 14(1): 78 - 90. [Abstract] [PDF]
|
|
|
|
 |
|
 L. Medrihan, E. Tantalaki, G. Aramuni, V. Sargsyan, I. Dudanova, M. Missler, and W. Zhang Early Defects of GABAergic Synapses in the Brain Stem of a MeCP2 Mouse Model of Rett Syndrome J Neurophysiol, January 1, 2008; 99(1): 112 - 121. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. A. Mitchell, L. J. Gentet, J. Dempster, and D. Belelli GABAA and glycine receptor-mediated transmission in rat lamina II neurones: relevance to the analgesic actions of neuroactive steroids J. Physiol., September 15, 2007; 583(3): 1021 - 1040. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. A. Prenosil, E. M. Schneider Gasser, U. Rudolph, R. Keist, J.-M. Fritschy, and K. E. Vogt Specific Subtypes of GABAA Receptors Mediate Phasic and Tonic Forms of Inhibition in Hippocampal Pyramidal Neurons J Neurophysiol, August 1, 2006; 96(2): 846 - 857. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Ponomarev, R. Maiya, M. T. Harnett, G. L. Schafer, A. E. Ryabinin, Y. A. Blednov, H. Morikawa, S. L. Boehm II, G. E. Homanics, A. Berman, et al. Transcriptional Signatures of Cellular Plasticity in Mice Lacking the {alpha}1 Subunit of GABAA Receptors J. Neurosci., May 24, 2006; 26(21): 5673 - 5683. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. W. J. Bosman, K. Heinen, S. Spijker, and A. B. Brussaard Mice Lacking the Major Adult GABAA Receptor Subtype Have Normal Number of Synapses, But Retain Juvenile IPSC Kinetics Until Adulthood J Neurophysiol, July 1, 2005; 94(1): 338 - 346. [Abstract] [Full Text] [PDF]
|
|
) and after (
) zolpidem application. Insets: average sIPSCs, control (thick line) and in the presence of zolpidem (thin line), normalized to peak current. D, the fraction of zolpidem-sensitive neurons increases between pn6 and pn21. Some neurons showed a significant but relatively small increase in