Archives of Insect Biochemistry and Physiology 18:147-I 57 (1 991)
Functional Assay for GABA Receptor Subtypes of a Cockroach Giant lnterneuron Bernard Hue Laboratoire de Newrophysiologie, Cedex, France
URA CNRS 622 UnzversitC d’Angers, 49045 Angers
-
y Aminobutyric acid (GABA) receptors were examined in the cockroach central nervous system (CNS) using the single fiber-oil gap method applied to an identified giant interneuron. Short-lasting pressure application of 10 rnM GABA developed a multiphasic response composed of a fast hyperpolarization followed by a transient depolarizing component and a stable hyperpolarization. This triphasic characteristic shape of the response was modified according to the dose of GABA injected or bath-applied and to the precise localization of the injection within the dendritic area. The transient depolarizing phase showed a negative reversal potential of -70 mV. Both hyperpolarizing phases reversed at a more negative level ranging to - 80 mV. A positive shift of these values was caused by a decrease in external chloride concentration. Bath-application of 0.1 m M picrotoxin (Ptx) decreased the depolarizing phase which was progressively replaced by a stable hyperpolarization. The transient depolarizing component desensitized quickly and was the most sensitive phase to Ptx action. The Ptx-resistant response reversed at a mean value of -100 rnV close to the equilibrium potential for potassium ions (EKC), suggesting that it was generated by a K+-channel coupled receptor. Although baclofen was unable to mimic the Ptx-resistant GABA response, the compound CGA 147823, known to bind with a high specificity to vertebrate GABAHreceptors, has been successfully used to reproduce the Ptx-resistantGABA response. It i s suggested that, in addition to GABA receptors linked to chloride channels, the insect CNS possesses GABA receptors sharing ionic characteristics of GABAHreceptors especially those located i n the vertebrate CNS, although they are insensitive to baclofen. Key words: insect CNS, synaptic receptor, fiber-oil gap method
Acknowledgments: 8. Hue thanks CIBA-GEIGY for the generous supply of CGA 147823 and
T. Piek for his helpful comments. Received March 21,1991; accepted June10,1991. Address reprint requests to Professor Bernard Hue, Laboratoire de Neurophysiologie, URA CNRS 611, FacultC de Medecine, rue haute de reculee, 49045 Angers Cedex, France. 0 1991 Wiley-Liss, Inc.
148
Hue
INTRODUCTION
The inhibitory neurotransmitter, GABA", is widely distributed in the insect CNS. The existence of only one type of CABA receptor resembling the vertebrate GABAA receptor has been recently reviewed [l]; however, some disparity in GABA receptor distribution or ionic mechanisms has been previously reported [2-41. The insect's GABA receptor is linked to a Ptx-sensitivechlorideion channel and possesses modulatory sites for benzodiazepines and barbiturates 151, but its insensitivity to bicuculline 16-81 supports the views that its pharmacological properties differ from either of the vertebrate GABA receptor subtypes. Electrophysiological and pharmacological studies on GABA receptors have been mainly performed on neuronal cell bodies studied in situ [4,9] or in culture [5,10]. Moreover, GABA has been identified as an inhibitory neurotransmitter at the cercal-nerve GI synapse in the A6 ganglion of the cockroach, Peripluneta arnericuna [11,12].GABA receptors are implicated in presynaptic and postsynaptic inhibition [11,13]. The present study extends previous investigations [8, 141 carrried on cercalnerve GI synapses, This preparation allows examination of responses induced by pressure application of drugs [15]. So, we have recorded membrane potential changes due to pressure application of GABA within the neuropile of the cockroach's A6 ganglion. The induced multiphasic GABA response overlaps a Ptx-resistant hyperpolarization which has been reproduced with 3-aminopropyl-phosphonous acid (CGA 147823), a new potent GABABagonist [16]. MATERIALS AND METHODS
Single fiber-oil gap experiments were performed using the cercal-nerve GI synapses located within the neuropile of the A6 ganglion of the CNS of the cockroach P. urnericuna [11]. All experimental data were obtained on the GI no. 2 visually identified during microdissection. Anatomical details of the dendritic tree have been revealed using the cobalt blackfilling method described previously on the same preparation [17]. Postsynaptic resting membrane potential was continuously monitored on a pen-chart recorder. Changes in postsynaptic membrane resistance were measured by modifications to hyperpolarizing square current pulses applied through the membrane by means of a balanced Wheatstone bridge connected in the recording circuit. The A6 ganglion was desheathed and superfused at a constant rate with a saline containing (in mM): NaCl208; KCl3.1; CaClz 10.0; NaHC03 2.0; sucrose 26; pH 7.2. In a series of experiments, the chloride content of the normal saline was reduced by replacement of the sodium chloride with equimolar sodium acetate or propionate. GABA and related agonists were injected within the A6 neuropile onto the GI dendrites using broken micropipettes (tip diameter 20 pm) connected to a pneumatic pressure ejecting system (Neuro Phore BH-2, Medical System Corp., Greenvale, NY, USA) as described earlier [15]. Direct microapplication of GABA to the dendrites was preferentially used on account of *Abbreviations used: A6 = Sixth abdominal; CNS = central nervous system; EPSP = excitatory postsynaptic potential; CABA = y-aminobutyric acid; GI = giant interneuron; IPSP = inhibitory postsynaptic potential; Ptx = picrotoxin; l 7 X = tetrodotoxin.
Cockroach CABA Receptor Subtypes
149
the high peripheral glial uptake of GABA [18]. Furthermore, pressure injections allow precise local applications of drugs and make easier the reversibility of the induced effects. In control experiments, pressure microapplications of saline never produced any effects on the synaptic transmission. All experiments were performed at room temperature (20°C). The antagonists (Ptx, baclofen) were dissolved in the saline and externally applied when necessary. Pharmacological compounds were obtained from Sigma Chemical Co. (St Louis, MO, USA). CGA 147823 was kindly provided by CIBA-GEIGY Ltd., (Basel, Switzerland) for use in this study. RESULTS Physiological Effects of Pressure AppIication of GABA Short-lasting microapplication (300 ms, 15 psi) of GABA at concentrations ranging from 0.1-1 mM within the A6 ganglion neuropile (close to the main dendritic branches of the GI) induced a triphasic response in the GI postsynaptic membrane. A fast hyperpolarization preceded a slow depolarizing wave followed by a slow hyperpolarizing phase (Fig. 1A). Concurrently, amplitude of unitary EPSPs was drastically decreased [2]. In this paper no attempt was made to further analyze this latter effect or changes in postsynaptic membrane A
C
B
control
c
I
A
A
5 Psi
10 P s i
15 Psi
Dv4fPrlc1 A
A
A
Fig. 1. Effects of CABA on the resting potential of the GI2 postsynaptic membrane. A, C, D: 1 mM CABA solution was applied by pressure injection (indicated by black triangles) within the neuropile of the A6 ganglion. B: 20 mM CABA was externally applied (black bar) onto the A6 ganglion. Temperature of the bath was lowered to 10°C in order to prevent excessive GABA uptake. Note in C that TTX does not suppress the action of GABA and in the dose-dependent shape of the response when the injection pressure is increased. In order to prevent desensitization of GABA receptors, the microapplications of GABA in D were separated by a period of time of 20 min. Vertical scale: A, B, C, I mV; 0,2 mV; horizontal scale: AC, 15 s; B , 2 min; D, 30 s.
150
Hue
conductance which have been previously studied [2,12]. Similar responses were also observed when a high concentration (10 mM) of GABA was externally applied (Fig. 1B) at a relatively low temperature (10OC) in order to reduce glial uptake of GABA [MI.Obviously in the latter case, the time course of the different waves was slower. These similarities show that the mode of application of GABA does not interfere with the main shape of the response. In the presence of 1 pM TTX, the triphasic GABA response was not altered as shown in Figure 1C, suggesting that GABA acted directly on synaptic receptors of the GI no. 2 under test. Further investigations were carried out in order to more precisely establish the dose dependence of the response. Figure 1D shows that low doses of GABA injected during 300 ms at 5 psi produced only a small hyperpolarization. Higher doses (10 and 15 psi) activated a transient depolarizing component which was more pronounced at higher pressure applications. Increased doses of GABA, applied through the superfusing saline, produced comparable effects (not illustrated). Dendritic CABA Responses
The multiphasic GABA response, as previously described, was only recorded when the tip of the injection micropipette was carefully positioned close to the center of the dendritic area (Fig. 2), i.e., close to the initial part of the main dendritic branches; whereas, at the periphery of the dendritic tree the GABA response becomes mainly depolarizing. No effect was seen when GABA
P
a
-
200pm
15 s
Fig. 2. Dependence of GABA responses on the application site. Left: A camera lucida drawing of the C12, filled with cobalt chloride, within the neuropile of the A6 ganglion i s represented. A trachea lying on the ventral side of the ganglion was used as a landmark for the precise microinjection of GABA. Both cercal nerves (p: posterior part of the ganglion) and connectives were cut. Right: Four traces correspond to the microinjection of GABA indicated by the black triangle (pulse duration, 300 rns; pressure, 15 psi; depth, 200 bm) at four different sites localized by the place of a micropipette. Note the lack of GABA response when the injection is done within the anterior part (a) of the neuropile, close to the initial part of the giant axon.
Cockroach CABA Receptor Subtypes
151
A
L 15 5 -80nV
-4
1
Fig. 3. Effect of shift of the resting potential on the shape of the triphasic response induced by a 300 ms CABA pulse (black triangle). The control resting potential ( - 75 mv) was evaluated after full depolarization of the postsynaptic membrane by superfusion with saline containing 210 m M KCI. A: The amplitudes were measured at the peak of each phase (indicated by A,0, 0). 6: The amplitudeswere plotted on the graph vs. the shift of the resting potential.
was applied within the anterior part of the A6 ganglion containing the initial segment of the giant axon. This latter observation indicates that GABA receptors are restricted to the GI dendritic tree. Earlier studies [11,12,19] performed on the same GI system have demonstrated that IPSPs were due to the release of GABA from cercal-nerve X endings. Changing the chloride concentration in the saline could reverse the hyperpolarization induced by bath-applied GABA as well as evoked IPSP. The reversal potential of these two responses was similar, around - 80 mV, suggesting that they were supported by the same chloride current. Further experiments were performed in an attempt to precisely determine the ionic mechanisms underlying the three components of the GAEL4 response. By passing hyperpolarizing current through the membrane the reversal potential of the GABA waves has been estimated. As indicated on the graph (Fig. 3) the depolarizing phase had the least negative reversal potential ( - 70 mV), both hyperpolarizing waves reversed at a more negative level ( - 79 mV and - 80 mV). These reversal potential values were consistent between GI no. 2 tested (n = 5). Lowering the amount of chloride in the saline (222-166 mM) caused both GABA hyperpolarizing phases to fade whereas the depolarizing phase was enhanced (Fig. 4A). Finally, Ptx, the well-known GABA-dependent chloride ionophore blocker [20,21], was bath-applied at 0.1 mM which fully blocked the chloride-dependent IPSP [11,19]. The time-course of the Ptx action on the triphasic GABA response was not quantified and is not presented here, but a slow decrease of the depolarizing phase occurred and was progressively replaced by a unique hyperpolarizing response which was resistant to Ptx action (Fig. 4B). During long-lasting (30 s) microapplication of GABA, postsynaptic mem-
152
Hue
A
6 control
n
mV Fig. 4. Effects of low chloride saline, 110 mM (A) and 0.1 mM Ptx (6) on the triphasic response induced by a 300 ms CABA pulse (black squares). Note that Ptx suppresses the depolarizing phase.
brane conductance was measured as changes in short (50ms) membrane hyperpolarizations. It was observed (Fig. 5) that conductance changes paralleled changes in resting membrane potential, i. e., the maximal conductance increase coincided with the peak of the transient depolarization and a steady-state value was reached. Increasing doses of Ptx (up to 10 pM) primarily affected the maximal conductance concomitant with the peak of the transient depolarizingphase without apparent effect on the steady-state. The steady-state conductance was markedly reduced up to 0.1 FM, although the stable GABA-induced hyperpolarization remained unaffected or else slightly enhanced. Such disparity in GABA-mediated chloride channels has been recently detected in unitary conductances analyzed in dissociated adult cockroach neurons [22]. The Ptxresistant GABA response has been reversed at a mean value of - 100 mV close to the equilibrium potential for potassium ions. This was illustrated in Figure 6 for a representative experiment. GABA Receptor Multiplicity
The vertebrate GABABreceptor was originally defined as being activated by baclofen and unaffected by the convulsant alkaloid, bicuculline [23,24]. In the insect CNS, biochemical [25] and electrophysiologicalexperiments performed on identified soma1 neurons [26] provided evidence that baclofen was inactive on GABA receptors, suggesting that the insect CNS does not possess GABAB receptors as pharmacologically identified in vertebrates. Using a double-barreled micropipette filled with 1 mM GABA and 1 mM baclofen, both were injected one after another within the A6 neuropile in order to compare their putative effects. Results summarized in Figure 7 show that baclofen does not mimic the Ptx-resistant GABA response whatever the sequence of application. It can be concluded that baclofen is inactive in reproducing the Ptx-resistant GABA response. Further investigations were made using a highly selective and potent GABAB agonist, 3-aminopropyl-phosphonousacid (CGA 147823). This agonist belongs to the group of y-aminopropyl-phosphonous acids, which bind with a high
Cockroach GABA Receptor Subtypes
153
control
10pM P t x
O.lmM Ptx
I 2 0 mY
20
5
Fig. 5. Effects of Ptx on CABA-mediated responses appreciated on resting potential and membrane conductance changes measured as hyperpolarizing current pulses (downward deflection) applied through the postsynaptic membrane. Note that in control conditions, long-lasting microapplication of GABA (black bar) develops a membrane conductance increase reaching a peak corresponding to the top of the transient depolarization before stabilization to a lower level. Ptx (10FM) suppresses both the peak conductance change and the transient depolarization uncovering a long-lasting hyperpolarization accompanied by a stable decrease i n membrane conductance. The long-lasting hyperpolarization is preserved (Ptx-resistantCABA response) up to 0.1 m M Ptx although changes in membrane conductance are drastically reduced.
B
A = 2
I
b
V
20s
-3
I
Fig. 6. Effect of changing the resting potential (Vm) on the Ptx-resistant CABA response. A: Recordings of CABA responses indicated in presence of 0.1 m M Ptx. B: Plot of amplitude of the GABA response vs. the Vm value. Recordings are issued from a representative experiment. Injection of CABA i s indicated by the black triangle.
154
Hue
control
+m Baclofen
GABA
r Baclofen
GABA
1 2mV 20s Fig. 7. Comparative effects of microinjection of CABA and baclofen (double-barreled micropipette) on the resting potential of the GI2 in control conditions and after Ptx application. Note that baclofen was unable to induce a hyperpolarization as was seen with GABA less than 0.1 mM Ptx treatment. Injections of GABA and baclofen are indicated by black squares.
specificity to GABAB receptors of vertebrate CNS [27]. It has been recently reported that this agonist inhibits the heart beat of the marine arachnid Limulus polyphemus, but fails to activate the locust thoracic soma1 GABA receptors [161. Using the double-barreled micropipette described above, a hyperpolarizing wave was readily induced in response to microinjection of 1mM CGA 147823 either in control conditions or in presence of 10-100 pM Ptx (see Fig. 8). Moreover, a microinjection of GABA during the microapplication of CGA 147823 produced a small hyperpolarization followed by a slow depolarizing wave indicating that CGA 147823 does not interact with the GABA receptor-chloride ionophore complex.
A
A
I
GABA
I
CGA 147823
I CGA 147823+GABA
B Ptx
G-J I
GABA
I
20
5
I
CGA 147823
Fig. 8. Comparative effects of CABA and CGA 147823 (indicated by black bars) on the resting potential in normal conditions (A: saline) and on Ptx treatment (B).The hyperpolarization induced by microinjection of CCA 147823 persists after Ptx treatment of the A6 ganglion and mimics the action of GABA. Note that CCA 147823 microapplication does not prevent the development of a biphasic CABA response in normal conditions.
Cockroach CABA Receptor Subtypes
155
DISCUSSION
Multiphasic GABA response have been described in mammalian pyramidal neurons [28-301 and in cultured mouse spinal neurons [31]. However, the present results are the first to identify regional differences in GABA receptor distribution in insect CNS dendrites and to characterize receptor-channel coupling both physiologically and pharmacologically. Three components, termed fast hyperpolarization, transient depolarization, and slow hyperpolarization of various amplitude were observed in response to either pressure injection or external application of GABA.These results indicate: first, a certain heterogeneity in the nature of GABA receptors yielding the total response, and second, to account for the major expression of the fast hyperpolarizing phase close to the center of the dendritic tree, a predominant localization of "fast" hyperpolarizing GABA receptors on the main branches of the GI dendritic tree. Furthermore, it has been demonstrated that the shape of the GABA response was also directly correlated to the dose of GABA injected or externally applied. Spreading of pressure-injected GABA could account for this dose-dependent effect. However, a similar phenomenon was observed by increasing doses of GABA externally applied, leading to the conclusion that GABA binds to two classes of GABA receptors which display different densities and affinities according to their distribution on the dendritic tree. The reversal potentials of the three components of the GABA-mediated response indicate that GABA gates differentionic currents. The fast hyperpolarizing phase reversed close to Ea= EIpsr [11,19] indicating that a chloride ion current is involved in generating the IPSP. This agrees well with the preceding observations performed on the same neuron [12] for which the amplitude of pure GABA-mediated hyperpolarization was correlated to the external chloride concentration, Thus, synaptic inhibitory neurotransmission within the neuropile of the A6 ganglion could be mediated by such GABA receptors linked to chloride channels inducing fast hyperpolarization. Referring to the classification adopted on rat hippocampal pyramidal cells [281, extrasynaptic receptors could be implicated in the production of the transient depolarizing phase. Its reversal potential ( - 70 mV) does not fit with EIpsp ( - 82 mV) but the large conductance increase, the high sensitivity to Ptx, and the dependence upon external chloride concentration confirm that these GABA receptorsgate chlorideion channels. The differentvalues ( - 70 mV and - 82 mV) of these reversal potentials would require, as it has been suggested on rat pyramidal cells (281, different types of transmembrane chloride pump mechanisms producing a non-homogeneous repartition of chloride ions within the different parts of the dendritic branches. At least, two receptor subtypes of GABAgated chloride channels can be distinguished on account of differences in chloride conductance increase, GABA, and Ptx sensitivity and distribution on the dendrites. Doses of baclofen 1,000 times greater than those active on vertebrate GABAB receptors are inactive on insect soma1 GABA receptors [26]. Thus, it was concluded that no GABAB receptors exist on these cells. Experiments reported here confirm the insensitivity of insect GABA receptors to baclofen. However, the new and potent GABABagonist CGA 147823 employed in this study was
156
Hue
able to reproduce the Ptx-resistant GABA response even in presence of 0.1 mM Ptx, suggesting that CGA 147823 works at a site remote from the GABA receptor chloride ionophore complex. Although no convincing results were obtained by increasing the potassium content of the saline, it could be postulated, that the Ptx-resistant GABA response is potassium ion dependent. The present study therefore suggests that insect CNS possesses a class of GABA receptors which display ionic and certain pharmacological characteristics of GABABreceptors in vertebrate CNS [29]. LITERATURE CITED 1. Sattelle DB: GABA receptors of insects. Adv Insect Physiol22,l (1990). 2. Hue B, Pelhate M, Chanelet J: Pre- and postsynaptic effects of taurine an GABA in the cockroach central nervous system. Can J Neurol Sci 6,243 (1979). 3. Roberts CJ, Krosgaard-Larsen P, Walker RJ: Studies of the action of GABA, muscimol and related compounds on Periplaneta and Limtilus neurons. Comp Biochem Physiol69C, 7 (1981). 4. Wafford KA, Sattelle DB: Effects of amino-acid neurotransmitter candidates on an identified insect motoneurone. Neurosci Lett 63, 135 (1986). 5. Lees G, Beadle DJ, Neumann R, Benson J: Responses to GABA by isolated insect neuronal somata: pharmacology and modulation by a benzodiazepine and a barbiturate. Brain Res 402,267 (1987). 6. Lummis SCR, Sattelle DB: Binding sites for y-aminobutyric acid and benzodiazepines in the central nervous system of insects. Pestic Sci 16,695 (1985). 7. Benson JA: Bicuculline blocks the response to acetylcholine and nicotine but not to muscarine or GABA in isolated insect neuronal somata. Brain Res 458,65 (1988). 8. Buckingham S, Sattelle DB, Hue 6: Synaptic and extrasynaptic actions of bicuculline on identified insect neurones. J Exp Biol (In press). 9. Kerkut GA, Pitman RM, Walker RJ: Iontophoretic application of acetylcholine and GABA onto insect central neurones. Comp Biochem Physiol31,611(1969). 10. Shimahara T, Pichon Y, Lees G, Beadle CA, Beadle DJ: Gamma-aminobutyric acid receptors on cultured cockroach brain neurones. j Exp Biol131,231 (1987). 11. Callec jJ: Synaptic transmission in the central nervous system of insects. In: Insect Neurobiology. Treherne JE, ed. North-Holland American Elsevier, New York, pp 119-185 (1974). 12. Callec JJ: Synaptic transmission in the central nervous system. In: Comprehensive Insect Physiology, Biochemistry and Pharmacology. Kerkut GA,Gilbert LI, eds. Pergamon Press, London, Vol. 5, pp 139-179 (1985). 13. Hue B, Callec JJ: Presynaptic inhibition in the cercal-afferent giant interneurone synapses of the cockroach, Periplaneta americana. j Insect Physiol29, 741 (1983). 14. Malecot CO, Hue B, Buckingham SD, Sattelle DB: GABA receptors in the nervous system of an insect Periplaneta americana. Proceedings of the 1st International GABABSymposium, Cambridge, p 81 (1989). 15. Piek T, Hue B, Mony L, Nakajima T, Pelhate M, Yasuhara T: Block of synaptic transmission in insect CNS by toxins from the venom of the wasp Meguscoliaflwifrons (fab.). Comp Biochem Physiol87C, 287 (1987). 16. Benson JA: A novel GABA receptor in the heart of a primitive arthropod Limulus polyphemus. J Exp Bioll47,421 (1989). 17. Harrow ID, Hue B, Pelhate M, Sattelle DB: Cockroach giant interneurones stained by cobaltbackfilling of dissected axons. J Exp Biol84, 341 (1980). 18. Hue B, Gabriel A, Le Patezour A Autoradiographic localization of [3H]-GABA accurnulation in the sixth abdominal ganglion of the cockroach, Periplaneta americana L. J Insect Physiol 28, 753 (1982). 19. Hue B: Electrophysiologie et pharmacologie de la transmission synaptique dans le systeme nerveux central de la blatte, Periplaneta americana. T h h e d’Etat, Universite Angers (1983). 20. Ticku M, Olsen RW: y-Aminobutyric-stimulated chloride permeability in crayfish muscle. Biochem Biophys Acta 464,519 (1977).
Cockroach GABA Receptor Subtypes
157
21. Constanti A: The "mixed" effect of picrotoxin on the GABA dosekonductance relation recorded from lobster muscle. Neuropharmacology 17,159 (1978). 22. Malecot CO, Sattelle DB: Single channel recordings from insect neuronal GABA-activated chloride channels. J Exp Biol151,495 (1990). 23. Hill DR, Bowery NG: 3H-bacoflenand 3H-GABA bind to bicuculline-insensitive GABABbinding sites in rat brain. Nature 290, 149 (1981). 24. Bowery NG, Doble A, Hill DR, Hudson AL, Shaw JS, Turnbell MJ, Warrington R: Bicucullineinsensitive GABA receptors on peripheral autonomic nerve terminals. Eur J Pharmacol71, 53 (1981). 25. Turner AJ, Whittle SR: Biochemical dissection of the aminobutyrate synapse. Biochem J 209, 29 (1983). 26. Sattelle 8, Pinnock RD, Wafford KA, David JA: GABA receptors on the cell-boy membrane of an identified insect motor neuron. Proc R SOCLond B232,443 (1988). 27. Dingwall JG, Ehrenfreund J, Hall RG, Jack J: Synthesis of y-aminopropylphosphonous acids using hypophosphorous acid synthons. Phosphorus Sulfur 30,571 (1987). 28. Alger BE, Nicoll RA: Pharmacological evidence for two kinds of GABA receptor on rat hippocampal pyramidal cells studied in vitro. J Physiol328,125 (1982). 29. Newberry NR, Nicoll RA: Comparison of the action of baclofen with y-aminobutyric acid on rat hippocampal pyramidal cells in vitro. J Physiol360,161 (1985). 30. Connors BW, Malenka RC, Silva LR: Two inhibitory postsynaptic potentials, andd GAB& and GABAHreceptor-mediated responses in neocortex of rat and cat. J Physiol406,443 (1988). 31. Barker JL, Harrison NL, Owen DG: Pharmacology and Physiology of C1- conductances activated by GABA in cultured mammalian central neurons. In: Chloride Channels and Carriers in Nerve, Muscle and Glial Cells. Alavarez-LeefrnansFJ, Russell JM, eds. Plenum Press, New York and London, pp 237-297 (1990).
Functional Assay for GABA Receptor Subtypes of a Cockroach Giant lnterneuron Bernard Hue Laboratoire de Newrophysiologie, Cedex, France
URA CNRS 622 UnzversitC d’Angers, 49045 Angers
-
y Aminobutyric acid (GABA) receptors were examined in the cockroach central nervous system (CNS) using the single fiber-oil gap method applied to an identified giant interneuron. Short-lasting pressure application of 10 rnM GABA developed a multiphasic response composed of a fast hyperpolarization followed by a transient depolarizing component and a stable hyperpolarization. This triphasic characteristic shape of the response was modified according to the dose of GABA injected or bath-applied and to the precise localization of the injection within the dendritic area. The transient depolarizing phase showed a negative reversal potential of -70 mV. Both hyperpolarizing phases reversed at a more negative level ranging to - 80 mV. A positive shift of these values was caused by a decrease in external chloride concentration. Bath-application of 0.1 m M picrotoxin (Ptx) decreased the depolarizing phase which was progressively replaced by a stable hyperpolarization. The transient depolarizing component desensitized quickly and was the most sensitive phase to Ptx action. The Ptx-resistant response reversed at a mean value of -100 rnV close to the equilibrium potential for potassium ions (EKC), suggesting that it was generated by a K+-channel coupled receptor. Although baclofen was unable to mimic the Ptx-resistant GABA response, the compound CGA 147823, known to bind with a high specificity to vertebrate GABAHreceptors, has been successfully used to reproduce the Ptx-resistantGABA response. It i s suggested that, in addition to GABA receptors linked to chloride channels, the insect CNS possesses GABA receptors sharing ionic characteristics of GABAHreceptors especially those located i n the vertebrate CNS, although they are insensitive to baclofen. Key words: insect CNS, synaptic receptor, fiber-oil gap method
Acknowledgments: 8. Hue thanks CIBA-GEIGY for the generous supply of CGA 147823 and
T. Piek for his helpful comments. Received March 21,1991; accepted June10,1991. Address reprint requests to Professor Bernard Hue, Laboratoire de Neurophysiologie, URA CNRS 611, FacultC de Medecine, rue haute de reculee, 49045 Angers Cedex, France. 0 1991 Wiley-Liss, Inc.
148
Hue
INTRODUCTION
The inhibitory neurotransmitter, GABA", is widely distributed in the insect CNS. The existence of only one type of CABA receptor resembling the vertebrate GABAA receptor has been recently reviewed [l]; however, some disparity in GABA receptor distribution or ionic mechanisms has been previously reported [2-41. The insect's GABA receptor is linked to a Ptx-sensitivechlorideion channel and possesses modulatory sites for benzodiazepines and barbiturates 151, but its insensitivity to bicuculline 16-81 supports the views that its pharmacological properties differ from either of the vertebrate GABA receptor subtypes. Electrophysiological and pharmacological studies on GABA receptors have been mainly performed on neuronal cell bodies studied in situ [4,9] or in culture [5,10]. Moreover, GABA has been identified as an inhibitory neurotransmitter at the cercal-nerve GI synapse in the A6 ganglion of the cockroach, Peripluneta arnericuna [11,12].GABA receptors are implicated in presynaptic and postsynaptic inhibition [11,13]. The present study extends previous investigations [8, 141 carrried on cercalnerve GI synapses, This preparation allows examination of responses induced by pressure application of drugs [15]. So, we have recorded membrane potential changes due to pressure application of GABA within the neuropile of the cockroach's A6 ganglion. The induced multiphasic GABA response overlaps a Ptx-resistant hyperpolarization which has been reproduced with 3-aminopropyl-phosphonous acid (CGA 147823), a new potent GABABagonist [16]. MATERIALS AND METHODS
Single fiber-oil gap experiments were performed using the cercal-nerve GI synapses located within the neuropile of the A6 ganglion of the CNS of the cockroach P. urnericuna [11]. All experimental data were obtained on the GI no. 2 visually identified during microdissection. Anatomical details of the dendritic tree have been revealed using the cobalt blackfilling method described previously on the same preparation [17]. Postsynaptic resting membrane potential was continuously monitored on a pen-chart recorder. Changes in postsynaptic membrane resistance were measured by modifications to hyperpolarizing square current pulses applied through the membrane by means of a balanced Wheatstone bridge connected in the recording circuit. The A6 ganglion was desheathed and superfused at a constant rate with a saline containing (in mM): NaCl208; KCl3.1; CaClz 10.0; NaHC03 2.0; sucrose 26; pH 7.2. In a series of experiments, the chloride content of the normal saline was reduced by replacement of the sodium chloride with equimolar sodium acetate or propionate. GABA and related agonists were injected within the A6 neuropile onto the GI dendrites using broken micropipettes (tip diameter 20 pm) connected to a pneumatic pressure ejecting system (Neuro Phore BH-2, Medical System Corp., Greenvale, NY, USA) as described earlier [15]. Direct microapplication of GABA to the dendrites was preferentially used on account of *Abbreviations used: A6 = Sixth abdominal; CNS = central nervous system; EPSP = excitatory postsynaptic potential; CABA = y-aminobutyric acid; GI = giant interneuron; IPSP = inhibitory postsynaptic potential; Ptx = picrotoxin; l 7 X = tetrodotoxin.
Cockroach CABA Receptor Subtypes
149
the high peripheral glial uptake of GABA [18]. Furthermore, pressure injections allow precise local applications of drugs and make easier the reversibility of the induced effects. In control experiments, pressure microapplications of saline never produced any effects on the synaptic transmission. All experiments were performed at room temperature (20°C). The antagonists (Ptx, baclofen) were dissolved in the saline and externally applied when necessary. Pharmacological compounds were obtained from Sigma Chemical Co. (St Louis, MO, USA). CGA 147823 was kindly provided by CIBA-GEIGY Ltd., (Basel, Switzerland) for use in this study. RESULTS Physiological Effects of Pressure AppIication of GABA Short-lasting microapplication (300 ms, 15 psi) of GABA at concentrations ranging from 0.1-1 mM within the A6 ganglion neuropile (close to the main dendritic branches of the GI) induced a triphasic response in the GI postsynaptic membrane. A fast hyperpolarization preceded a slow depolarizing wave followed by a slow hyperpolarizing phase (Fig. 1A). Concurrently, amplitude of unitary EPSPs was drastically decreased [2]. In this paper no attempt was made to further analyze this latter effect or changes in postsynaptic membrane A
C
B
control
c
I
A
A
5 Psi
10 P s i
15 Psi
Dv4fPrlc1 A
A
A
Fig. 1. Effects of CABA on the resting potential of the GI2 postsynaptic membrane. A, C, D: 1 mM CABA solution was applied by pressure injection (indicated by black triangles) within the neuropile of the A6 ganglion. B: 20 mM CABA was externally applied (black bar) onto the A6 ganglion. Temperature of the bath was lowered to 10°C in order to prevent excessive GABA uptake. Note in C that TTX does not suppress the action of GABA and in the dose-dependent shape of the response when the injection pressure is increased. In order to prevent desensitization of GABA receptors, the microapplications of GABA in D were separated by a period of time of 20 min. Vertical scale: A, B, C, I mV; 0,2 mV; horizontal scale: AC, 15 s; B , 2 min; D, 30 s.
150
Hue
conductance which have been previously studied [2,12]. Similar responses were also observed when a high concentration (10 mM) of GABA was externally applied (Fig. 1B) at a relatively low temperature (10OC) in order to reduce glial uptake of GABA [MI.Obviously in the latter case, the time course of the different waves was slower. These similarities show that the mode of application of GABA does not interfere with the main shape of the response. In the presence of 1 pM TTX, the triphasic GABA response was not altered as shown in Figure 1C, suggesting that GABA acted directly on synaptic receptors of the GI no. 2 under test. Further investigations were carried out in order to more precisely establish the dose dependence of the response. Figure 1D shows that low doses of GABA injected during 300 ms at 5 psi produced only a small hyperpolarization. Higher doses (10 and 15 psi) activated a transient depolarizing component which was more pronounced at higher pressure applications. Increased doses of GABA, applied through the superfusing saline, produced comparable effects (not illustrated). Dendritic CABA Responses
The multiphasic GABA response, as previously described, was only recorded when the tip of the injection micropipette was carefully positioned close to the center of the dendritic area (Fig. 2), i.e., close to the initial part of the main dendritic branches; whereas, at the periphery of the dendritic tree the GABA response becomes mainly depolarizing. No effect was seen when GABA
P
a
-
200pm
15 s
Fig. 2. Dependence of GABA responses on the application site. Left: A camera lucida drawing of the C12, filled with cobalt chloride, within the neuropile of the A6 ganglion i s represented. A trachea lying on the ventral side of the ganglion was used as a landmark for the precise microinjection of GABA. Both cercal nerves (p: posterior part of the ganglion) and connectives were cut. Right: Four traces correspond to the microinjection of GABA indicated by the black triangle (pulse duration, 300 rns; pressure, 15 psi; depth, 200 bm) at four different sites localized by the place of a micropipette. Note the lack of GABA response when the injection is done within the anterior part (a) of the neuropile, close to the initial part of the giant axon.
Cockroach CABA Receptor Subtypes
151
A
L 15 5 -80nV
-4
1
Fig. 3. Effect of shift of the resting potential on the shape of the triphasic response induced by a 300 ms CABA pulse (black triangle). The control resting potential ( - 75 mv) was evaluated after full depolarization of the postsynaptic membrane by superfusion with saline containing 210 m M KCI. A: The amplitudes were measured at the peak of each phase (indicated by A,0, 0). 6: The amplitudeswere plotted on the graph vs. the shift of the resting potential.
was applied within the anterior part of the A6 ganglion containing the initial segment of the giant axon. This latter observation indicates that GABA receptors are restricted to the GI dendritic tree. Earlier studies [11,12,19] performed on the same GI system have demonstrated that IPSPs were due to the release of GABA from cercal-nerve X endings. Changing the chloride concentration in the saline could reverse the hyperpolarization induced by bath-applied GABA as well as evoked IPSP. The reversal potential of these two responses was similar, around - 80 mV, suggesting that they were supported by the same chloride current. Further experiments were performed in an attempt to precisely determine the ionic mechanisms underlying the three components of the GAEL4 response. By passing hyperpolarizing current through the membrane the reversal potential of the GABA waves has been estimated. As indicated on the graph (Fig. 3) the depolarizing phase had the least negative reversal potential ( - 70 mV), both hyperpolarizing waves reversed at a more negative level ( - 79 mV and - 80 mV). These reversal potential values were consistent between GI no. 2 tested (n = 5). Lowering the amount of chloride in the saline (222-166 mM) caused both GABA hyperpolarizing phases to fade whereas the depolarizing phase was enhanced (Fig. 4A). Finally, Ptx, the well-known GABA-dependent chloride ionophore blocker [20,21], was bath-applied at 0.1 mM which fully blocked the chloride-dependent IPSP [11,19]. The time-course of the Ptx action on the triphasic GABA response was not quantified and is not presented here, but a slow decrease of the depolarizing phase occurred and was progressively replaced by a unique hyperpolarizing response which was resistant to Ptx action (Fig. 4B). During long-lasting (30 s) microapplication of GABA, postsynaptic mem-
152
Hue
A
6 control
n
mV Fig. 4. Effects of low chloride saline, 110 mM (A) and 0.1 mM Ptx (6) on the triphasic response induced by a 300 ms CABA pulse (black squares). Note that Ptx suppresses the depolarizing phase.
brane conductance was measured as changes in short (50ms) membrane hyperpolarizations. It was observed (Fig. 5) that conductance changes paralleled changes in resting membrane potential, i. e., the maximal conductance increase coincided with the peak of the transient depolarization and a steady-state value was reached. Increasing doses of Ptx (up to 10 pM) primarily affected the maximal conductance concomitant with the peak of the transient depolarizingphase without apparent effect on the steady-state. The steady-state conductance was markedly reduced up to 0.1 FM, although the stable GABA-induced hyperpolarization remained unaffected or else slightly enhanced. Such disparity in GABA-mediated chloride channels has been recently detected in unitary conductances analyzed in dissociated adult cockroach neurons [22]. The Ptxresistant GABA response has been reversed at a mean value of - 100 mV close to the equilibrium potential for potassium ions. This was illustrated in Figure 6 for a representative experiment. GABA Receptor Multiplicity
The vertebrate GABABreceptor was originally defined as being activated by baclofen and unaffected by the convulsant alkaloid, bicuculline [23,24]. In the insect CNS, biochemical [25] and electrophysiologicalexperiments performed on identified soma1 neurons [26] provided evidence that baclofen was inactive on GABA receptors, suggesting that the insect CNS does not possess GABAB receptors as pharmacologically identified in vertebrates. Using a double-barreled micropipette filled with 1 mM GABA and 1 mM baclofen, both were injected one after another within the A6 neuropile in order to compare their putative effects. Results summarized in Figure 7 show that baclofen does not mimic the Ptx-resistant GABA response whatever the sequence of application. It can be concluded that baclofen is inactive in reproducing the Ptx-resistant GABA response. Further investigations were made using a highly selective and potent GABAB agonist, 3-aminopropyl-phosphonousacid (CGA 147823). This agonist belongs to the group of y-aminopropyl-phosphonous acids, which bind with a high
Cockroach GABA Receptor Subtypes
153
control
10pM P t x
O.lmM Ptx
I 2 0 mY
20
5
Fig. 5. Effects of Ptx on CABA-mediated responses appreciated on resting potential and membrane conductance changes measured as hyperpolarizing current pulses (downward deflection) applied through the postsynaptic membrane. Note that in control conditions, long-lasting microapplication of GABA (black bar) develops a membrane conductance increase reaching a peak corresponding to the top of the transient depolarization before stabilization to a lower level. Ptx (10FM) suppresses both the peak conductance change and the transient depolarization uncovering a long-lasting hyperpolarization accompanied by a stable decrease i n membrane conductance. The long-lasting hyperpolarization is preserved (Ptx-resistantCABA response) up to 0.1 m M Ptx although changes in membrane conductance are drastically reduced.
B
A = 2
I
b
V
20s
-3
I
Fig. 6. Effect of changing the resting potential (Vm) on the Ptx-resistant CABA response. A: Recordings of CABA responses indicated in presence of 0.1 m M Ptx. B: Plot of amplitude of the GABA response vs. the Vm value. Recordings are issued from a representative experiment. Injection of CABA i s indicated by the black triangle.
154
Hue
control
+m Baclofen
GABA
r Baclofen
GABA
1 2mV 20s Fig. 7. Comparative effects of microinjection of CABA and baclofen (double-barreled micropipette) on the resting potential of the GI2 in control conditions and after Ptx application. Note that baclofen was unable to induce a hyperpolarization as was seen with GABA less than 0.1 mM Ptx treatment. Injections of GABA and baclofen are indicated by black squares.
specificity to GABAB receptors of vertebrate CNS [27]. It has been recently reported that this agonist inhibits the heart beat of the marine arachnid Limulus polyphemus, but fails to activate the locust thoracic soma1 GABA receptors [161. Using the double-barreled micropipette described above, a hyperpolarizing wave was readily induced in response to microinjection of 1mM CGA 147823 either in control conditions or in presence of 10-100 pM Ptx (see Fig. 8). Moreover, a microinjection of GABA during the microapplication of CGA 147823 produced a small hyperpolarization followed by a slow depolarizing wave indicating that CGA 147823 does not interact with the GABA receptor-chloride ionophore complex.
A
A
I
GABA
I
CGA 147823
I CGA 147823+GABA
B Ptx
G-J I
GABA
I
20
5
I
CGA 147823
Fig. 8. Comparative effects of CABA and CGA 147823 (indicated by black bars) on the resting potential in normal conditions (A: saline) and on Ptx treatment (B).The hyperpolarization induced by microinjection of CCA 147823 persists after Ptx treatment of the A6 ganglion and mimics the action of GABA. Note that CCA 147823 microapplication does not prevent the development of a biphasic CABA response in normal conditions.
Cockroach CABA Receptor Subtypes
155
DISCUSSION
Multiphasic GABA response have been described in mammalian pyramidal neurons [28-301 and in cultured mouse spinal neurons [31]. However, the present results are the first to identify regional differences in GABA receptor distribution in insect CNS dendrites and to characterize receptor-channel coupling both physiologically and pharmacologically. Three components, termed fast hyperpolarization, transient depolarization, and slow hyperpolarization of various amplitude were observed in response to either pressure injection or external application of GABA.These results indicate: first, a certain heterogeneity in the nature of GABA receptors yielding the total response, and second, to account for the major expression of the fast hyperpolarizing phase close to the center of the dendritic tree, a predominant localization of "fast" hyperpolarizing GABA receptors on the main branches of the GI dendritic tree. Furthermore, it has been demonstrated that the shape of the GABA response was also directly correlated to the dose of GABA injected or externally applied. Spreading of pressure-injected GABA could account for this dose-dependent effect. However, a similar phenomenon was observed by increasing doses of GABA externally applied, leading to the conclusion that GABA binds to two classes of GABA receptors which display different densities and affinities according to their distribution on the dendritic tree. The reversal potentials of the three components of the GABA-mediated response indicate that GABA gates differentionic currents. The fast hyperpolarizing phase reversed close to Ea= EIpsr [11,19] indicating that a chloride ion current is involved in generating the IPSP. This agrees well with the preceding observations performed on the same neuron [12] for which the amplitude of pure GABA-mediated hyperpolarization was correlated to the external chloride concentration, Thus, synaptic inhibitory neurotransmission within the neuropile of the A6 ganglion could be mediated by such GABA receptors linked to chloride channels inducing fast hyperpolarization. Referring to the classification adopted on rat hippocampal pyramidal cells [281, extrasynaptic receptors could be implicated in the production of the transient depolarizing phase. Its reversal potential ( - 70 mV) does not fit with EIpsp ( - 82 mV) but the large conductance increase, the high sensitivity to Ptx, and the dependence upon external chloride concentration confirm that these GABA receptorsgate chlorideion channels. The differentvalues ( - 70 mV and - 82 mV) of these reversal potentials would require, as it has been suggested on rat pyramidal cells (281, different types of transmembrane chloride pump mechanisms producing a non-homogeneous repartition of chloride ions within the different parts of the dendritic branches. At least, two receptor subtypes of GABAgated chloride channels can be distinguished on account of differences in chloride conductance increase, GABA, and Ptx sensitivity and distribution on the dendrites. Doses of baclofen 1,000 times greater than those active on vertebrate GABAB receptors are inactive on insect soma1 GABA receptors [26]. Thus, it was concluded that no GABAB receptors exist on these cells. Experiments reported here confirm the insensitivity of insect GABA receptors to baclofen. However, the new and potent GABABagonist CGA 147823 employed in this study was
156
Hue
able to reproduce the Ptx-resistant GABA response even in presence of 0.1 mM Ptx, suggesting that CGA 147823 works at a site remote from the GABA receptor chloride ionophore complex. Although no convincing results were obtained by increasing the potassium content of the saline, it could be postulated, that the Ptx-resistant GABA response is potassium ion dependent. The present study therefore suggests that insect CNS possesses a class of GABA receptors which display ionic and certain pharmacological characteristics of GABABreceptors in vertebrate CNS [29]. LITERATURE CITED 1. Sattelle DB: GABA receptors of insects. Adv Insect Physiol22,l (1990). 2. Hue B, Pelhate M, Chanelet J: Pre- and postsynaptic effects of taurine an GABA in the cockroach central nervous system. Can J Neurol Sci 6,243 (1979). 3. Roberts CJ, Krosgaard-Larsen P, Walker RJ: Studies of the action of GABA, muscimol and related compounds on Periplaneta and Limtilus neurons. Comp Biochem Physiol69C, 7 (1981). 4. Wafford KA, Sattelle DB: Effects of amino-acid neurotransmitter candidates on an identified insect motoneurone. Neurosci Lett 63, 135 (1986). 5. Lees G, Beadle DJ, Neumann R, Benson J: Responses to GABA by isolated insect neuronal somata: pharmacology and modulation by a benzodiazepine and a barbiturate. Brain Res 402,267 (1987). 6. Lummis SCR, Sattelle DB: Binding sites for y-aminobutyric acid and benzodiazepines in the central nervous system of insects. Pestic Sci 16,695 (1985). 7. Benson JA: Bicuculline blocks the response to acetylcholine and nicotine but not to muscarine or GABA in isolated insect neuronal somata. Brain Res 458,65 (1988). 8. Buckingham S, Sattelle DB, Hue 6: Synaptic and extrasynaptic actions of bicuculline on identified insect neurones. J Exp Biol (In press). 9. Kerkut GA, Pitman RM, Walker RJ: Iontophoretic application of acetylcholine and GABA onto insect central neurones. Comp Biochem Physiol31,611(1969). 10. Shimahara T, Pichon Y, Lees G, Beadle CA, Beadle DJ: Gamma-aminobutyric acid receptors on cultured cockroach brain neurones. j Exp Biol131,231 (1987). 11. Callec jJ: Synaptic transmission in the central nervous system of insects. In: Insect Neurobiology. Treherne JE, ed. North-Holland American Elsevier, New York, pp 119-185 (1974). 12. Callec JJ: Synaptic transmission in the central nervous system. In: Comprehensive Insect Physiology, Biochemistry and Pharmacology. Kerkut GA,Gilbert LI, eds. Pergamon Press, London, Vol. 5, pp 139-179 (1985). 13. Hue B, Callec JJ: Presynaptic inhibition in the cercal-afferent giant interneurone synapses of the cockroach, Periplaneta americana. j Insect Physiol29, 741 (1983). 14. Malecot CO, Hue B, Buckingham SD, Sattelle DB: GABA receptors in the nervous system of an insect Periplaneta americana. Proceedings of the 1st International GABABSymposium, Cambridge, p 81 (1989). 15. Piek T, Hue B, Mony L, Nakajima T, Pelhate M, Yasuhara T: Block of synaptic transmission in insect CNS by toxins from the venom of the wasp Meguscoliaflwifrons (fab.). Comp Biochem Physiol87C, 287 (1987). 16. Benson JA: A novel GABA receptor in the heart of a primitive arthropod Limulus polyphemus. J Exp Bioll47,421 (1989). 17. Harrow ID, Hue B, Pelhate M, Sattelle DB: Cockroach giant interneurones stained by cobaltbackfilling of dissected axons. J Exp Biol84, 341 (1980). 18. Hue B, Gabriel A, Le Patezour A Autoradiographic localization of [3H]-GABA accurnulation in the sixth abdominal ganglion of the cockroach, Periplaneta americana L. J Insect Physiol 28, 753 (1982). 19. Hue B: Electrophysiologie et pharmacologie de la transmission synaptique dans le systeme nerveux central de la blatte, Periplaneta americana. T h h e d’Etat, Universite Angers (1983). 20. Ticku M, Olsen RW: y-Aminobutyric-stimulated chloride permeability in crayfish muscle. Biochem Biophys Acta 464,519 (1977).
Cockroach GABA Receptor Subtypes
157
21. Constanti A: The "mixed" effect of picrotoxin on the GABA dosekonductance relation recorded from lobster muscle. Neuropharmacology 17,159 (1978). 22. Malecot CO, Sattelle DB: Single channel recordings from insect neuronal GABA-activated chloride channels. J Exp Biol151,495 (1990). 23. Hill DR, Bowery NG: 3H-bacoflenand 3H-GABA bind to bicuculline-insensitive GABABbinding sites in rat brain. Nature 290, 149 (1981). 24. Bowery NG, Doble A, Hill DR, Hudson AL, Shaw JS, Turnbell MJ, Warrington R: Bicucullineinsensitive GABA receptors on peripheral autonomic nerve terminals. Eur J Pharmacol71, 53 (1981). 25. Turner AJ, Whittle SR: Biochemical dissection of the aminobutyrate synapse. Biochem J 209, 29 (1983). 26. Sattelle 8, Pinnock RD, Wafford KA, David JA: GABA receptors on the cell-boy membrane of an identified insect motor neuron. Proc R SOCLond B232,443 (1988). 27. Dingwall JG, Ehrenfreund J, Hall RG, Jack J: Synthesis of y-aminopropylphosphonous acids using hypophosphorous acid synthons. Phosphorus Sulfur 30,571 (1987). 28. Alger BE, Nicoll RA: Pharmacological evidence for two kinds of GABA receptor on rat hippocampal pyramidal cells studied in vitro. J Physiol328,125 (1982). 29. Newberry NR, Nicoll RA: Comparison of the action of baclofen with y-aminobutyric acid on rat hippocampal pyramidal cells in vitro. J Physiol360,161 (1985). 30. Connors BW, Malenka RC, Silva LR: Two inhibitory postsynaptic potentials, andd GAB& and GABAHreceptor-mediated responses in neocortex of rat and cat. J Physiol406,443 (1988). 31. Barker JL, Harrison NL, Owen DG: Pharmacology and Physiology of C1- conductances activated by GABA in cultured mammalian central neurons. In: Chloride Channels and Carriers in Nerve, Muscle and Glial Cells. Alavarez-LeefrnansFJ, Russell JM, eds. Plenum Press, New York and London, pp 237-297 (1990).
