- Email: [email protected]
Perpetual inhibitory activity in mammalian brain slices generated by spontaneous GABA release
Brain Research, 545 (1991) 142-150 © 1991 Elsevier Science Publishers B.V. (Biomedical Division) 0006-8993/91/$03.50 ADONIS 000689939116467U
142
BRES 16467
Perpetual inhibitory activity in mammalian brain slices generated by spontaneous GABA release Thomas S. Otis, Kevin J. Staley and Istvan Mody Department of Neurology and Neurological Sciences M016, Stanford University School of Medicine, Stanford, CA 94305 (U.S.A.) (Accepted 23 October 1990)
Key words: Brain slice; Dentate gyrus; ~,-Aminobutyric acid; Hippocampus; Neurotransmitter release; Spontaneous inhibitory postsynaptic current; Synaptic transmission; Whole-cell patch clamp
Miniature spontaneous inhibitory postsynaptic currents (slPSCs) mediated by GABAA receptors were recorded using whole-cell patch clamp recordings in rat brain slices maintained in vitro at 34 _+ 1 °C. We have found that firing of action potentials by principal neurons or by GABAergic interneurons is not necessary to the generation of sIPSCs since they persist in the presence of 1-5/tM tetrodotoxin (TrX). The average frequency of the discrete sIPSCs exhibits a large cell-to-cell variability and is between 5-15 Hz. The amplitudes of the sIPSCs depend on the difference between the membrane potential and the equilibrium potential for Cl- (Eo). Generally, 70-80 mV away from E a , sIPSCs have a mean amplitude of 30-80 pA (i.e. peak conductance of 400-1000 pS) with an average decay time constant of 5.8 ms. Accordingly, unitary single sIPSCs arise from the simultaneous activation of no more than 20 GABA A receptor/channels. The perpetual barrage of spontaneous GABAergic activity is very likely to be a critical factor in the regulation of neuronal excitability and the mechanism of action of several neuroactive compounds. INTRODUCTION N o r m a l functioning of the central nervous system (CNS) requires the existence of a delicate balance between excitation and inhibition. The principal inhibitory n e u r o t r a n s m i t t e r of the m a m m a l i a n CNS is ~aminobutyric acid ( G A B A ) acting via G A B A A and G A B A B receptors 7'15'29m. Relatively small changes in G A B A e r g i c neurotransmission can have striking effects on the functional state of nerve cells, resulting in p r o f o u n d alterations of neuronal excitability39'4°. In fact, m a n y c o m p o u n d s used clinically in the treatment of nervous system disorders alter the activity of G A B A A receptor/channels 7,15,29,39-41,46. D e c a d e s ago, the term synaptic noise was introduced to describe the steady b o m b a r d m e n t of m a m m a l i a n nerve cells by spontaneous excitatory or inhibitory miniature synaptic potentials. A t the time, the origin of synaptic noise was attributed to the release of neurotransmitter substances through continual firing of nerve cells8'16'24. It was later shown that b l o c k a d e of action potential firing did not prevent the occurrence of miniature potentials at certain synapses leading to a re-examination of neurotransmitter release mechanisms in the m a m m a l i a n CNS 3'
Previous studies using sharp m i c r o e l e c t r o d e s to record intracellularly in h i p p o c a m p a l slices have described spontaneous inhibitory postsynaptic potentials 1'28'36 or currents 11"17"21 in chloride-loaded neurons. These studies have concluded that most, if not all, spontaneous inhibitory activity stems from n e u r o n a l action potential firing 1A1'28'36. Once action potentials are abolished, e.g. by the N a ÷ channel blocker t e t r o d o t o x i n (T-FX), spontaneous inhibitory activity b e c o m e s infrequent and of very small amplitude 11. O n e o f the inherent drawbacks while recording with sharp m i c r o e l e c t r o d e s is the presence of a significant leak conductance due to perforation of the neuronal m e m b r a n e . By introducing a relatively large shunt, this conductance m a y r e n d e r small a m p l i t u d e changes in m e m b r a n e current o r voltage undetectable. Tight seal (>109 ~2) patch recordings provide an extremely small a m o u n t of m e m b r a n e leak c o m p a r e d to conventional sharp m i c r o e l e c t r o d e recordings 33. We have a d a p t e d a whole cell patch recording m e t h o d 4'18 to record from brain slices m a i n t a i n e d in a conventional recording c h a m b e r at 34 + 1 °C. The resolution of this m e t h o d has enabled us to investigate previously u n d e t e c t a b l e small amplitude miniature s p o n t a n e o u s inhibitory postsynaptic currents (slPSCs).
12,30
Correspondence: I. Mody, Department of Neurology and Neurological Sciences M016, Stanford University School of Medicine, Stanford, CA 94305, U.S.A.
143 record slPSCs, the following intracellular solutions were used (in mM): (1) KCI 140, MgCI 2 2, HEPES 10; (2) CsCI 140, "MgCI2 2, HEPES 10; (3) K-gluconate 130, KCI 10, MgCI 2 2, HEPES 10. Some intracellular solutions also contained 11 mM EGTA or BAPTA and 1 mM CaC12 which had no noticeable effect on the slPSCs. All intracellular media were adjusted to a pH of 7.2 with KOH or CsOH and had an osmolarity of 250--270 mOsm. Recordings were done using an Axoclamp-2A in continuous SEVC mode or an Axopatch-lD amplifier with >80% series resistance compensation. Data were low pass filtered at 2-3 kHz, digitized at 6-12.5 kHz and analyzed using the Strathclyde Electrophysiology Software (courtesy of J. Dempster).
MATERIALS AND METHODS
Hippocampal slices (400-450/~m) or coronal brain sections (400 ~m) containing the hippocampus were prepared at 2-4 °C from adult (200-350 g) male Wistar rats on a Mcllwain tissue chopper or a Lancer Series 1000 Vibratome respectively. The slices were incubated in a storage chamber at 32 °C for at least 1 h before recordings were started. The recordings were done at 34 + 1 °C in an interface type chamber using a nylon net to cover the slices. The net disrupted the surface tension providing a thin ( < 100/~m) film of fluid over the top of the brain tissue. Oxygenated (95%02/ 5%CO2) artificial cerebrospinal fluid (ACSF) was perfused at a rate of 2.5-3 mi/min. The composition of the ACSF was (in mM): NaCI 126, NaHCO 3 26, NaH2PO 4 1.25, KCI 2.5, CaCI 2 2, MgCI 2 2, o-glucose 10 (303 mOsm). In most cases the ACSF also contained 1-5 ~ M tetrodotoxin (TTX, Calbiochem). Patch-electrodes (2-2.5 /~m tip diameter; 3-5 M.Q) were pulled in two stages from borosilicate thin-walled capillaries (o.d. 1.5 mm; TWl50F-4; WPI) on a Narashige PP-83 electrode puller. In various experiments, the electrodes were filled with different intracellular solutions. To
RESULTS O u r r e c o r d i n g s f r o m d e n t a t e g y r u s g r a n u l e cells ( n = 67), p y r a m i d a l cells o f t h e h i p p o c a m p a l C A 1 r e g i o n (n = 15) o r s o m a t o s e n s o r y c o r t e x ( n = 8), a n d P u r k i n j e cells
Control
1/~M TTX
2S
M e a n A m p l i t u d e - 33.4 * 1 6.1 pA
M e a n A m p l i t u d e - 6 0 . 4 + 40.3 pA a o
~
s
4
0
0
60 •
SM
150
200
0
250
50
•
100 150 Peak Amplitude (pA)
Peak AnNplltuda(pA)
lo I
200
250
10
M e a n - 74.1 ms
Mean - 130.5 m s
'HIII
i--
ae
IlO0 Inter-Event interval I m s )
8oo
ii
o •
200
400
Inler-Event Interval Imsl
Fig. 1. Blocking action potential-dependent neuronal activity with TTX in a dentate gyrus granule cell reduces average slPSC amplitude and frequency, but does not abolish slPSCs. Upper records show inward slPSCs recorded at -55 mV before (left trace) and 10 min after (right trace) perfusion of 1/tM TTX. Note the 30-pA reduction in the mean baseline inward current, presumably resulting from TTX blockade of large synaptic currents and the diminished frequency of TTX-resistant slPSCs, lntracellular solution contained: 130 mM CsCI, 11 mM BAPTA, 10 mM HEPES, 2 mM MgCI2 and 1.1 mM CaC! 2 at a pH of 7.2. The local anesthetic derivative QX314 (10 mM; Astra Pharmaceuticals) was also included in the patch pipette to block breakthrough Na ÷ spikes in the postsynaptic cell. Middle panel shows amplitude histograms constructed from 2-min time periods including the above traces. The mean amplitude + S.D. in control was 60.4 + 40.3 pA, while in 1 ~uM TTX it was 33.4 + 16.1 pA. Histograms in lower panel represent exponential frequency distributions of the intervals between two consecutive slPSCs recorded from the identical 2-min time periods. The mean inter-event interval in control was T = 74.1 ms, for a total number of intervals N = 1616, while in the presence of 1/tM TTX the mean value was T = 130.5 ms for N = 916. In both control and TTX histograms, binwidth (BW) is 10 ms and the fitted exponential probability density function (solid line) has the form: (N x BW/T) x exp(-t/T).
144 of the cerebellum (n = 2) indicate the presence of spontaneous inhibitory currents mediated by G A B A A receptors through the opening of CI- channels. Most recordings were done in conditions of symmetrical C1concentrations (140-145 mM) on either side of the neuronal membrane. Thus, the reversal potential of currents through G A B A A ionophores was around 0 mV and the spontaneous inhibitory postsynaptic currents (slPSCs) were uniformly inward at holding potentials close to the normal resting membrane potential of neurons (-60 to -80 mV). In several experiments, decreasing the intracellular concentration of CI- shifted the reversal potential of the currents to more negative values.
Effect of tetrodotoxin Previous studies have reported the persistence of spontaneous synaptic currents in the absence of action potentials in cultured mammalian neurons ~4 and have alluded to TFX-resistant spontaneous synaptic currents in hippocampal slices ~1'18. However, a quantification of the T F X effect is not available for spontaneous synaptic events in the mammalian CNS. We have investigated the effect of bath-applied TTX up to concentrations of 5/~M and found that 1 /~M was sufficient to block all action potential activity in every cell examined in our preparation. As shown in Fig. 1, T I ' X (1 /tM) reduced the amplitude of the spontaneous synaptic currents and prolonged the inter-event intervals by 76% (i.e. reduced their frequency by about 43%). With a CI driving force o f - 7 0 to -80 mV, large (>200 pA) spontaneous currents were abolished by TTX. When individual events were averaged, no differences were found in the rise times or decay time constants of T-FX-sensitive or qTX-insensitive spontaneous activity. As indicated by the exponential probability density functions fitted to the distributions of the inter-event intervals 47, both "l'TX-sensitive and TTXresistant activity appeared to be randomly generated. In cells with very stable holding currents (e.g. Fig. 1) it was also possible to detect a change in the mean holding current following perfusion of TTX due to the reduction in mean slPSC amplitude and diminished frequency. In the example shown in Fig. 1, a net mean outward current of 30 pA was discernible following the blockade of the larger spontaneous inward currents by q T X corresponding to a net conductance decrease of 0.54 nS. Contribution of spontaneous IPSCs to the resting membrane potential and conductance In order to determine the effective contribution of the spontaneous q~X-resistant synaptic currents to the resting membrane potential we have carried out experiments in current clamp recordings while setting E a at various
potentials. Fig. 2 illustrates such an experiment, whereby virtue of the C1- concentration of the intracellular solution, Eel was set to -55 mV or to 0 mV. Depending on neuronal input resistance (120-280 MK2), the spontaneous IPSPs ranged from 0.05 to 1.5 mV at a membrane potential -15 mV away from E o and from 1 to 15 mV at about 70-80 mV away from E o . It is thus clear that depending on the in situ E o in a given cell and the resting membrane potential, the TFX-resistant slPSCs will be expressed as membrane potential oscillations of various amplitudes. Perhaps even more important is the steady conductance mechanism which persists in the presence of such continuous synaptic activity. Fig. 3 shows the effect of perfusing the competitive G A B A A receptor antagonist bicuculline methiodide (BMI, 5/tM). The net outward current produced after the blockade of the spontaneous events was 60 pA which at the holding potential o f - 6 0
Ecj = - 5 5 m V
Ecj -- 0 m V
R M P = -73 m V
R M P = -72 m V
~,,~J
"~r~v~ '
I~,
/
I
Ii ~1 I L
i
__i I 10ms
mV
Fig. 2. Whole-cell current clamp recordings of 1q'X-resistant spontaneous postsynaptic potentials in two dentate gyrus granule cells with different CI- reversal potentials. Left panel: the intracellular fill contained 135 mM K-gluconate and 5 mM KCI in addition to 2 mM MgCI 2 and 10 m M H E P E S (ph = 7.2). The slPSPs reversed at the E o o f - 5 5 mV. This E o is very close to the reversal potential of evoked IPSPs recorded in these neurons with sharp microelectrodes without dialysis of the cell's interior. In each of the 308 ms long traces the average membrane potential is 0.2 to 0.5 mV more depolarized than the first 10 ms of baseline. Right panel: in another granule cell E o was set to 0 mV by filling the patch electrode with 140 mM KC1, 2 mM MgCI 2 and 10 m M H E P E S (pH 7.2). The averages of the membrane potential traces shown are 2.5 to 5.2 mV more depolarized than the event-free first 10 ms of the recordings. Note the marked baseline noise in this recording compared to the recording on the left. This is probably due to greater basal CI- channel activity at a potential further away from ECl-
145 m V corresponds to a conductance decrease of 1 nS. In general, depending on the frequency of occurrence, the
frequency of TTX-resistant slPSCs. The n e u r o n in Fig. 3 had one of the shortest m e a n inter-event intervals (67.6
net change in conductance following blockade of the slPSCs was 3 0 - 5 0 % of the peak conductance calculated from the amplitude of the averaged events. Submicromolar (0.5-0.8 /~M) concentrations of BMI were also effective in abolishing the slPSCs (n = 3). It is noteworthy that such low concentrations of antagonist have less than significant effects on the amplitude of evoked IPSPs 2"9. We have also used picrotoxin (10-50/~M) to block the slPSCs to demonstrate the C1- channel dependence of the spontaneous events (n = 3).
ms corresponding to an average frequency of 14.8 Hz). Nevertheless, as indicated by the m o n o e x p o n e n t i a l nature of the probability density function of the inter-event intervals (Fig. 3B) and the excellent fit to a Poisson distribution of the n u m b e r of events per successive 120 ms intervals (Fig. 3C), these rapidly occurring slPSCs appear to have a r a n d o m occurrence 47.
Decay time constants and rise times Average slPSCs were obtained by collecting between 400 and 700 single events during a period of 2-5 min. A t
Figs. 1 and 3 also reveal the large variability in the
A
-
5 pM BICUCULLINE
b
t a
a
b
Is0 40./50m=
i 8
C 1111 10e
--
~
0
e.MVen-67.6m:
• 100
200 ~ 400 InlervRI(me)
Inter-Event
! i
500
0 1 2 I 4 II e ? n No,ofEventsInConeecutlve120 me IIIne
Fig. 3. A: perfusion of 5/zM bicuculline blocks TTX-resistant sIPSCs. Uppermost record shows the transmembrane current monitored during wash in of 5/~M bicuculline methiodide. The holding potential was -55 mV, and the patch pipette contained 140 mM CsC1, 10 HEPES, and 2 MgCI2. All extracellular solutions contained 1/tM TTX. An expanded sweep taken just before the switch from control solution (marked by the asterisk and 'a') depicts discrete inward currents, which are completely abolished upon perfusion of 5/~M bicuculline ('b'). B: exponential frequency distribution of the form described in Fig 1. Data collected from the same cell as above in a 2-min period just prior to point a. Binwidth was 5 ms and the mean inter-event interval was T --- 67.6 ms (total number of events = 873). The fitted exponential (solid line) has the form: (N x BW/T) x exp(-t/T). C: frequency distribution of slPSCs in 400 consecutive 120-ms time bins taken from the same period as in B. The continuous curve is a Poisson distribution of the form: 400 x (rex/X!) :,< exp(-m) where X is the number of slPSCs per bin and m is the mean number of slPSCs per consecutive 120-ms time bins (rn = 1.74).
146
A
a
200. •-
-''-"
• :/""
=1
l • • "
.'z~',.'l-.
r - 4x£-04
t" • ~
•
.
ms
""
i.~;~T:~.~.~.s. oi
o
".-~
.
lb
~
1"5
/0
(ms) C
D
125 I00 o (o eq
v
E
4
r - I~E-02
Ip
E
75
i~
2
:....
,
.
.
-...:._ . - . .
;
.
ID
o
•
o
" . " , . ~ M m k ~ . ' , , r.'-
25, 0 0.0
o.,,
1.o 10-90~
1.5
2.0
£5
o
~,
1"o
Rise Time ( m s )
" 1"5
2"o
7 (ms)
Fig. 4. Spontaneous inhibitory postsynaptic currents (sIPSCs) recorded in a dentate gyrus granule cell in the presence of 1/~M TTX at a holding potential of -90 mV. A: 3 individual slPSCs are depicted from 431 such events recorded over a period of 6 min. The 3 synaptic currents have different amplitudes (112 pA, 60 pA and 24 pA) but very similar decay time constants: 5.18 ms, 5.00 ms and 4.96 ms respectively. The mean amplitude of the slPSCs in this cell was 67.5 + 1.7 pA and the average of the decay time constants was 5.58 + 0.1 ms (+ S.E.M.; n = 431) B: graph showing the lack of correlation between decay time constant and amplitude of slPSCs (correlation coefficient = 4 x 10-4). C: histogram of the rise times defined as the time required for the slPSC to grow from 10% to 90% of its peak amplitude. Mean rise time was 0.75 + 0.02 ms (+ S.E.M.; n = 431). D: the decay time constant was not correlated with the rise time of the slPSCs (correlation coefficient = 8 x 10-2).
-55 to -90 mV holding potentials, the size of the average slPSCs ranged between 30-90 pA (e.g. Figs. 1, 4, 5 and 6). In individual neurons there was a large variability in the decay time constants of single slPSCs (e.g. Fig. 4B). Nevertheless, in most cells the average slPSC decayed monoexponentially with a time constant of 3.7-7.2 ms (e.g. Fig. 4). Only in 3 out of 92 neurons the average of 100-400 slPSCs exhibited a double exponential decay with a fast time constant of <1.5 ms (0.85-1.42 ms) and a second slower time constant of 4.5-6.6 ms (e.g. Fig.
A
~ _ --
Y
1 ,°p,'
5A).
Depending on the holding potential, the amplitude of individual TTX-resistant slPSCs ranged between 15-200 pA (e.g. Fig. 4B). Large or small amplitude events had
a
15 I--1
Control
50/a~ Penb~rb. u)
o oo
Fig. 5. A: the average slPSC recorded at a holding potential of -80 mV in a granule cell in the presence of 1/~M T/'X is significantly prolonged following perfusion of 50 pM Na-pentobarbital. The decay phase of the average spontaneous current could be best fitted by a double exponential with the time constants: 1.42 and 6.11 ms. In the presence of the drug both decay time constants more than doubled to 2.92 and 13.46 ms respectively. The amplitude of the average slPSC (39 pA) was not altered by pentobarbital. B: the drug had no significant effect on the frequency distribution of slPSC amplitudes.
10
bJ
0
-100
-i~
-~
-4o
slPSC A m p l i t u d e (pA)
-2o
147
A +35 mV
. . . . . . . . . . .
-90mV ~ a
. . . . . . . . . . . .
.
.
.
.
.
.
~-~
=
.
.
.
.
.
.
.
5
msI10pA
C 17.7 • 3.5
-18.3 i 5.3
m
+
m
-32.6 :t 10.5
to
ID
/
10
:1oo
-io
40
-4o
-=o
0
o
slPSC A m p l i t u d e ( 1 ~ )
slPS.C Amplitude (pA)
lOO
Fig. 6. A: averaged sIPSCs recorded in a dentate gyrus granule cell with a 140 mM CsCI containing patch electrode at two different holding potentials (-90 mV and +35 mV). Note the rectification of the amplitude as well as the decay time constant at the depolarized holding potential. B,C: amplitude distributions of sIPSCs at the two different holding potentials (-90 mV in B. and +35 mV in C.). Both histograms were best fitted by the sum of two normal (Gaussian) distributions with the indicated means and standard deviations. Note that the mean of the smallest unitary events in B. 18.3 pA, which, considering the sIPSC reversal potential of 0 mV and a single GABAg channel current of 1.8 pA, corresponds to the opening of approximately 10 GABAA channels.
similar decay time constants (Fig. 4A,B), indicating adequate control of the membrane voltage 26'5°. The rise-times (measured between 10 and 90% of peak amplitude) of the sIPSCs were <1 ms (Fig. 1C) and did not correlate with the decay time constant (Fig. 4D). In each cell tested (n = 5) the TTX-resistant sIPSCs decayed in a similar fashion as the large spontaneous IPSCs recorded in the absence of TTX. When normalized to the same amplitude, the averages of the large amplitude TTX-sensitive events were essentially superimposable over the averages of the smaller amplitude TFX-resistant sIPSCs. Anesthetic drugs known to enhance the burst time of G A B A A receptor/channels 32'34 had a profound effect on the decay time constant of sIPSCs. Fig. 5 depicts the prolongation of an averaged response by perfusion of 50 p M Na-pentobarbital. The average prolongation of the decay time constant caused by 50 g M pentobarbital was 266 _+ 23% of control (mean + S.E.M.; n = 4). Similar results were obtained with the volatile anesthetics halothane and isoflurane applied at their clinically relevant concentrations 37.
Rectification of slPSC amplitudes and decay time constants G A B A A receptor channels are known for their outward rectification 22'49. This rectification manifests itself in both a non-linear voltage-dependence of the G A B A responses and an increase in the decay time constant of the IPSCs at depolarized potentials lt'13'43. The averaged TTX-resistant sIPSCs had similar rectifying properties. Fig. 6 displays the increase in amplitude and decay time constant of the averaged sIPSCs at depolarized holding potentials. This property of the sIPSCs would make the inhibitory events more effective when neurons are depolarized from their resting m e m b r a n e potentials. DISCUSSION Our results indicate that spontaneous synaptic currents persist when action potential activity is blocked by perfusion of TFX. Only the relatively large events (>200 pA) were abolished by TTX. These large spontaneous currents were probably similar to IPSCs previously described in studies using sharp microelectrode recordings H. In the presence of T F X , the miniature
148 slPSCs had several characteristics of a G A B A A receptormediated CI conductance: (1) they were sensitive to intracellular CI- concentrations and reversed around the Ec~; (2) they displayed an outward rectification in amplitudes and a voltage-dependence of the decay time constant (Fig. 6A) characteristic of both G A B A A channels 22'49 and GABAA-mediated IPSCsl1'13'43; (3) pentobarbital (50 ~tM), which prolongs IPSPs 38 by prolonging the bursting of G A B A A channels 32"34, significantly lengthened the decay time constants of slPSCs (Fig. 5A); (4) the competitive GABAA receptor antagonist bicuculline methiodide (BMI) 15 or the Ci- ionophore blocker picrotoxin completely abolished the slPSCs. The large cell-to-cell variability in the frequency of the slPSCs is probably due to the unequal number of GABAergic synapses on the recorded neurons. A neuron with a large number of inhibitory synapses would be expected to generate slPSCs more frequently. Although it would be tempting to make quantal predictions from the size of the TTX-resistant slPSCs, at the moment such analysis seems inappropriate. For example, multiple Gaussian distributions with different means could have been fitted to the amplitude histograms of Fig. 1, as it has been done for the events in Fig. 5. However, the tedious and often inaccurate fit of amplitude histograms to several normal curves is not by itself proof of the quantal nature of spontaneous events. Cumulative probability distributions have to be made in order to establish the presence of distinct subunits in a continuous distribution of amplitudes 47'4s. We have done such log-normal probability plots of slPSC amplitudes and shown a clearly concave shape of the curve in each cell examined. This indicates the presence of subunits with different mean amplitudes and variances 47'48, but we could not accurately separate and resolve the average size and variance of distinct event subpopulations. The most likely explanation for the encountered difficulty is, akin to the marked variability in slPSC frequencies, the large number of heterogeneous GABAergic synapses innervating each neuron. Further studies using low intensity direct stimulation of GABAergic interneurons are under way to unequivocally resolve the possible quantal nature of slPSCs. The exponential frequency distribution of the interevent intervals and the good Poisson fits to the number of events per consecutive non-overlapping time bins shown in Figs. 1 and 3B,C, indicate that the TTXresistant slPSCs occur randomly. Yet, the released G A B A responsible for the activation of the spontaneous currents did not conform to the classical release mechanism of neurotransmitter substances27: the miniature slPSCs were not dependent upon neuronal action poten-
tial firing and persisted in the presence of the Ca 2+ channel antagonist Cd 2÷ (100-200/~M, n = 6, not shown; and ref. 11). Thus, the release of G A B A responsible for activation of the slPSCs in our preparation resembles the TTX-insensitive Cd2÷-resistant basal G A B A release measured in vivo 51. It is conceivable that ongoing excitatory activity could have spontaneously depolarized inhibitory interneurons to release G A B A intermittently. Indeed, a recent study 31 demonstrated discrete TTX-resistant quisqualate receptor-mediated spontaneous excitatory postsynaptic currents (sEPSCs) in neocortical neurons. No such sEPSCs were noted in the hippocampal CA1 pyramidal cells or granule cells of the dentate gyrus (e.g. Fig. 3A). Nevertheless, these principal cells may spontaneously release an excitatory transmitter to act upon the inhibitory interneurons. We tested this hypothesis by blocking spontaneous excitatory neuronal activity with excitatory amino acid antagonists. Perfusion of 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX, 5 ~tM; n = 3) or o-2amino-5-phosphonovaleric acid (D-AP5, 20/~M; n = 4) had no significant effect on the frequency or average amplitude of the slPSCs (not shown). In light of these findings, it seems unlikely that an excitatory drive controls the spontaneous G A B A release mechanism from inhibitory interneurons. The amplitude and the peak conductance range of the miniature slPSPs indicates that just a few G A B A A channels contribute to a single spontaneous inhibitory current. Bearing in mind that in these neurons the main conductance state of a single G A B A A channel is between 25-30 pS 5'6'18'22, each channel would generate unitary currents of 1.7-2.7 pA at holding potentials o f - 7 0 to -90 mV. Hence the simultaneous opening of 10-20 GABAA channels is sufficient to cause spontaneous unitary events in the range of 20-30 pA. This number of channels is significantly smaller than has been estimated for the inhibitory synapses of Mauthner cells 19, and is an order of magnitude less than previous estimates in hippocampal neurons 11. From the decay time constants of slPSCs, further speculations may be made about the G A B A concentrations necessary to produce such spontaneous currents. We found that the decay of the slPSCs could be described by a single exponential in most cells. Based on the effect of agonist concentration on the kinetics of GABA-activated channels, the monoliganded G A B A A receptor has a less stable and shorter open state than the receptor occupied by two G A B A molecules 35. It has been demonstrated that G A B A concentrations of at least 5/~M are necessary to favor the biliganded receptor state which shows longer open times 35. It appears that such G A B A concentrations must be reached close to the release site in the synaptic cleft to produce the miniature
149 sIPSCs observed in our study. In summary, we have described a perpetual, background G A B A e r g i c inhibition in the m a m m a l i a n CNS analogous to the tonic N M D A receptor activity31'42 and the discrete activation of A M P A receptors recently reported 31. Previously, this tonic inhibition has, for the most part, r e m a i n e d undetected due to the inability to employ high resolution patch clamp recording techniques in brain slices. Such inhibitory activity could critically contribute to controlling the general state of excitability of m a m m a l i a n CNS neurons. For example, submicromolar doses of bicuculline cause the horizontal spread of epileptiform activity in the cerebral cortex 9. Such doses of G A B A A antagonists barely alter evoked inhibitory postsynaptic potentials 2'9 but completely abolish slPSCs. Thus, the transmission of excitation in neocortical aggregates may d e p e n d on the ongoing spontaneous inhibition, and the absence of slPSCs (as in submicromolar concentrations of bicuculline) may permit the generation and
REFERENCES 1 Alger, B.E. and Nicoll, R.A., Spontaneous inhibitory postsynaptic potentials in hippocampus: mechanism for tonic inhibition, Brain Research, 200 (1980) 195-200. 2 Alger, B.E. and Nicoll, R.A., Pharmacological evidence for two kinds of GABA receptor on rat hippocampal pyramidal cells studied in vitro, J. Physiol., 328 (1982) 125-141. 3 Blankenship, J.E. and Kuno, M., Analysis of spontaneous subthreshold activity in spinal motoneurons of the cat, J. Neurophysiol., 31 (1968) 195-209. 4 Blanton, M.G., LoTurco, J.J. and Kriegstein, A.R., Whole cell recording from neurons in slices of reptilian and mammalian cerebral cortex, J. Neurosci. Methods, 30 (1989) 203-210. 5 Bormann, J., Sakmann, B. and Seifert, W., Isolation of GABA-activated single-channel CI- currents in the soma membrane of rat hippocampal neurones, J. Physiol., 341 (1983) 9-10P. 6 Bormann, J., Electrophysiology of GABAA and GABA B receptor subtypes, Trends Neurosci., 11 (1988) 112-116. 7 Bowery, N.G. (Ed.), Actions and Interactions of GABA and Benzodiazepines, Raven Press, New York, 1983. 8 Brock, L.G., Coombs, J.S. and Eccles, J.C., The recording of potentials from motoneurones with an intracellular electrode, J. Physiol., 117 (1952) 431-460. 9 Chagnac-Amitai, Y. and Connors, B.W., Horizontal spread of synchronized activity in neocortex and its control by GABAmediated inhibition, J. Neurophsyiol., 61 (1989) 747-758. 10 Chen, Q.X., Stelzer, A., Kay, A.R. and Wong, R.K.S., GABAA receptor function is regulated by phosphorylation in acutely dissociated guinea-pig hippocampal neurones, J. Physiol., 420 (1990) 207-221. 11 Collingridge, G.L., Gage, P.W. and Robertson, B., Inhibitory post-synaptic currents in rat hippocampal CA1 neurones, J. Physiol., 356 (1984) 551-564. 12 Colomo, F. and Erulkar, S.D., Miniature synaptic potentials at frog spinal neurones in the presence of tetrodotoxin, J. Physiol., 199 (1968) 205-221. 13 Cull-Candy, S.G., Inhibitory synaptic currents in voltageclamped locust muscle fibres desensitized to their excitatory transmitter, Proc. R.Soc. Lond. Ser. B, 221 (1984) 375-383. 14 Cull-Candy, S.G. and Usowicz, M.M., Whole-cell current noise produced by excitatory and inhibitory amino acids in large
spreading of a b e r r a n t activity. It is also plausible that most drugs such as anesthetics, antiepileptics, barbiturates, or benzodiazepines, k n o w n to alter G A B A e r g i c neurotransmission, may exert their primary action on the perpetual spontaneous inhibitory activity. O u r study also opens new possibilities for studying the intracellular regulation of G A B A channels 1°'2°'23'25'44'45 and drug or neurotransmitter interactions with inhibitory mechanisms in the CNS. Future research will be aimed at determining if alterations of such 'silent' inhibitory activity take place following chronic changes in n e u r o n a l excitability7'29' 39-41,46
Acknowledgements. This work was supported in part by NIH Grants NS-12151, RR05353-27 and a Fellowship from the Klingenstein Fund to I.M.T.S.O. is a Howard Hughes Predoctoral Fellow and K.J.S. is a Dana Postdoctoral Fellow. We would like to thank Dr. J. Dempster for providing the Strathclyde Electrophysiology Software and J. T. Palmer for expert technical assistance.
cerebellar neurones of the rat, J. Physiol., 415 (1989) 533-553. 15 Curtis, D.R., Duggan, A.W., Felix, D. and Johnston, G.A.R., GABA, bicucuiline and central inhibition, Nature, 226 (1970) 1222-1224. 16 Eccles, J.C., The mechanism of synaptic transmission, Ergebn. Physiol., 51 (1961) 299-430. 17 Edwards, EA. and Gage, P.W., Seasonal changes in inhibitory currents in rat hippocampus, Neurosci. Lett., 84 (1988) 266-270. 18 Edwards, F.A., Konnerth, A., Sakmann, B. and Takahashi, T., A thin slice preparation for patch clamp recordings from neurones of the mammalian central nervous system, Pflagers Arch., 414 (1989) 600-612. 19 Faber, D.S. and Korn, H., Transmission at a central inhibitory synapse. I: Magnitude of unitary postsynaptic conductance change and kinetics of channel activation, J. Neurophysiol., 48 (1982) 654-677. 20 Feltz, A., Demeneix, B., Feltz, A., Taleb, O., Trouslard, J., Bossu, J.-L. and Dupont, J.-L., Intracellular effectors and modulators of GABA-A and GABA-B receptors: a commentary, Biochimie, 69 (1987) 395-406. 21 Gage, P.W. and Robertson, B., Prolongation of inhibitory postsynaptic currents by pentobarbitone, haiothane and ketamine in CA1 pyramidal cells in rat hippocampus, Br. J. Pharmacol., 85 (1985) 675-681. 22 Gray, R. and Johnston, D., Rectification of single GABA-gated chloride channels in adult hippocampal neurons, J. Neurophysiol., 54 (1985) 134-141. 23 Gyenes, M., Farrant, M. and Farb, D.H., 'Run-down' of GABA A receptor function during whole-cell recording: a possible role for phosphorylation, Mol. Pharmacol., 34 (1988) 719-723. 24 Hubbard, J.I., Stenhouse, D. and Eccles, R.M., Origin of synaptic noise, Science, 157 (1967) 33'0-331. 25 Inoue, M., Oomura, Y., Yakushiji, T. and Akaike, N., Intracellular calcium ions decrease the affinity of the GABA receptor, Nature, 324 (1986) 156-158. 26 Johnston, D. and Brown, T.H., Interpretation of voltage-clamp measurements in hippocampal neurons, J. Neurophysiol., 50 (1983) 464-486. 27 Katz, B., The Release of Neurotransmitter Substances, Liverpool University Press, Liverpool, 1969. 28 King, G.L., Knox, J.J. and Dingledine, R., Reduction of inhibition by a benzodiazepine antagonist, Ro15-1788, in the rat
150 hippocampal slice, Neuroscience, 15 (1985) 371-378. 29 Krogsgaard-Larsen, P., Scheel-Kruger, J. and Kofod, H. (Eds.), GABA-Neurotransrnitters, Munkgaard Press, Copenhagen, 1979. 30 Kuno, M., Quantum aspects of central and ganglionic synaptic transmission in vertebrates, Physiol. Rev., 51 (1971) 647-678. 31 LoTurco, J.J., Mody, I. and Kriegstein, A.R., Differential activation of glutamate receptors by spontaneously released transmitter in slices of neocortex, Neurosci. Lett., 114 (1990) 265-271. 32 Macdonald, R.L., Twyman, R.E., Rogers, C.J. and Weddle, M.G., Pentobarbital regulation of the kinetic properties of GABA receptor chloride channels, Adv. Biochem. Psychopharmacol., 45 (1988) 61-71. 33 Marty, A. and Neher, E., Tight-seal whole-cell recording. In B. Sakmann and E. Neher (Eds.), Single-Channel Recording, Plenum Press, New York, 1983, pp. 107-122. 34 Mathers, D.A., The GABA A receptor: new insights from single-channel recording, Synapse, 1 (1987) 96-101. 35 Mathers, D.A. and Wang, Y., Effect of agonist concentration on the lifetime of GABA-activated membrane channels in spinal cord neurons, Synapse, 2 (1988) 627-632. 36 Miles, R. and Wong, R.K.S., Unitary inhibitory synaptic potentials in the guinea-pig hippocampus in vitro, J. Physiol., 356 (1984) 97-113. 37 Mody, I., Tanelian, D.L. and Maclver, M.B., Halothane enhances tonic neuronal inhibition by elevating intracellular calcium, Brain Research, in press. 38 Nicoll, R.A., Eccles, J.C., Oshima, T. and Rubia, R., Prolongation of hippocampal inhibitory postsynaptic potentials by barbiturates, Nature, 258 (1975) 625-627. 39 Prince, D.A. and Connors, B.W. Mechanism of interictal epileptogenesis. In A.V. Delgado-Escueta, A.A. Ward, D.M. Woodbury and R.J. Porter (Eds.), Basic Mechanisms of epilepsies; Molecular and Cellular Approaches, (Advances in Neurology, Vol. 44), Raven Press, New York, 1986, pp. 275-299. 40 Roberts, E. Failure of GABAergic inhibition: a key to local and
41 42
43 44
45
46 47
48
49 50
51
global seizures. In A.V. Delgado-Escueta, A.A. Ward, D.M. Woodbury and R.J. Porter (Eds.), Basic Mechanisms of epilepsies; Molecular and Cellular Approaches, (Advances in Neurology, Vol. 44), Raven Press, New York, 1986, pp. 319-341. Roberts, E., Chase, T.N. and Tower, D.B. (Eds.), GABA in Nervous System Function, Raven Press, New York, 1976. Sah, P., Hestrin, S. and NicoU, R.A., Tonic activation of NMDA receptors by ambient glutamate enhances excitability of neurons, Science, 246 (1989) 815-818. Segal, M. and Barker, J.L., Rat hippocampal neurons in culture: voltage-clamp analysis of inhibitory synaptic connections, J. Neurophysiol., 52 (1984) 469-487. Sigel, E. and Baur, R., Activation of protein kinase C differentially modulates neuronal Na +, Ca 2+ and ~,-aminobutyrate type A channels, Proc. Natl. Acad. Sci. U.S.A., 85 (1988) 6192-6196. Stelzer, A., Kay, A.R. and Wong, R.K.S., GABAA-receptor function in hippocampal cells is maintained by phosphorylation factors, Science, 241 (1988) 339-341. Tallman, J.F., Paul, S.M., Skolnick, P. and Gallager, D.W., Receptors for the age of anxiety: pharmacology of the benzodiazepines, Science, 207 (1980) 274-281. Van der Kloot, W., Kita, H. and Cohen, I., The timing of miniature end plate potentials, Progr. Neurobiol., 4 (1975) 269-326. Van der Kloot, W., Statistical and graphical methods for testing the hypothesis that quanta are made up of subunits, J. Neurosci. Methods, 27 (1989) 81-89. Weiss, D.S., Membrane potential modulates the activation of GABA-gated channels, J. Neurophysiol., 59 (1988) 514-527. Weiss, D.S., Barnes, Jr., E.M. and Hablitz, J.J., Whole-cell and single channel recordings of GABA-gated currents in cultured chick cerebral neurons, J. Neurophysiol., 59 (1988) 495-513. Westerink, B.H.C. and De Vries, J.B., On the origin of extracellular GABA collected by brain microdialysis and assayed by a simplified on-line method, Naunyn-Schmiedeberg's Arch. Pharmacol., 339 (1989) 603-607.
142
BRES 16467
Perpetual inhibitory activity in mammalian brain slices generated by spontaneous GABA release Thomas S. Otis, Kevin J. Staley and Istvan Mody Department of Neurology and Neurological Sciences M016, Stanford University School of Medicine, Stanford, CA 94305 (U.S.A.) (Accepted 23 October 1990)
Key words: Brain slice; Dentate gyrus; ~,-Aminobutyric acid; Hippocampus; Neurotransmitter release; Spontaneous inhibitory postsynaptic current; Synaptic transmission; Whole-cell patch clamp
Miniature spontaneous inhibitory postsynaptic currents (slPSCs) mediated by GABAA receptors were recorded using whole-cell patch clamp recordings in rat brain slices maintained in vitro at 34 _+ 1 °C. We have found that firing of action potentials by principal neurons or by GABAergic interneurons is not necessary to the generation of sIPSCs since they persist in the presence of 1-5/tM tetrodotoxin (TrX). The average frequency of the discrete sIPSCs exhibits a large cell-to-cell variability and is between 5-15 Hz. The amplitudes of the sIPSCs depend on the difference between the membrane potential and the equilibrium potential for Cl- (Eo). Generally, 70-80 mV away from E a , sIPSCs have a mean amplitude of 30-80 pA (i.e. peak conductance of 400-1000 pS) with an average decay time constant of 5.8 ms. Accordingly, unitary single sIPSCs arise from the simultaneous activation of no more than 20 GABA A receptor/channels. The perpetual barrage of spontaneous GABAergic activity is very likely to be a critical factor in the regulation of neuronal excitability and the mechanism of action of several neuroactive compounds. INTRODUCTION N o r m a l functioning of the central nervous system (CNS) requires the existence of a delicate balance between excitation and inhibition. The principal inhibitory n e u r o t r a n s m i t t e r of the m a m m a l i a n CNS is ~aminobutyric acid ( G A B A ) acting via G A B A A and G A B A B receptors 7'15'29m. Relatively small changes in G A B A e r g i c neurotransmission can have striking effects on the functional state of nerve cells, resulting in p r o f o u n d alterations of neuronal excitability39'4°. In fact, m a n y c o m p o u n d s used clinically in the treatment of nervous system disorders alter the activity of G A B A A receptor/channels 7,15,29,39-41,46. D e c a d e s ago, the term synaptic noise was introduced to describe the steady b o m b a r d m e n t of m a m m a l i a n nerve cells by spontaneous excitatory or inhibitory miniature synaptic potentials. A t the time, the origin of synaptic noise was attributed to the release of neurotransmitter substances through continual firing of nerve cells8'16'24. It was later shown that b l o c k a d e of action potential firing did not prevent the occurrence of miniature potentials at certain synapses leading to a re-examination of neurotransmitter release mechanisms in the m a m m a l i a n CNS 3'
Previous studies using sharp m i c r o e l e c t r o d e s to record intracellularly in h i p p o c a m p a l slices have described spontaneous inhibitory postsynaptic potentials 1'28'36 or currents 11"17"21 in chloride-loaded neurons. These studies have concluded that most, if not all, spontaneous inhibitory activity stems from n e u r o n a l action potential firing 1A1'28'36. Once action potentials are abolished, e.g. by the N a ÷ channel blocker t e t r o d o t o x i n (T-FX), spontaneous inhibitory activity b e c o m e s infrequent and of very small amplitude 11. O n e o f the inherent drawbacks while recording with sharp m i c r o e l e c t r o d e s is the presence of a significant leak conductance due to perforation of the neuronal m e m b r a n e . By introducing a relatively large shunt, this conductance m a y r e n d e r small a m p l i t u d e changes in m e m b r a n e current o r voltage undetectable. Tight seal (>109 ~2) patch recordings provide an extremely small a m o u n t of m e m b r a n e leak c o m p a r e d to conventional sharp m i c r o e l e c t r o d e recordings 33. We have a d a p t e d a whole cell patch recording m e t h o d 4'18 to record from brain slices m a i n t a i n e d in a conventional recording c h a m b e r at 34 + 1 °C. The resolution of this m e t h o d has enabled us to investigate previously u n d e t e c t a b l e small amplitude miniature s p o n t a n e o u s inhibitory postsynaptic currents (slPSCs).
12,30
Correspondence: I. Mody, Department of Neurology and Neurological Sciences M016, Stanford University School of Medicine, Stanford, CA 94305, U.S.A.
143 record slPSCs, the following intracellular solutions were used (in mM): (1) KCI 140, MgCI 2 2, HEPES 10; (2) CsCI 140, "MgCI2 2, HEPES 10; (3) K-gluconate 130, KCI 10, MgCI 2 2, HEPES 10. Some intracellular solutions also contained 11 mM EGTA or BAPTA and 1 mM CaC12 which had no noticeable effect on the slPSCs. All intracellular media were adjusted to a pH of 7.2 with KOH or CsOH and had an osmolarity of 250--270 mOsm. Recordings were done using an Axoclamp-2A in continuous SEVC mode or an Axopatch-lD amplifier with >80% series resistance compensation. Data were low pass filtered at 2-3 kHz, digitized at 6-12.5 kHz and analyzed using the Strathclyde Electrophysiology Software (courtesy of J. Dempster).
MATERIALS AND METHODS
Hippocampal slices (400-450/~m) or coronal brain sections (400 ~m) containing the hippocampus were prepared at 2-4 °C from adult (200-350 g) male Wistar rats on a Mcllwain tissue chopper or a Lancer Series 1000 Vibratome respectively. The slices were incubated in a storage chamber at 32 °C for at least 1 h before recordings were started. The recordings were done at 34 + 1 °C in an interface type chamber using a nylon net to cover the slices. The net disrupted the surface tension providing a thin ( < 100/~m) film of fluid over the top of the brain tissue. Oxygenated (95%02/ 5%CO2) artificial cerebrospinal fluid (ACSF) was perfused at a rate of 2.5-3 mi/min. The composition of the ACSF was (in mM): NaCI 126, NaHCO 3 26, NaH2PO 4 1.25, KCI 2.5, CaCI 2 2, MgCI 2 2, o-glucose 10 (303 mOsm). In most cases the ACSF also contained 1-5 ~ M tetrodotoxin (TTX, Calbiochem). Patch-electrodes (2-2.5 /~m tip diameter; 3-5 M.Q) were pulled in two stages from borosilicate thin-walled capillaries (o.d. 1.5 mm; TWl50F-4; WPI) on a Narashige PP-83 electrode puller. In various experiments, the electrodes were filled with different intracellular solutions. To
RESULTS O u r r e c o r d i n g s f r o m d e n t a t e g y r u s g r a n u l e cells ( n = 67), p y r a m i d a l cells o f t h e h i p p o c a m p a l C A 1 r e g i o n (n = 15) o r s o m a t o s e n s o r y c o r t e x ( n = 8), a n d P u r k i n j e cells
Control
1/~M TTX
2S
M e a n A m p l i t u d e - 33.4 * 1 6.1 pA
M e a n A m p l i t u d e - 6 0 . 4 + 40.3 pA a o
~
s
4
0
0
60 •
SM
150
200
0
250
50
•
100 150 Peak Amplitude (pA)
Peak AnNplltuda(pA)
lo I
200
250
10
M e a n - 74.1 ms
Mean - 130.5 m s
'HIII
i--
ae
IlO0 Inter-Event interval I m s )
8oo
ii
o •
200
400
Inler-Event Interval Imsl
Fig. 1. Blocking action potential-dependent neuronal activity with TTX in a dentate gyrus granule cell reduces average slPSC amplitude and frequency, but does not abolish slPSCs. Upper records show inward slPSCs recorded at -55 mV before (left trace) and 10 min after (right trace) perfusion of 1/tM TTX. Note the 30-pA reduction in the mean baseline inward current, presumably resulting from TTX blockade of large synaptic currents and the diminished frequency of TTX-resistant slPSCs, lntracellular solution contained: 130 mM CsCI, 11 mM BAPTA, 10 mM HEPES, 2 mM MgCI2 and 1.1 mM CaC! 2 at a pH of 7.2. The local anesthetic derivative QX314 (10 mM; Astra Pharmaceuticals) was also included in the patch pipette to block breakthrough Na ÷ spikes in the postsynaptic cell. Middle panel shows amplitude histograms constructed from 2-min time periods including the above traces. The mean amplitude + S.D. in control was 60.4 + 40.3 pA, while in 1 ~uM TTX it was 33.4 + 16.1 pA. Histograms in lower panel represent exponential frequency distributions of the intervals between two consecutive slPSCs recorded from the identical 2-min time periods. The mean inter-event interval in control was T = 74.1 ms, for a total number of intervals N = 1616, while in the presence of 1/tM TTX the mean value was T = 130.5 ms for N = 916. In both control and TTX histograms, binwidth (BW) is 10 ms and the fitted exponential probability density function (solid line) has the form: (N x BW/T) x exp(-t/T).
144 of the cerebellum (n = 2) indicate the presence of spontaneous inhibitory currents mediated by G A B A A receptors through the opening of CI- channels. Most recordings were done in conditions of symmetrical C1concentrations (140-145 mM) on either side of the neuronal membrane. Thus, the reversal potential of currents through G A B A A ionophores was around 0 mV and the spontaneous inhibitory postsynaptic currents (slPSCs) were uniformly inward at holding potentials close to the normal resting membrane potential of neurons (-60 to -80 mV). In several experiments, decreasing the intracellular concentration of CI- shifted the reversal potential of the currents to more negative values.
Effect of tetrodotoxin Previous studies have reported the persistence of spontaneous synaptic currents in the absence of action potentials in cultured mammalian neurons ~4 and have alluded to TFX-resistant spontaneous synaptic currents in hippocampal slices ~1'18. However, a quantification of the T F X effect is not available for spontaneous synaptic events in the mammalian CNS. We have investigated the effect of bath-applied TTX up to concentrations of 5/~M and found that 1 /~M was sufficient to block all action potential activity in every cell examined in our preparation. As shown in Fig. 1, T I ' X (1 /tM) reduced the amplitude of the spontaneous synaptic currents and prolonged the inter-event intervals by 76% (i.e. reduced their frequency by about 43%). With a CI driving force o f - 7 0 to -80 mV, large (>200 pA) spontaneous currents were abolished by TTX. When individual events were averaged, no differences were found in the rise times or decay time constants of T-FX-sensitive or qTX-insensitive spontaneous activity. As indicated by the exponential probability density functions fitted to the distributions of the inter-event intervals 47, both "l'TX-sensitive and TTXresistant activity appeared to be randomly generated. In cells with very stable holding currents (e.g. Fig. 1) it was also possible to detect a change in the mean holding current following perfusion of TTX due to the reduction in mean slPSC amplitude and diminished frequency. In the example shown in Fig. 1, a net mean outward current of 30 pA was discernible following the blockade of the larger spontaneous inward currents by q T X corresponding to a net conductance decrease of 0.54 nS. Contribution of spontaneous IPSCs to the resting membrane potential and conductance In order to determine the effective contribution of the spontaneous q~X-resistant synaptic currents to the resting membrane potential we have carried out experiments in current clamp recordings while setting E a at various
potentials. Fig. 2 illustrates such an experiment, whereby virtue of the C1- concentration of the intracellular solution, Eel was set to -55 mV or to 0 mV. Depending on neuronal input resistance (120-280 MK2), the spontaneous IPSPs ranged from 0.05 to 1.5 mV at a membrane potential -15 mV away from E o and from 1 to 15 mV at about 70-80 mV away from E o . It is thus clear that depending on the in situ E o in a given cell and the resting membrane potential, the TFX-resistant slPSCs will be expressed as membrane potential oscillations of various amplitudes. Perhaps even more important is the steady conductance mechanism which persists in the presence of such continuous synaptic activity. Fig. 3 shows the effect of perfusing the competitive G A B A A receptor antagonist bicuculline methiodide (BMI, 5/tM). The net outward current produced after the blockade of the spontaneous events was 60 pA which at the holding potential o f - 6 0
Ecj = - 5 5 m V
Ecj -- 0 m V
R M P = -73 m V
R M P = -72 m V
~,,~J
"~r~v~ '
I~,
/
I
Ii ~1 I L
i
__i I 10ms
mV
Fig. 2. Whole-cell current clamp recordings of 1q'X-resistant spontaneous postsynaptic potentials in two dentate gyrus granule cells with different CI- reversal potentials. Left panel: the intracellular fill contained 135 mM K-gluconate and 5 mM KCI in addition to 2 mM MgCI 2 and 10 m M H E P E S (ph = 7.2). The slPSPs reversed at the E o o f - 5 5 mV. This E o is very close to the reversal potential of evoked IPSPs recorded in these neurons with sharp microelectrodes without dialysis of the cell's interior. In each of the 308 ms long traces the average membrane potential is 0.2 to 0.5 mV more depolarized than the first 10 ms of baseline. Right panel: in another granule cell E o was set to 0 mV by filling the patch electrode with 140 mM KC1, 2 mM MgCI 2 and 10 m M H E P E S (pH 7.2). The averages of the membrane potential traces shown are 2.5 to 5.2 mV more depolarized than the event-free first 10 ms of the recordings. Note the marked baseline noise in this recording compared to the recording on the left. This is probably due to greater basal CI- channel activity at a potential further away from ECl-
145 m V corresponds to a conductance decrease of 1 nS. In general, depending on the frequency of occurrence, the
frequency of TTX-resistant slPSCs. The n e u r o n in Fig. 3 had one of the shortest m e a n inter-event intervals (67.6
net change in conductance following blockade of the slPSCs was 3 0 - 5 0 % of the peak conductance calculated from the amplitude of the averaged events. Submicromolar (0.5-0.8 /~M) concentrations of BMI were also effective in abolishing the slPSCs (n = 3). It is noteworthy that such low concentrations of antagonist have less than significant effects on the amplitude of evoked IPSPs 2"9. We have also used picrotoxin (10-50/~M) to block the slPSCs to demonstrate the C1- channel dependence of the spontaneous events (n = 3).
ms corresponding to an average frequency of 14.8 Hz). Nevertheless, as indicated by the m o n o e x p o n e n t i a l nature of the probability density function of the inter-event intervals (Fig. 3B) and the excellent fit to a Poisson distribution of the n u m b e r of events per successive 120 ms intervals (Fig. 3C), these rapidly occurring slPSCs appear to have a r a n d o m occurrence 47.
Decay time constants and rise times Average slPSCs were obtained by collecting between 400 and 700 single events during a period of 2-5 min. A t
Figs. 1 and 3 also reveal the large variability in the
A
-
5 pM BICUCULLINE
b
t a
a
b
Is0 40./50m=
i 8
C 1111 10e
--
~
0
e.MVen-67.6m:
• 100
200 ~ 400 InlervRI(me)
Inter-Event
! i
500
0 1 2 I 4 II e ? n No,ofEventsInConeecutlve120 me IIIne
Fig. 3. A: perfusion of 5/zM bicuculline blocks TTX-resistant sIPSCs. Uppermost record shows the transmembrane current monitored during wash in of 5/~M bicuculline methiodide. The holding potential was -55 mV, and the patch pipette contained 140 mM CsC1, 10 HEPES, and 2 MgCI2. All extracellular solutions contained 1/tM TTX. An expanded sweep taken just before the switch from control solution (marked by the asterisk and 'a') depicts discrete inward currents, which are completely abolished upon perfusion of 5/~M bicuculline ('b'). B: exponential frequency distribution of the form described in Fig 1. Data collected from the same cell as above in a 2-min period just prior to point a. Binwidth was 5 ms and the mean inter-event interval was T --- 67.6 ms (total number of events = 873). The fitted exponential (solid line) has the form: (N x BW/T) x exp(-t/T). C: frequency distribution of slPSCs in 400 consecutive 120-ms time bins taken from the same period as in B. The continuous curve is a Poisson distribution of the form: 400 x (rex/X!) :,< exp(-m) where X is the number of slPSCs per bin and m is the mean number of slPSCs per consecutive 120-ms time bins (rn = 1.74).
146
A
a
200. •-
-''-"
• :/""
=1
l • • "
.'z~',.'l-.
r - 4x£-04
t" • ~
•
.
ms
""
i.~;~T:~.~.~.s. oi
o
".-~
.
lb
~
1"5
/0
(ms) C
D
125 I00 o (o eq
v
E
4
r - I~E-02
Ip
E
75
i~
2
:....
,
.
.
-...:._ . - . .
;
.
ID
o
•
o
" . " , . ~ M m k ~ . ' , , r.'-
25, 0 0.0
o.,,
1.o 10-90~
1.5
2.0
£5
o
~,
1"o
Rise Time ( m s )
" 1"5
2"o
7 (ms)
Fig. 4. Spontaneous inhibitory postsynaptic currents (sIPSCs) recorded in a dentate gyrus granule cell in the presence of 1/~M TTX at a holding potential of -90 mV. A: 3 individual slPSCs are depicted from 431 such events recorded over a period of 6 min. The 3 synaptic currents have different amplitudes (112 pA, 60 pA and 24 pA) but very similar decay time constants: 5.18 ms, 5.00 ms and 4.96 ms respectively. The mean amplitude of the slPSCs in this cell was 67.5 + 1.7 pA and the average of the decay time constants was 5.58 + 0.1 ms (+ S.E.M.; n = 431) B: graph showing the lack of correlation between decay time constant and amplitude of slPSCs (correlation coefficient = 4 x 10-4). C: histogram of the rise times defined as the time required for the slPSC to grow from 10% to 90% of its peak amplitude. Mean rise time was 0.75 + 0.02 ms (+ S.E.M.; n = 431). D: the decay time constant was not correlated with the rise time of the slPSCs (correlation coefficient = 8 x 10-2).
-55 to -90 mV holding potentials, the size of the average slPSCs ranged between 30-90 pA (e.g. Figs. 1, 4, 5 and 6). In individual neurons there was a large variability in the decay time constants of single slPSCs (e.g. Fig. 4B). Nevertheless, in most cells the average slPSC decayed monoexponentially with a time constant of 3.7-7.2 ms (e.g. Fig. 4). Only in 3 out of 92 neurons the average of 100-400 slPSCs exhibited a double exponential decay with a fast time constant of <1.5 ms (0.85-1.42 ms) and a second slower time constant of 4.5-6.6 ms (e.g. Fig.
A
~ _ --
Y
1 ,°p,'
5A).
Depending on the holding potential, the amplitude of individual TTX-resistant slPSCs ranged between 15-200 pA (e.g. Fig. 4B). Large or small amplitude events had
a
15 I--1
Control
50/a~ Penb~rb. u)
o oo
Fig. 5. A: the average slPSC recorded at a holding potential of -80 mV in a granule cell in the presence of 1/~M T/'X is significantly prolonged following perfusion of 50 pM Na-pentobarbital. The decay phase of the average spontaneous current could be best fitted by a double exponential with the time constants: 1.42 and 6.11 ms. In the presence of the drug both decay time constants more than doubled to 2.92 and 13.46 ms respectively. The amplitude of the average slPSC (39 pA) was not altered by pentobarbital. B: the drug had no significant effect on the frequency distribution of slPSC amplitudes.
10
bJ
0
-100
-i~
-~
-4o
slPSC A m p l i t u d e (pA)
-2o
147
A +35 mV
. . . . . . . . . . .
-90mV ~ a
. . . . . . . . . . . .
.
.
.
.
.
.
~-~
=
.
.
.
.
.
.
.
5
msI10pA
C 17.7 • 3.5
-18.3 i 5.3
m
+
m
-32.6 :t 10.5
to
ID
/
10
:1oo
-io
40
-4o
-=o
0
o
slPSC A m p l i t u d e ( 1 ~ )
slPS.C Amplitude (pA)
lOO
Fig. 6. A: averaged sIPSCs recorded in a dentate gyrus granule cell with a 140 mM CsCI containing patch electrode at two different holding potentials (-90 mV and +35 mV). Note the rectification of the amplitude as well as the decay time constant at the depolarized holding potential. B,C: amplitude distributions of sIPSCs at the two different holding potentials (-90 mV in B. and +35 mV in C.). Both histograms were best fitted by the sum of two normal (Gaussian) distributions with the indicated means and standard deviations. Note that the mean of the smallest unitary events in B. 18.3 pA, which, considering the sIPSC reversal potential of 0 mV and a single GABAg channel current of 1.8 pA, corresponds to the opening of approximately 10 GABAA channels.
similar decay time constants (Fig. 4A,B), indicating adequate control of the membrane voltage 26'5°. The rise-times (measured between 10 and 90% of peak amplitude) of the sIPSCs were <1 ms (Fig. 1C) and did not correlate with the decay time constant (Fig. 4D). In each cell tested (n = 5) the TTX-resistant sIPSCs decayed in a similar fashion as the large spontaneous IPSCs recorded in the absence of TTX. When normalized to the same amplitude, the averages of the large amplitude TTX-sensitive events were essentially superimposable over the averages of the smaller amplitude TFX-resistant sIPSCs. Anesthetic drugs known to enhance the burst time of G A B A A receptor/channels 32'34 had a profound effect on the decay time constant of sIPSCs. Fig. 5 depicts the prolongation of an averaged response by perfusion of 50 p M Na-pentobarbital. The average prolongation of the decay time constant caused by 50 g M pentobarbital was 266 _+ 23% of control (mean + S.E.M.; n = 4). Similar results were obtained with the volatile anesthetics halothane and isoflurane applied at their clinically relevant concentrations 37.
Rectification of slPSC amplitudes and decay time constants G A B A A receptor channels are known for their outward rectification 22'49. This rectification manifests itself in both a non-linear voltage-dependence of the G A B A responses and an increase in the decay time constant of the IPSCs at depolarized potentials lt'13'43. The averaged TTX-resistant sIPSCs had similar rectifying properties. Fig. 6 displays the increase in amplitude and decay time constant of the averaged sIPSCs at depolarized holding potentials. This property of the sIPSCs would make the inhibitory events more effective when neurons are depolarized from their resting m e m b r a n e potentials. DISCUSSION Our results indicate that spontaneous synaptic currents persist when action potential activity is blocked by perfusion of TFX. Only the relatively large events (>200 pA) were abolished by TTX. These large spontaneous currents were probably similar to IPSCs previously described in studies using sharp microelectrode recordings H. In the presence of T F X , the miniature
148 slPSCs had several characteristics of a G A B A A receptormediated CI conductance: (1) they were sensitive to intracellular CI- concentrations and reversed around the Ec~; (2) they displayed an outward rectification in amplitudes and a voltage-dependence of the decay time constant (Fig. 6A) characteristic of both G A B A A channels 22'49 and GABAA-mediated IPSCsl1'13'43; (3) pentobarbital (50 ~tM), which prolongs IPSPs 38 by prolonging the bursting of G A B A A channels 32"34, significantly lengthened the decay time constants of slPSCs (Fig. 5A); (4) the competitive GABAA receptor antagonist bicuculline methiodide (BMI) 15 or the Ci- ionophore blocker picrotoxin completely abolished the slPSCs. The large cell-to-cell variability in the frequency of the slPSCs is probably due to the unequal number of GABAergic synapses on the recorded neurons. A neuron with a large number of inhibitory synapses would be expected to generate slPSCs more frequently. Although it would be tempting to make quantal predictions from the size of the TTX-resistant slPSCs, at the moment such analysis seems inappropriate. For example, multiple Gaussian distributions with different means could have been fitted to the amplitude histograms of Fig. 1, as it has been done for the events in Fig. 5. However, the tedious and often inaccurate fit of amplitude histograms to several normal curves is not by itself proof of the quantal nature of spontaneous events. Cumulative probability distributions have to be made in order to establish the presence of distinct subunits in a continuous distribution of amplitudes 47'4s. We have done such log-normal probability plots of slPSC amplitudes and shown a clearly concave shape of the curve in each cell examined. This indicates the presence of subunits with different mean amplitudes and variances 47'48, but we could not accurately separate and resolve the average size and variance of distinct event subpopulations. The most likely explanation for the encountered difficulty is, akin to the marked variability in slPSC frequencies, the large number of heterogeneous GABAergic synapses innervating each neuron. Further studies using low intensity direct stimulation of GABAergic interneurons are under way to unequivocally resolve the possible quantal nature of slPSCs. The exponential frequency distribution of the interevent intervals and the good Poisson fits to the number of events per consecutive non-overlapping time bins shown in Figs. 1 and 3B,C, indicate that the TTXresistant slPSCs occur randomly. Yet, the released G A B A responsible for the activation of the spontaneous currents did not conform to the classical release mechanism of neurotransmitter substances27: the miniature slPSCs were not dependent upon neuronal action poten-
tial firing and persisted in the presence of the Ca 2+ channel antagonist Cd 2÷ (100-200/~M, n = 6, not shown; and ref. 11). Thus, the release of G A B A responsible for activation of the slPSCs in our preparation resembles the TTX-insensitive Cd2÷-resistant basal G A B A release measured in vivo 51. It is conceivable that ongoing excitatory activity could have spontaneously depolarized inhibitory interneurons to release G A B A intermittently. Indeed, a recent study 31 demonstrated discrete TTX-resistant quisqualate receptor-mediated spontaneous excitatory postsynaptic currents (sEPSCs) in neocortical neurons. No such sEPSCs were noted in the hippocampal CA1 pyramidal cells or granule cells of the dentate gyrus (e.g. Fig. 3A). Nevertheless, these principal cells may spontaneously release an excitatory transmitter to act upon the inhibitory interneurons. We tested this hypothesis by blocking spontaneous excitatory neuronal activity with excitatory amino acid antagonists. Perfusion of 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX, 5 ~tM; n = 3) or o-2amino-5-phosphonovaleric acid (D-AP5, 20/~M; n = 4) had no significant effect on the frequency or average amplitude of the slPSCs (not shown). In light of these findings, it seems unlikely that an excitatory drive controls the spontaneous G A B A release mechanism from inhibitory interneurons. The amplitude and the peak conductance range of the miniature slPSPs indicates that just a few G A B A A channels contribute to a single spontaneous inhibitory current. Bearing in mind that in these neurons the main conductance state of a single G A B A A channel is between 25-30 pS 5'6'18'22, each channel would generate unitary currents of 1.7-2.7 pA at holding potentials o f - 7 0 to -90 mV. Hence the simultaneous opening of 10-20 GABAA channels is sufficient to cause spontaneous unitary events in the range of 20-30 pA. This number of channels is significantly smaller than has been estimated for the inhibitory synapses of Mauthner cells 19, and is an order of magnitude less than previous estimates in hippocampal neurons 11. From the decay time constants of slPSCs, further speculations may be made about the G A B A concentrations necessary to produce such spontaneous currents. We found that the decay of the slPSCs could be described by a single exponential in most cells. Based on the effect of agonist concentration on the kinetics of GABA-activated channels, the monoliganded G A B A A receptor has a less stable and shorter open state than the receptor occupied by two G A B A molecules 35. It has been demonstrated that G A B A concentrations of at least 5/~M are necessary to favor the biliganded receptor state which shows longer open times 35. It appears that such G A B A concentrations must be reached close to the release site in the synaptic cleft to produce the miniature
149 sIPSCs observed in our study. In summary, we have described a perpetual, background G A B A e r g i c inhibition in the m a m m a l i a n CNS analogous to the tonic N M D A receptor activity31'42 and the discrete activation of A M P A receptors recently reported 31. Previously, this tonic inhibition has, for the most part, r e m a i n e d undetected due to the inability to employ high resolution patch clamp recording techniques in brain slices. Such inhibitory activity could critically contribute to controlling the general state of excitability of m a m m a l i a n CNS neurons. For example, submicromolar doses of bicuculline cause the horizontal spread of epileptiform activity in the cerebral cortex 9. Such doses of G A B A A antagonists barely alter evoked inhibitory postsynaptic potentials 2'9 but completely abolish slPSCs. Thus, the transmission of excitation in neocortical aggregates may d e p e n d on the ongoing spontaneous inhibition, and the absence of slPSCs (as in submicromolar concentrations of bicuculline) may permit the generation and
REFERENCES 1 Alger, B.E. and Nicoll, R.A., Spontaneous inhibitory postsynaptic potentials in hippocampus: mechanism for tonic inhibition, Brain Research, 200 (1980) 195-200. 2 Alger, B.E. and Nicoll, R.A., Pharmacological evidence for two kinds of GABA receptor on rat hippocampal pyramidal cells studied in vitro, J. Physiol., 328 (1982) 125-141. 3 Blankenship, J.E. and Kuno, M., Analysis of spontaneous subthreshold activity in spinal motoneurons of the cat, J. Neurophysiol., 31 (1968) 195-209. 4 Blanton, M.G., LoTurco, J.J. and Kriegstein, A.R., Whole cell recording from neurons in slices of reptilian and mammalian cerebral cortex, J. Neurosci. Methods, 30 (1989) 203-210. 5 Bormann, J., Sakmann, B. and Seifert, W., Isolation of GABA-activated single-channel CI- currents in the soma membrane of rat hippocampal neurones, J. Physiol., 341 (1983) 9-10P. 6 Bormann, J., Electrophysiology of GABAA and GABA B receptor subtypes, Trends Neurosci., 11 (1988) 112-116. 7 Bowery, N.G. (Ed.), Actions and Interactions of GABA and Benzodiazepines, Raven Press, New York, 1983. 8 Brock, L.G., Coombs, J.S. and Eccles, J.C., The recording of potentials from motoneurones with an intracellular electrode, J. Physiol., 117 (1952) 431-460. 9 Chagnac-Amitai, Y. and Connors, B.W., Horizontal spread of synchronized activity in neocortex and its control by GABAmediated inhibition, J. Neurophsyiol., 61 (1989) 747-758. 10 Chen, Q.X., Stelzer, A., Kay, A.R. and Wong, R.K.S., GABAA receptor function is regulated by phosphorylation in acutely dissociated guinea-pig hippocampal neurones, J. Physiol., 420 (1990) 207-221. 11 Collingridge, G.L., Gage, P.W. and Robertson, B., Inhibitory post-synaptic currents in rat hippocampal CA1 neurones, J. Physiol., 356 (1984) 551-564. 12 Colomo, F. and Erulkar, S.D., Miniature synaptic potentials at frog spinal neurones in the presence of tetrodotoxin, J. Physiol., 199 (1968) 205-221. 13 Cull-Candy, S.G., Inhibitory synaptic currents in voltageclamped locust muscle fibres desensitized to their excitatory transmitter, Proc. R.Soc. Lond. Ser. B, 221 (1984) 375-383. 14 Cull-Candy, S.G. and Usowicz, M.M., Whole-cell current noise produced by excitatory and inhibitory amino acids in large
spreading of a b e r r a n t activity. It is also plausible that most drugs such as anesthetics, antiepileptics, barbiturates, or benzodiazepines, k n o w n to alter G A B A e r g i c neurotransmission, may exert their primary action on the perpetual spontaneous inhibitory activity. O u r study also opens new possibilities for studying the intracellular regulation of G A B A channels 1°'2°'23'25'44'45 and drug or neurotransmitter interactions with inhibitory mechanisms in the CNS. Future research will be aimed at determining if alterations of such 'silent' inhibitory activity take place following chronic changes in n e u r o n a l excitability7'29' 39-41,46
Acknowledgements. This work was supported in part by NIH Grants NS-12151, RR05353-27 and a Fellowship from the Klingenstein Fund to I.M.T.S.O. is a Howard Hughes Predoctoral Fellow and K.J.S. is a Dana Postdoctoral Fellow. We would like to thank Dr. J. Dempster for providing the Strathclyde Electrophysiology Software and J. T. Palmer for expert technical assistance.
cerebellar neurones of the rat, J. Physiol., 415 (1989) 533-553. 15 Curtis, D.R., Duggan, A.W., Felix, D. and Johnston, G.A.R., GABA, bicucuiline and central inhibition, Nature, 226 (1970) 1222-1224. 16 Eccles, J.C., The mechanism of synaptic transmission, Ergebn. Physiol., 51 (1961) 299-430. 17 Edwards, EA. and Gage, P.W., Seasonal changes in inhibitory currents in rat hippocampus, Neurosci. Lett., 84 (1988) 266-270. 18 Edwards, F.A., Konnerth, A., Sakmann, B. and Takahashi, T., A thin slice preparation for patch clamp recordings from neurones of the mammalian central nervous system, Pflagers Arch., 414 (1989) 600-612. 19 Faber, D.S. and Korn, H., Transmission at a central inhibitory synapse. I: Magnitude of unitary postsynaptic conductance change and kinetics of channel activation, J. Neurophysiol., 48 (1982) 654-677. 20 Feltz, A., Demeneix, B., Feltz, A., Taleb, O., Trouslard, J., Bossu, J.-L. and Dupont, J.-L., Intracellular effectors and modulators of GABA-A and GABA-B receptors: a commentary, Biochimie, 69 (1987) 395-406. 21 Gage, P.W. and Robertson, B., Prolongation of inhibitory postsynaptic currents by pentobarbitone, haiothane and ketamine in CA1 pyramidal cells in rat hippocampus, Br. J. Pharmacol., 85 (1985) 675-681. 22 Gray, R. and Johnston, D., Rectification of single GABA-gated chloride channels in adult hippocampal neurons, J. Neurophysiol., 54 (1985) 134-141. 23 Gyenes, M., Farrant, M. and Farb, D.H., 'Run-down' of GABA A receptor function during whole-cell recording: a possible role for phosphorylation, Mol. Pharmacol., 34 (1988) 719-723. 24 Hubbard, J.I., Stenhouse, D. and Eccles, R.M., Origin of synaptic noise, Science, 157 (1967) 33'0-331. 25 Inoue, M., Oomura, Y., Yakushiji, T. and Akaike, N., Intracellular calcium ions decrease the affinity of the GABA receptor, Nature, 324 (1986) 156-158. 26 Johnston, D. and Brown, T.H., Interpretation of voltage-clamp measurements in hippocampal neurons, J. Neurophysiol., 50 (1983) 464-486. 27 Katz, B., The Release of Neurotransmitter Substances, Liverpool University Press, Liverpool, 1969. 28 King, G.L., Knox, J.J. and Dingledine, R., Reduction of inhibition by a benzodiazepine antagonist, Ro15-1788, in the rat
150 hippocampal slice, Neuroscience, 15 (1985) 371-378. 29 Krogsgaard-Larsen, P., Scheel-Kruger, J. and Kofod, H. (Eds.), GABA-Neurotransrnitters, Munkgaard Press, Copenhagen, 1979. 30 Kuno, M., Quantum aspects of central and ganglionic synaptic transmission in vertebrates, Physiol. Rev., 51 (1971) 647-678. 31 LoTurco, J.J., Mody, I. and Kriegstein, A.R., Differential activation of glutamate receptors by spontaneously released transmitter in slices of neocortex, Neurosci. Lett., 114 (1990) 265-271. 32 Macdonald, R.L., Twyman, R.E., Rogers, C.J. and Weddle, M.G., Pentobarbital regulation of the kinetic properties of GABA receptor chloride channels, Adv. Biochem. Psychopharmacol., 45 (1988) 61-71. 33 Marty, A. and Neher, E., Tight-seal whole-cell recording. In B. Sakmann and E. Neher (Eds.), Single-Channel Recording, Plenum Press, New York, 1983, pp. 107-122. 34 Mathers, D.A., The GABA A receptor: new insights from single-channel recording, Synapse, 1 (1987) 96-101. 35 Mathers, D.A. and Wang, Y., Effect of agonist concentration on the lifetime of GABA-activated membrane channels in spinal cord neurons, Synapse, 2 (1988) 627-632. 36 Miles, R. and Wong, R.K.S., Unitary inhibitory synaptic potentials in the guinea-pig hippocampus in vitro, J. Physiol., 356 (1984) 97-113. 37 Mody, I., Tanelian, D.L. and Maclver, M.B., Halothane enhances tonic neuronal inhibition by elevating intracellular calcium, Brain Research, in press. 38 Nicoll, R.A., Eccles, J.C., Oshima, T. and Rubia, R., Prolongation of hippocampal inhibitory postsynaptic potentials by barbiturates, Nature, 258 (1975) 625-627. 39 Prince, D.A. and Connors, B.W. Mechanism of interictal epileptogenesis. In A.V. Delgado-Escueta, A.A. Ward, D.M. Woodbury and R.J. Porter (Eds.), Basic Mechanisms of epilepsies; Molecular and Cellular Approaches, (Advances in Neurology, Vol. 44), Raven Press, New York, 1986, pp. 275-299. 40 Roberts, E. Failure of GABAergic inhibition: a key to local and
41 42
43 44
45
46 47
48
49 50
51
global seizures. In A.V. Delgado-Escueta, A.A. Ward, D.M. Woodbury and R.J. Porter (Eds.), Basic Mechanisms of epilepsies; Molecular and Cellular Approaches, (Advances in Neurology, Vol. 44), Raven Press, New York, 1986, pp. 319-341. Roberts, E., Chase, T.N. and Tower, D.B. (Eds.), GABA in Nervous System Function, Raven Press, New York, 1976. Sah, P., Hestrin, S. and NicoU, R.A., Tonic activation of NMDA receptors by ambient glutamate enhances excitability of neurons, Science, 246 (1989) 815-818. Segal, M. and Barker, J.L., Rat hippocampal neurons in culture: voltage-clamp analysis of inhibitory synaptic connections, J. Neurophysiol., 52 (1984) 469-487. Sigel, E. and Baur, R., Activation of protein kinase C differentially modulates neuronal Na +, Ca 2+ and ~,-aminobutyrate type A channels, Proc. Natl. Acad. Sci. U.S.A., 85 (1988) 6192-6196. Stelzer, A., Kay, A.R. and Wong, R.K.S., GABAA-receptor function in hippocampal cells is maintained by phosphorylation factors, Science, 241 (1988) 339-341. Tallman, J.F., Paul, S.M., Skolnick, P. and Gallager, D.W., Receptors for the age of anxiety: pharmacology of the benzodiazepines, Science, 207 (1980) 274-281. Van der Kloot, W., Kita, H. and Cohen, I., The timing of miniature end plate potentials, Progr. Neurobiol., 4 (1975) 269-326. Van der Kloot, W., Statistical and graphical methods for testing the hypothesis that quanta are made up of subunits, J. Neurosci. Methods, 27 (1989) 81-89. Weiss, D.S., Membrane potential modulates the activation of GABA-gated channels, J. Neurophysiol., 59 (1988) 514-527. Weiss, D.S., Barnes, Jr., E.M. and Hablitz, J.J., Whole-cell and single channel recordings of GABA-gated currents in cultured chick cerebral neurons, J. Neurophysiol., 59 (1988) 495-513. Westerink, B.H.C. and De Vries, J.B., On the origin of extracellular GABA collected by brain microdialysis and assayed by a simplified on-line method, Naunyn-Schmiedeberg's Arch. Pharmacol., 339 (1989) 603-607.