JOURNALOF NEUROPHYSIOLOGY Vol. 66, No. 3, September 199 1. Printed in U.S.A.

Inhibitory Transmission in the Basolateral Amygdala DONALD

G. RAINNIE,

EFTIHIA

K. ASPRODINI,

AND

PATRICIA

SHINNICK-GALLAGHER

Departmentof Pharmacologyand Toxicology, Universityof TexasMedical Branch, Galveston,Texas 77550 SUMMARY

AND

CONCLUSIONS

1. Intracellular recording techniques were used to characterize synaptic inhibitory postsynaptic potentials (IPSPs)recorded from neurons of the basolateral nucleus of the amygdala (BLA). Bipolar electrodes positioned in the stria terminalis (ST) or lateral amygdala (LA) were used to evoke synaptic responsesat a frequency of 0.25 Hz. 2. Two synaptic waveforms having IPSP components could be evoked by electrical stimulation of either pathway: a biphasic, excitatory postsynaptic potential (EPSP),fast-IPSP (f-IPSP) waveform, and a multiphasic, EPSP,f-IPSP, and subsequent slow-IPSP (s-IPSP)waveform. Expression of either waveform was dependent on the site of stimulation. ST stimulation evoked a similar number of biphasic (45%) and multiphasic (50%) synaptic responses. In contrast, stimulation of the LA pathway evoked mainly (80%) multiphasic synaptic responses. 3. Both the f- and s-IPSP elicited by ST stimulation could be reduced in amplitude in the presence of the glutamatergic, Nmethyl-D-aspartate (NMDA) antagonist, (DL)-2-amino-5-phosphonovaleric acid (APV, 50 PM), and were abolished by the glutamatergic, non-NMDA antagonist, 6-cyano-7-nitroquinoxaline2,3-dione (CNQX, 10 ELM).In contrast, a CNQX-resistant f-IPSP was evoked with LA stimulation and abolished by subsequent addition of bicuculline methiodide (BMI), a y-aminobutyric acid (GABA,) receptor antagonist, suggesting direct inhibition of BLA neurons by GABAergic LA interneurons. The sensitivity of the s-IPSPsand the f-IPSPs to glutamatergic antagonists suggeststhe presence of feed-forward inhibition onto BLA neurons. 4. The f-IPSP possessedcharacteristics of potentials mediated by GABA, receptors linked to Cl- channels, namely, a reversal potential of -70 mV, a decrease in membrane resistance (13.5 Mfi) recorded at -60 mV, a block by BMI, and potentiation by sodium pentobarbital (NaPB). 5. The s-IPSPwas associated with a resistance decrease of 4.5 MS2,a reversal potential of -95 mV, and was reversibly depressed (w 66%) by 2-hydroxy-saclofen ( 100 PM), suggesting activation of GABA, receptors. 6. The large resistance change associated with the f-IPSP, its temporal overlap with evoked EPSPs, and the development of both spontaneous and evoked burst firing in the presence of BMI suggeststhat the f-IPSP determines the primary state of excitability in BLA neurons. 7. NaPB ( 100 PM) not only increased f-IPSP amplitude and duration but also reduced the amplitude and duration of the GABA,-mediated s-IPSP, and the observation that only a f-IPSP is expressed in the presence of CNQX on stimulation of the LA pathway suggeststhat a heterogeneous population of GABA interneurons is involved in BLA synaptic transmission. 8. Superfusion of NaPB or 2-hydroxy-saclofen (2-OH-SAC) resulted in the inhibition of spontaneous EPSPsand IPSPs suggesting presynaptic inhibition of tonic glutamate release onto BLA neurons. 9. These experiments provide evidence for the existence of two electrophysiologically distinct GABAergic IPSPsin the BLA. The degree of expression of these two IPSPsis determined by the input 0022-3077/9

1 $1 SO Copyright

pathway, hence only feed-forward inhibition occurs via the ST input, whereas both feed-forward and direct inhibition occurs via the LA pathway. The f-IPSP has been shown to determine the primary state of excitability of BLA neurons, therefore any alteration of the inhibitory drive into the BLA via either pathway will determine neuronal excitability within the nucleus. This may explain the sensitivity of the BLA to the kindling phenomenon.

INTRODUCTION

The basolateral nucleus of the amygdala (BLA) has been shown to contain moderate amounts of y-aminobutyric acid (GABA) (Ben Ari et al. 1976) and GABA immunoreactivity (Ottersen et al. 1986). With the use of Golgi-Kopsch staining, McDonald (1982) categorized BLA neurons morphologically into three classes (see Rainnie et al. 199 1). Class II neurons, comprising - 5% of the cells impregnated (McDonald 1982), have been identified as GABAergic (McDonald 1985) and form axosomatic or axodendritic synapses with Class I neurons, the principal cell type. Lesioning of the stria terminalis (ST) had no effect on GABA content in the basolateral nucleus, suggesting these neurons are intrinsic (Ben-Ari et al. 1976). Electrophysiological studies showed that electrical stimulation of amygdaloid afferents elicited excitatory-inhibitory responses in some studies (Le Gal La Salle 1976; Prelevic et al. 1976), whereas in other reports strong inhibitory responses were most often observed (Ben-Ari et al. 1974). Activation of other amygdaloid structures also caused inhibition, suggesting the presence of a rich inhibitory-intraamygdaloid network (Le Gal La Salle 1976). Furthermore, excitatory responses to glutamate were curtailed by a shortlatency inhibitory response (Le Gal La Salle 1976), suggesting the inhibition may be due to recurrent collaterals (Le Gal La Salle and Ben-Ari 198 I). Microiontophoretic application of GABA in the amygdala suppressed spontaneous or glutamate-induced discharge (Ben-Ari and Kelly 1976), and long-lasting inhibitory postsynaptic potentials (IPSPs) have been recorded in response to stimulation of the ST and lateral nucleus in the amygdala in vivo (de Molina et al. 198 1) and in vitro (Takagi and Yamamoto 198 l), a component of which was found to be sensitive to bicuculline (Takagi and Yamamoto 198 1). Altogether these data suggest that GABA is a major inhibitory transmitter in the BLA. Recently, we reported that, in a model of epilepsy, kindling, GABAergic IPSPs are absent in the contralateral amygdala, suggesting an impairment of GABAergic function (Gean et al. 1989). The purpose of this study was to characterize, physiologically and pharmacologically, the GABA-mediated synaptic transmission in the

0 199 1 The American

Physiological

Society

999

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.

1000

BLA occurring via two pathways: lateral amygdala (LA).

RAINNIE

I) the ST, and 2) the

METHODS

Intracellular current-clamp recordings were obtained from BLA neurons, and the IPSPs evoked by stimulation of either the ST or the LA nucleus were studied. Bipolar stainless steel electrodes (Kopf Instruments, SNE-200X; effective resistance, 50 Mfi) were used to stimulate electrically (150-ps square-wave pulses) the ST or LA. For the majority of experiments, stimulus intensities, generated by a Grass S88 stimulator, were adjusted to threshold for peak orthodromic IPSP activation. In all experiments five individual synaptic potentials were averaged. Figures illustrated and statistics are both of averaged responses unless otherwise stated. For further details of the methods, see the preceding paper (Rainnie et al. 199 1). Drugs used in this study were as follows: 100 PM sodium pentobarbitd (NaPB; Robinson Laboratory, San Francisco, CA; made fresh before each experiment); 2-hydroxy-saclofen Q-OHSAC, 3-amino~2(4-chlorophenyl)-2-hydroxy-propylsulfonic acid; Research Biochemicals, Natick, MA; stored frozen in 500-~1 lmM. aliquots); 30 pm bicuculline methiodide (BMI; Sigma Chemicals, St. Louis, MO); 10 PM &cyano-7-nitroquinoxalline-2,3dione (CNQX; Tocris Neuramin, Essex, England; and 50 CLM (DL)-2-amino-$phosphonovaleric acid(APV; Research Biochemicals, Natick, MA). RESULTS

Evoked 1PSP.s

ET AL.

STRIA

A CONTROL

CNQX

10 ~JM

+

4 8 mV

BMT 30 plul

6oms

I

LA

B CONTROL

CNQX

10 PM

CNQX 10 pM + BMI 30 pM

FIG. 1. Stimulation of the LA but not ST pathway evoked a CNQX-resistant f-IPSP in BLA neurons. A: in control artificial cerebrospinal fluid (aCSF), orthodromic ST stimulation (20 V) evoked a multiphasic waveform (top). Six minutes after the introduction of CNQX ( 10 PM), the same stimulus intensity elicited a s-EPSP (mid&e). Addition of BMI (30 PM) resulted in a slight enhancement of the s-EPSP. B: in the same neuron, LA stimulation (10 V) evoked a biphasic waveform (top). In the presence of CNQX, a f-IPSP component could still be observed after LA stimulation (middle). Subsequent addition of BMI (bottom) abolished the f-IPSP and unmasked a pronounced s-EPSP. Membrane potential was held at -60 mV throughout.

As described in the preceding paper, electrical stimulation of either the ST or LA elicited both bi- and multiphasic synaptic waveforms (Fig. 1, A and B, top traces). Contamination of IPSPs by excitatory postsynaptic potentials (EPSPs) could not be avoided because both the fast and slow IPSP (f- and s-IPSP, respectively) were reduced in amplitude by superfusion of the glutamatergic antagonists CNQX and APV. Of the two pathways, the IPSPs evoked via the ST were the most sensitive as shown in Fig. 1A tude (-5.3 t 0.6 mV) in 26.1 t 0.8 ms (n = 9) and had a (middEe), where CNQX (10 PM) abolished both the fast and duration of 60.2 t 10.1 ms (n = 9) when recorded at -60 slow components. Frequently, however, a CNQX-resistant mV. On the other hand, the multiphasic waveform consistf-IPSP was observed after stimulation of the LA pathway ing of an EPSP followed by a f-IPSP and a subsequent s(Fig. lB, middle). This CNQX-resistant f-IPSP could be IPSP (Fig. 2C) was observed after 50% (n = 1 l/22) of ST blocked by the addition of the GABA, receptor antagonist stimulations and 80% of LA stimulations. The s-IPSP had a BMI (30 PM) to the superfusate revealing the presence of a peak amplitude of (-3.9 t 0.5 mV) and a time to peak of s-EPSP (Fig. 1B, bottom). Because of the temporal overlap 118.1 t 4.6 ms (n = 9) and was characterized by a proof the EPSP and f-IPSP and the sensitivity of the f-IPSP to longed hyperpolarization lasting 45 1.8 t 32.3 ms (n = 9). glutamatergic antagonists, the properties of the f-IPSP The amplitudes of both the f- and s-IPSP were dependent could not be studied in isolation. For this reason the dura- on the intensity of stimulation (Fig. 2, B and D). Increases tion of the f-IPSP was defined as the time from when the in the stimulus intensity caused an increase in the peak falling phase of the f-EPSP crossed the baseline to when the amplitude of the f-IPSP in biphasic waveforms elicited by membrane potential returned to the baseline at the terminastimulation of the ST (Fig. 2, A and B). A similar relationtion of the f-IPSP. In addition, there were two further differship between stimulus intensity and the amplitude of the ences between the IPSPs evoked from the two pathways. s-IPSP was observed in neurons having multiphasic waveFirst, the voltage required to achieve maximal IPSP ampliforms (Fig. 2, C and D). In these neurons low stimulus tudes differed significantly [ST = 16.1 t 1.9 V (mean -+ SE), intensities evoked EPSPs followed by f-IPSPs, whereas n = 15; LA = 11.7 t 0.9 V, n = 13; P = 0.051. Second, the higher stimulus intensities evoked s-IPSPs that followed ffrequency of occurrence of either postsynaptic waveform IPSPS. depended on the site of stimulation. Specifically, the biphaSeparation of the f-IPSP from the s-IPSP was more easily sic potential consisting of an EPSP that was terminated observed with membrane polarization after DC current insharply by the onset of a f-IPSP (Fig. 2A) was observed after jection (Fig. 3A). Both the f- (A) and s-IPSP (A) increased in 45% (n = 10/22) of ST stimulations and only 20% (n = amplitude with membrane depolarization from the RMP 4/20) of LA stimulations. The f-IPSP reached peak ampli(-69 mV). In most cells a depolarizing wave (Fig. 3A, &) was

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.

INHIBITORY

TRANSMISSION

IN THE BLA

1001

Stimulus

B 0.0;

10 :

@&

Intensity 20 :

:

(V) 30 :

:

40 :

:

t

A---A f-Ipsp \

A \

A\A

A

-10.01

0.01\ Ij \ 0

-2.o-

10

20

30

A \

A-A

f-lesp

A-A

S-IPSP

A

-4.o-

A\A\A,

-6.0-g

A-A

-8.0 -10.0

40

1

FIG. 2. Amplitudes of the evoked f- and s-IPSP are dependent on stimulus intensity. A: in a neuron showing only a biphasic response to ST stimulation, increasing stimulus intensity (S-40 V) caused a graded increase in the amplitude of the f-IPSP (A). B: plot of peak f-IPSP amplitude recorded in A as a function of stimulus intensity. C: in another neuron, ST stimulation evoked a multiphasic waveform having both f- (A triangle) and s- (A) IPSPs whose amplitudes were dependent on stimulus intensity (5-40 V). D: plot of IPSP amplitude as a function of stimulus intensity at the time points shown in C. Membrane potential was held at -60 mV in both A and B. Note different calibrations in A and C.

A

C

Membrane -100

-90 I1

%0 mV -4

--

Potential -80 ,I A-A A-A

I_

(mV) -70 ,, _ t-Imp s-IPSP

-60 .

. 10.0 -8.0 Y 6.0 4.0 2.0

-100 #

:

-90 :

:

-80 :

A-A

A

:

-70 :

f

-60 :

t+sp

\ A\ -70

= $ 3 0.

- 10.0

8.0 6.0

% g

4.0

3

A\ /\

-2.0

3

-4.0 -6.0

FIG. 3. Voltage dependence of evoked IPSPs in BLA neurons. A: in this neuron, ST stimulation (15 V) evoked a multiphasic waveform showing both f- (A) and a s- (A) IPSP components when held at -60 mV with DC current injection. Holding the membrane at gradually more hyperpolarized potentials caused the f-IPSP to reverse polarity between -70 and -80 mV and the s-IPSP to null at -90 mV. Note the apparent depolarizing wave (4) between the peaks of the f- and s-IPSP. B: in another neuron, following the same protocol as described in A, the f-IPSP component of the biphasic response to ST stimulation ( 12 V) reversed polarity close to -70 mV. C: plot of mean f- (A) and s- (A) IPSP amplitudes as a function of membrane potential. D: a plot of mean biphasic f-IPSP amplitude as a function of membrane potential.

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.

1002

RAINNIE

apparent between the peaks of the f- and s-IPSP; it is unclear whether this phenomenon may represent the concurrent decay phase of the f-IPSP and appearance of the sIPSP, or is itself an active synaptic component. On membrane hyperpolarization the f-IPSP decreased in amplitude, nulled at -70 mV, and reversed in polarity becoming a depolarizing potential, which increased in amplitude with further membrane hyperpolarization (Fig. 3B).In contrast, the s-IPSP showed a gradual reduction in amplitude with membrane hyperpolarization and had a reversal potential at or near -90 mV (Fig. 3, A and C). When mean f- and s-IPSP amplitudes were plotted as a function of membrane potential (Fig. 3C), different reversal potentials for the fIPSP (-70 t 1.8 mV; n = 9) and the s-IPSP (-95 t 0.9 mV; n = 9) were consistently observed. The characteristics of the f-IPSP, i.e., time to peak and peak amplitude, were similar in both bi- and multiphasic waveforms, and we tested whether the reversal potential of the f-IPSP recorded in a biphasic synaptic response was similar to that observed in a multiphasic potential. The fIPSP recorded in a biphasic potential also increased in amplitude with membrane depolarization from the RMP (-72 mV), showed a null potential of -70 mV, and reversed to a depolarizing potential, which increased in amplitude with membrane hyperpolarization (Fig. 3B). When the mean fIPSP amplitude was plotted as a function of membrane potential (Fig. 30), a similar plot to that of the f-IPSP of the multiphasic waveform was obtained. Thus it would appear

ET AL.

that the f-IPSPs in both the bi- and multiphasic waveforms are identical. Both types of IPSP were accompanied by an decrease in membrane resistance, but the magnitude of the resistance change was much higher at the peak of the f-IPSP (13.5 t 1.5 MQ; n = 6) than at the peak of the s-IPSP (4.5 t 0.5 MQ; n = 5) when recorded at -60 mV.

SpontaneouslyoccurringIPSPs Spontaneous IPSPs were recorded in 58% (n = 57) of the neurons examined and occurred at a frequency of 17 t 3/min (n = 12). The spontaneous IPSPs appeared to consist of two different populations; one having a fast time course, whereas the other exhibited a much slower time course. The fast time course IPSP was most frequently observed and had a mean amplitude of 2.9 t 0.6 mV (n = 15) when recorded at -60 mV. This spontaneous IPSP was similar in appearance to the evoked f-IPSP. That is, it increased in amplitude with membrane depolarization from the RMP, had a null potential about -75 mV, and reversed polarity at membrane potentials more negative than -75 mV (Fig. 4A). Occasionally, a slower IPSP was observed after the spontaneous f-IPSP (Fig. 4A, \ ). This slower spontaneous IPSP had characteristics similar to those of the evoked sIPSP, namely, the spontaneous s-IPSP increased in amplitude with membrane depolarization from the RMP, gradually decreased in amplitude with membrane hyperpolarization, and reversed in polarity close to -90 mV. When the Membrane

0-O

Spont.

a..**0. 1 *.

A***

4mV

n-0

A s-IPSP

ii 0.

n

(mV)

s-IPSP

l *.

n-

Potential

-u

Spont.

f-IPSP

Spent,

s-IPSP

l .

l *.

IS -I FIG. 4. Voltage dependence of both fast and slow spontaneous IPSPs recorded in BLA neurons. A: membrane polarization with DC current injection revealed a clear separation of the 2 IPSP components to occur at -80 mV. Here the fast component had reversed polarity, whereas the slow component ( \ ) was still recorded as a hyperpolarizing potential. B: peak IPSP amplitudes, at different membrane potentials, were measured at the time to peak for the 2 components recorded at -80 mV. Data obtained from the neuron illustrated in A. Peak amplitude of spontaneous f- and s-IPSPs (mand q , respectively) plotted as a function of membrane potential revealed a reversal potential of -75 mV for the fast and -95 mV for the slow resembles that of evoked f- and s-IPSPs ( 0). component. C: voltage dependence of spontaneous f- and s-IPSPs (-) l

l

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.

INHIBITORY

TRANSMISSION

amplitudes of the f- and s-IPSP (m and q , respectively) were plotted as a function of membrane potential, a clear separation between the reversal potentials was observed (Fig. 4B). The similarities between the voltage dependence of the spontaneous f- and s-IPSPs (m and 0) to those of the evoked f- and s-IPSP (A and A) are shown in Fig. 4C, where IPSP amplitude was plotted as a function of membrane potential.

IN THE BLA

1003

A

-I

10 mV so ma

Eflectsof NaPB on evokedIPSPs The similarity in voltage sensitivity and fast onset of the evoked f-IPSP to those reported in other cortical regions (for review see Krnjevic 1984) indicated that this potential may be mediated by GABA, receptor activation resulting in an increased chloride conductance (Kelly et al. 1969). This receptor has been shown to be specifically antagonized by bicuculline (BIC) (Curtis et al. 197 1). However, addition of BMI (30 PM) to the superfusate resulted in the generation of burst firing in BLA neurons (Fig. 54). To characterize the f-IPSP further, we tested the effects of NaPB on evoked f-IPSP amplitude because NaPB has previously been demonstrated to potentiate BIC-sensitive IPSPs observed in other CNS regions (Alger and Nicoll 1982a,b; Ransom and Barker 1975; Tseng and Haberly 1988). A typical effect of NaPB superfusion on evoked synaptic responses in BLA neurons is shown in Fig. 5B. In this neuron, held at -60 mV with DC current injection, stimulation of the ST evoked an EPSP followed by a f-IPSP (A). Addition of NaPB to the superfusate resulted in an increase not only in the amplitude of the f-IPSP (control: 6.2 mV; NaPB: -8.3 mV) but also in its duration (control: 74 ms; NaPB: 420 ms). When f-IPSP amplitude was plotted as a function of membrane potential (Fig. 5C), NaPB caused an increase in the f-IPSP at all membrane potentials tested. In those cells tested, NaPB increased the f-IPSP amplitude by 18% (control: -7.7 t 1.1 mV; NaPB: -8.6 t 1.2 mV; n = 5) and f-IPSP duration by 197% (control: 124 t 15 ms; NaPB: 322 t 33 ms; n = 5). The increase in amplitude and duration was fully reversible on washout of the drug (not shown). Superfusing NaPB caused a membrane hyperpolarization (1.6 t 0.6 mV; n = 5), in neurons at a holding potential of -60 mV.

Efects of 2-OH-SAC on evokedIPSPs In the hippocampus the late IPSP has been shown to be mediated by the activation of postsynaptic GABA, receptors that are coupled to a potassium conductance by a pertussis toxin-sensitive G-protein (Andrade et al. 1986; Hablitz and Thalmann 1987; Newberry and Nicoll 1984). Recently, 2-OH-SAC (3.amino-2-(4.chlorophenyl)-2-hydroxy-propylsulfonic acid), a sulfonic analogue of the GABA, agonist baclofen, has been shown to possess antagonist activity at peripheral and central GABA, receptors (Kerr et al. 1988). Because of the similarities in the voltage dependence, reversal potential, time to peak, and duration of the evoked s-IPSP observed in BLA neurons to those of the late IPSP recorded from hippocampal neurons, we tested the sensitivity of the s-IPSP to 2-OH-SAC. The effect of 2-OH-SAC superfusion on the multiphasic

6

J

C

Membrane

Potential

CONTROL

4 nv 1

(mv)

-6.0

i

-8.0

,z

-10.0

FIG. 5. GABA, receptor activation mediates the f-IPSP. A: stimulation of the ST ( 10 V) evoked burst firing responses in BLA neurons treated with BMI (30 PM). B: in another neuron, NaPB ( 100 PM) enhanced both the amplitude and duration of the f-IPSP recorded after ST stimulation ( 15 V). Membrane potential, -65 mV in A and -60 mV in B. C a plot of f-IPSP amplitude as a function of membrane potentials showed that, when compared with control, NaPB (v) increased the amplitude and duration of the f-IPSP at all potentials tested (A). The reversal potential however, remained unaltered.

waveform evoked by ST stimulation is shown in Fig. 6A. In this neuron, holding the membrane potential at -60 mV to maximize the IPSP amplitudes, 15-V stimulation evoked an EPSP, an f-IPSP (peak amplitude, -8.1 mV; A), and a subsequent s-IPSP (peak amplitude, -5.9 mV; 0). After superfusion with 2-OH-SAC (100 PM), the amplitude of the s-IPSP was markedly reduced (peak amplitude, - 1.7 mV; +); in contrast, no effect was observed on the amplitude of the f-IPSP (peak amplitude, -8.3 mV; A). When the peak f-IPSP amplitude was plotted as a function of membrane potential (Fig. 6B), 2-OH-SAC had little effect on the amplitude of the f-IPSP at all membrane potentials tested, nor was there an alteration of the reversal potential. The s-IPSP, however, was reduced in amplitude at potentials more depolarized than the RMP (-69 mV) and nulled around -80 mV. The plot of membrane input resistance as a function of membrane potential (Fig. 6C) showed that 2-OH-SAC reduced input resistance at all membrane potentials tested and that the reduction was larger at more

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.

1004

RAINNIE

A

n

Membrane 1

:

-90

;

f

Potential -60

:::r

(mV)

-70

-60 ;

;4.0 1

Membrane

Potential

-60 I

-70. .

-90 I 0-O 0-0

A-A A-A 0-O l -e

100 pM

J

-60 mV

-100

2-OH-SAC,

ET AL.

control f-lPSP 2-OH-SAC f-IPSP control S-BP P-OH-SAC s-IPSP

-

control 2-OH-SAC

FIG. 6. GABA, receptor activation mediates the s-IPSP. A: LA stimulation (15 V) evoked a multiphasic response in control aCSF. Super-fusion with 2-OH-SAC (100 PM) selectively blocked the s-IPSP (+ and 0) without affecting the f-IPSP component (A and A). Membrane potential, -60 mV. B: plot of IPSP amplitude as a function of membrane potential showed that the f-IPSP amplitude before (A) and during (A) the application of 2-OH-SAC was unaltered at all membrane potentials tested (A and A are superimposed). The reversal potential also remained unchanged. In contrast, the s-IPSP was depressed at potentials below -70 mV in the presence of 2-OH-SAC (+) as compared with control values (0). C: a plot membrane input resistance as a function of membrane potential; 2-OH-SAC decreased resistance at all potentials tested; the decrease was more apparent at -60 mV than at more hyperpolarized levels.

(mV) --

-60. I

0

0

0 P

t

25.0 z 3

-

120.0

depolarized levels. The 9.9% reduction in input resistance increased the amplitude and duration of the f-IPSP, as (control: 38.5 t 4.8 MQ; 2-OH-SAC: 34.0 t 3.0 MQ; n = 5), mentioned above, but also markedly reduced the ampliobserved at -60 mV, may account for the mean 17.1% re- tude and duration of the s-IPSP. duction (control: -7.7 t 0.3 mV; 2-OH-SAC: -6.4 t 0.6 mV; n = 5) in the f-IPSP amplitude. It cannot, however, Effectsof NaPB and 2-OH-SAC on spontaneous fully account for the 64.3% reduction (control: -5.5 + _ 0.4 postsynapticpotentials (PSPs) mV; 2-OH-SAC: -2.1 t 0.7 mV; n = 5) in the s-IPSP ampliSuperfusion of either NaPB or 2-OH-SAC completely tude observed at -60 mV. In common with NaPB, 2-OHblocked both spontaneous IPSPs and EPSPs (Fig. 8, A and SAC caused a 4.3 t 1.7-mV (n = 5) hyperpolarization, from B; top traces). This action appeared to be independent of a holding potential of -60 mV, in BLA neurons.

GABA, and GABA, interactions In most experiments the effects of NaPB were studied on those neurons showing only a biphasic response to stimulation of the ST or LA to avoid contamination of the tail of the f-IPSP by the onset of the s-IPSP. However, NaPB was superfused after washout of 2-OH-SAC in a few neurons exhibiting a multiphasic response (n = 3). An example is shown in Fig. 7 where stimulating the ST after washout of 2-OH-SAC (+) resulted in the reappearance of a clear multiphasic response (0). Superfusion of NaPB (A) not only

2-OH-SAC,

A

CONTROL

NaPB,

B

100 pM

CONTROL

100 pM

2-OH-SAC, A

100 pM

NaPB, 100 pM

7. Enhancement of the f-IPSP reduces the amplitude of the sIPSP. In the presence of 2-OH-SAC, LA stimulation ( 15 V) evoked biphasic response (+). On washout, a multiphasic synaptic potential was observed with the return of the s-IPSP (0). Addition of PBNa not only increased the amplitude and duration of the f-IPSP, but also dramatically reduced the amplitude and duration of the s-IPSP (A). Membrane potential, -60 mV. FIG.

5mV 2 set -I

FIG. 8. Both NaPB and 2-OH-SAC abolish spontaneous PSPs in BLA neurons. A: in this neuron, spontaneous f- and s-IPSPs were recorded at -60 mV (top). After addition of NaPB, no spontaneous potentials were recorded (bottom) at the same membrane potential. B: in another neuron, both excitatory and inhibitory spontaneous potentials were observed at -60 mV (top). Superfusing 2-OH-SAC abolished both excitatory and inhibitory spontaneous potentials (bottom; membrane potential, -60 mV).

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.

INHIBITORY

TRANSMISSION

the effects on membrane hyperpolarization shared by both compounds because returning the membrane potential to the predrug level (usually -60 mV) with DC current injection did not result in the reappearance of spontaneous PSPs (Fig. 8, A and B; bottom traces). DISCUSSION

The main conclusions drawn from this study are as follows: I) f- and s-IPSPs can be recorded independent of action potential activity in the impaled neuron and show a sensitivity to glutamate antagonists, indicating a degree of feed-forward inhibition onto BLA neurons; 2) direct intraamygdaloid inhibition of BLA neurons occurs via projections of LA GABAergic interneurons; 3) both the f-IPSP and the s-IPSP observed in BLA neurons after stimulation of the ST or LA appear to be mediated through activation of GABA receptors; 4) the f-IPSP is probably subserved by activation of GABA, receptors as it is associated with a large decrease in membrane resistance ( 13.5 MQ), showed a reversal potential of -70 mV, was blocked by the GABA, receptor antagonist BMI, and potentiated by NaPB; 5) the s-IPSP is probably mediated through activation of GABA, receptors as it is associated with a small decrease in resistance (4.5 Mn), showed a reversal potential of -90 mV, was depressed by the GABA, receptor antagonist 2-OH-SAC; 6) the GABA, receptor-mediated f-IPSP determines the primary state of excitability of BLA neurons because of its large conductance and temporal overlap with evoked EPSPs; 7) the block of the GABA,-mediated IPSP in the presence of NaPB suggests a possible heterogeneity of GABA interneurons involved in BLA synaptic transmission; and 8) the inhibition of spontaneous synaptic activity by both NaPB and 2-OH-SAC is probably mediated by an indirect action “upstream” from the impaled BLA neurons, possibly by potentiating presynaptic GABA inhibition. Ever since the original observations by Krnjevic and Schwartz (1967) suggesting that GABA could be the neurotransmitter mediating IPSPs recorded from cortical neurons, and that these potentials were chloride dependent (Kelly et al. 1969), the role of GABA in synaptic transmission has been well established. Subsequently, two GABA receptor subtypes have been identified: I) the BIC-sensitive GABA, receptor and 2) the BIC-insensitive GABA, receptor (Dunlap 198 1; Hill and Bower-y 198 1; for GABA reviews see Bormann 1988; Krogsgaard-Larsen 1988). In this study we have demonstrated that activation of both types of receptors can occur during synaptic transmission in BLA neurons. Involvement of GABA, receptor activation transmission in BLA neurons

in synaptic

GABA activation of the GABA, receptor and subsequent opening of a chloride channel can be modulated by the activation of distinct binding sites for benzodiazepines and barbiturates on the receptor itself (for a review see Bormann 1988; Gallager and Tallman 1990). Furthermore, binding of GABA can be blocked by the specific competi-

IN THE BLA

1005

tive antagonist bicuculline (BIC) (Curtis et al. 197 1; Frere et al. 1982; Nowak et al. 1982). We have demonstrated that the fast IPSP, observed after stimulation of either the ST or LA pathway, has characteristics similar to those reported for activation of GABA, receptors, namely, a fast onset (time to peak, 26.1 t 0.8 ms), a relatively short duration (60.2 t 10.1 ms), a reversal potential of -70 mV, and an associated large decrease in membrane resistance ( 13.5 t 1.5 Ma). As shown in Fig. 5A, superfusion with BMI, caused burst firing in BLA neurons, indicating that neuronal excitability was determined, in part, by a GABA,-mediated process. Further, in the presence of CNQX, the amplitude of the APV-sensitive s-EPSP was enhanced by BMI. This was most prominent after stimulation of the LA (see Fig. 1B) and suggested a CNQX-resistant GABA,-mediated shunt of the s-EPSP. Conversely, the amplitude and duration of the f-IPSP was potentiated by 100 PM NaPB, further supporting the hypothesis that the f-IPSP is a GABAA-mediated potential. This barbiturate anesthetic has been previously shown to potentiate exogenous GABA responses in spinal cord cultures (Ransom and Barker 1975) and IPSPs in CA1 neurons (Nicoll et al. 1975). Recent reports have demonstrated that NaPB acts directly at the GABA, receptor complex to increase the mean open and decrease the mean closed time of chloride channels (Akaike et al. 1990; Nakahiro et al. 1989; Twyman et al. 1989), hence increasing the apparent affinity of the GABA, receptor to GABA. Involvement of GABA, receptor activation transmission in BLA neurons

in synaptic

After the differentiation of the GABA receptor subtypes (Bowery et al. 1980; Hill and Bowery 198 l), the GABA, receptor has been shown to affect synaptic transmission by both a pre- (Anderson and Mitchell 1985; Ault and Nadler 1982; Blaxter and Carlen 1985; Johnston et al. 1980; Kato et al. 1982; Lanthorn and Cotman 198 1) and a postsynaptic mechanism (Newberry and Nicoll 1984). The GABA,-mediated postsynaptic hyperpolarization results from an increase in potassium conductance (Alger 1984), has a reversal potential of -90 mV (Gahwiler and Brown 1985; Newberry and Nicoll 1985), and is dependent on activation of a pertussis toxin-sensitive G-protein (Andrade et al. 1986). We have demonstrated that the s-IPSP, observed after stimulation of either the ST or LA pathway, decreased in amplitude with membrane hyperpolarization, had a reversal potential of -90 mV, and was inhibited 64% by the specific GABA, antagonist 2-OH-SAC (100 PM). These data suggest that the s-IPSP observed in BLA neurons is mediated mainly by activation of GABA, receptors, although a contribution by other neurotransmitters can not be overlooked. A 2-OH-SAC-sensitive GABA,-mediated s-IPSP has also been reported in CA 1 neurons of the hippocampus after stimulation of the stratum radiatum (Davies et al. 1990; Segal 1990). Stimulation of the Schaffer collateralcommissural pathway evoked a s-IPSP, which was inhibited by phaclofen (Dutar and Nicoll 1988a), a less potent antagonist at GABA, receptors (Kerr et al. 1987), and abolished by pertussis toxin (Dutar and Nicoll 1988b). Phaclo-

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.

1006

RAINNIE

fen has also been shown to be capable of partially blocking the late hyperpolarizing potential (LHP) observed in dorsal lateral septal neurons (Hasuo and Gallagher 1988). Interaction between the expression of GABA,- and GABA,-mediated potentials In those neurons displaying both a f-IPSP and a s-IPSP, NaPB caused not only the potentiation of the GABA, response, but also a decrease in the amplitude and duration of the s-IPSP. Several possible explanations could account for this phenomenon: 1) increasing a GABA,-mediated shunt, 2) enhancing a dendritic GABA, depolarizing potential, or 3) inhibiting GABAergic interneurons responsible for the s-IPSP. First, the large resistance decrease associated with GABA, receptor activation when potentiated by NaPB may be of sufficient magnitude and duration to shunt the s-IPSP. Second, NaPB has been shown to potentiate a late BIC-sensitive depolarizing potential in hippocampal pyramidal neurons seen in response to exogenous GABA application (Ransom and Barker 1975; Wong and Watkins 1982) and orthodromic stimulation of both the stratum radiatum (Alger and Nicoll 1979, 1982b) and the mossy fiber tract (Thalmann 1988). This depolarizing potential has been proposed to be due to GABA release in the dendritic field. If the s-IPSP was mediated by activation of receptors in the dendritic field, it may be possible that NaPB potentiates an intrinsic GABA,-mediated dendritic depolarizing potential, which then masks the s-IPSP. The absence of a depolarizing potential after NaPB superfusion in those neurons displaying only a f-IPSP might argue against the above hypothesis. It is possible, however, that in the neurons having only a f-IPSP, the GABA inter-neurons that are activated do not make synaptic contacts with the dendrites of BLA neurons, suggesting a possible heterogeneity of GABA interneurons. Alger and Nicoll (1982b) did not observe a depolarizing response in the presence of NaPB after activation of those interneurons responsible for GABA,-mediated feedback inhibition in the hippocampus. These previously published data support the idea of a heterogeneity of GABA interneurons. Finally, it is possible that the interneurons responsible for the f-IPSP make synaptic contacts with the excitatory input onto those interneurons mediating the s-IPSP. Activation of GABA, receptors on these excitatory terminals in the presence of NaPB would result in depression of the excitatory drive and hence a reduction of the s-IPSP amplitude. This circuitry also implies a heterogeneous population of GABA interneurons and suggests that those interneurons mediating the f-IPSP may also modulate the activity of the interneurons responsible for expression of the s-IPSP. Another study by O’Beirne and co-workers (1987) reported that NaPB enhanced a LHP, [a synaptic potential considered to be synonymous with the s-IPSP, also called the late inhibitory postsynaptic potential (l-IPSP) by other workers] in CA1 neurons. It is possible that the apparent increase in the LHP was due to the marked increase in both the amplitude and duration of the GABA,-mediated fIPSP, masking the actual effect of NaPB on the LHP. Indeed, a similar GABA,-mediated s-IPSP that is not only

ET

AL.

insensitive to, but also potentiated by, BIC has been reported in the CNS of rats (Connors et al. 1988; Crunelli et al. 1988; Hasuo and Gallagher 1988; Soltesz et al. 1989), turtle (Kriegstein and Connors 1986), and humans (McCormick 1989). The mechanism by which BIC may enhance the s-IPSP amplitude and duration is similar but opposite to those of NaPB attenuation of the s-IPSP mentioned above. Interestingly, Segal(l990) has demonstrated that 4-aminopyridine superfusion can elicit a s-IPSP in hippocampal neurons that is independent of f-IPSP activation, further indicating a distinction between those neurons mediating the f-IPSP and those mediating the s-IPSP. Actions of NaPB and 2-OH-SAC on membrane potential and spontaneous IPSPs in BLA neurons The observation that both NaPB and 2-OH-SAC hyperpolarize the membrane and block spontaneous IPSPs and EPSPs in BLA neurons appeared paradoxical in view of their opposite actions on postsynaptic GABAergic potentials. It is possible that these two compounds acting at different presynaptic sites ultimately cause the same postsynaptic response. Previous reports have indicated that 2-OHSAC has no effect on passive membrane properties (Kerr et al. 1988; Lambert et al. 1989); conversely, NaPB evoked a picrotoxin-sensitive (Macdonald and Barker 1978; Nicoll 1975), and a tetrodotoxin-resistant K+-mediated (O’Beirne et al. 1987) hyperpolarization. In our study, NaPB was shown to cause an inhibition of spontaneous PSPs. When the membrane potential was repolarized to the predrug level, spontaneous PSPs in BLA neurons did not reappear in the presence of either compound, indicating that this action was independent of membrane hyperpolarization. It is possible, therefore, that PSP inhibition occurred via a presynaptic mechanism. GABA, receptors have been reported to exist presynaptically (Anderson and Mitchell 1985; Bonanno et al. 1988; Kato et al. 1982) and act to inhibit the release of both excitatory amino acids (EAAs) and GABA from presynaptic terminals. Although there appears to be a consensus of opinion that the pre- and postsynaptic GABA, receptors are mediated through different ionic mechanisms, the pharmacology of the presynaptic GABA, receptor is unclear. In the CNS the baclofen-induced reduction of both inhibitory and excitatory transmission has been reported to be sensitive (Curtis et al. 1988; Kerr et al. 1988; Ong et al. 1990) and insensitive (Dolphin and Huston 1990) to 2-OH-SAC. These inconsistent findings may arise from a heterogeneity of presynaptic GABA, receptors on the terminals of inhibitory and excitatory axons. Indeed, in CA1 neurons, increasing stimulus intensity is capable of overcoming baclofen-induced inhibition of excitatory but not of inhibitory transmission (Peet and McLennan 1986). Furthermore, both GABA, and GABA, receptors have been demonstrated to modulate transmitter release at excitatory synapses (Miwa et al. 1990; Peng and Frank 1989a,b). In contrast, GABA release at inhibitory synapses has been shown to be reduced by GABA, but not GABA, agonists (Neal and Shah 1989; Pittaluga et al. 1987; Waldmeier et al. 1988). A similar GABA receptor distribution and heterogeneity within the BLA may account for the

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.

INHIBITORY

TRANSMISSION

inhibition of spontaneous PSPs caused by both 2-OH-SAC and NaPB. Membrane hyperpolarization induced by both NaPB and 2-OH-SAC may therefore result from a reduction of the tonic release of glutamate onto BLA neurons. Another possibility is that both compounds act directly as GABA agonists on excitatory terminals to reduce EAA release. Indeed, preliminary results from our laboratory and those of others (Curtis et al. 1988; Davies et al. 1990) indicate that 2-OH-SAC may act as a partial agonist at presynaptic GABA, receptors, thereby reducing EAA release. Furthermore, NaPB has been shown to suppress spontaneous synaptic activity in cultured mouse spinal cord neurons, possibly by reducing N and L calcium currents in presynaptic terminals (Gross and Macdonald 1988). Functional consequences of GABA, and GABA, receptor activation in the BLA Several other CNS structures have two types of IPSPs that closely resemble those of the BLA. These include the hippocampus (Alger 1984; Davies et al. 1990; Lambert et al. 1989; McCarren and Alger 1985; Turner 1990), lateral geniculate nucleus (Crunelli et al. 1988; Soltesz et al. 1989), lateral septum (Stevens et al. 1987), cortex, and neocortex (Connors et al. 1988; Howe et al. 1987; Tseng and Haberly 1988). Although the characteristics of the two different types of IPSPs reported in the above areas have distinctive characteristics for each cell type, the similarity between brain regions is striking. Consequently, it is reasonable to assume that the IPSPs subserve the same function throughout these regions. The large resistance decrease, and possible somatic and/ or initial dendrite origin of the f-IPSP, ensures that BLA neurons will be effectively inhibited from firing action potentials even in the face of strong excitatory stimuli. The fast onset and short duration of the IPSP also allows precise moment-to-moment adjustment of the cellular response to excitatory inputs. In contrast, the small conductance increase and long duration associated with the s-IPSP would indicate that this input may only regulate low-frequency excitatory input during times of normal neuronal functioning. A similar conclusion was reported by Soltesz et al. (1989), who showed that a GABA,-mediated IPSP could abolish low- but not high-frequency repetitive discharges in lateral geniculate neurons. As mentioned above, superfusion with BMI but not 2-OH-SAC results in burst firing in BLA neurons. It would appear that the f-IPSP is the primary determinant of synchronicity in the BLA. This observation concurs with those of Traub et al. (1987), who demonstrated that, in a computerized model of the CA3 region of the hippocampus, the f-IPSP is required to block burst activity, and the s-IPSP can block the full development of synchronization. Because synaptic transmission in the BLA has both a fast, GABA,-mediated and a slow, GABA,-mediated inhibitory component, any reduction in the expression of one or both of these potentials may move the basolateral nucleus toward a state favoring epileptogenesis. In conclusion, the complexity of inhibitory synaptic circuitry within the BLA suggests that modulation of this inhibition would play a key role in potential integrative properties of the nucleus.

IN THE BLA

1007

We thank Dr. Joel Gallagher for critical reading of the manuscript. This work was supported by National Institute of Neurological Disorders and Stroke Grant NS-24643 to P. Shinnick-Gallagher. Address reprint requests to P. Shinnick-Gallagher. Received 7 January 199 1; accepted in final form 17 April 199 1. REFERENCES AKAIKE, N., TOKUTOMI, N., AND IKEMOTO, Y. Augmentation of GABAinduced current in frog sensory neurons by pentobarbital. Am. J. Physiol. 258 (Cell Physiol. 27): C452-C460, 1990. ALGER, B. E. Characteristics of a slow hyperpolarising synaptic potential in rat hippocampal pyramidal cells in vitro. J. Neurophysiol. 52: 892-9 10, 1984. ALGER, B. E. AND NICOLL, R. A. GABA-mediated biphasic inhibitory responses in hippocampus. Nature Lond. 28 1: 3 15-3 17, 1979. ALGER, B. E. AND NICOLL, R. A. Feed-forward dendritic inhibition in rat hippocampal pyramidal cells studied in vitro. J. Physiol. Land. 328: 105-123, 1982a. 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. Lond. 328: 125-141, 1982b. ANDERSON, R. AND MITCHELL, R. Evidence for GABAB autoreceptors in median eminence. Eur. J. Pharmacol. 118: 355-358, 1985. ANDRADE, R., MALENKA, R. C., AND NICOLL, R. A. A G protein couples serotonin and GABA, receptors to the same channels in hippocampus. Science Wash. DC 234: 126 1- 1265, 1986. AULT, B. AND NADLER, J. V. Baclofen selectively inhibits transmission at synapses made by axons of CA3 pyramidal cells in the hippocampal slice (Abstract). J. Pharmacol. Exp. Ther. 223: 29 1, 1982. BEN-ARI, Y., KANAZAWA, I., AND ZIGMOND, R. E. Regional distribution of glutamate decarboxylase and GABA within the amygdaloid complex and stria terminalis system of the rat. J. Neurochem. 26: 1279-1283, 1976a. BEN-ARI, Y. AND KELLY, J. S. Dopamine evoked inhibition of single cells of the feline putamen and basolateral amygdala. J. Physiol. Lond. 256: 1-21, 1976b. BEN-ARI, Y., LE GAL LA SALLE, G., AND CHAMPAGNAT, J.-C. Lateral amygdala unit activity. I. Relationship between spontaneous and evoked activity. Electroencephalogr. Clin. Neurophysiol. 37: 449-46 I, 1974. BLAXTER, T. J. AND CARLEN, P. L. Pre- and postsynaptic effects of baclofen in the rat hippocampal slice. Brain Res. 34 1: 195- 199, 1985. BONANNO, G., FONTANA, G., AND RAITERI, M. Phaclofen antagonizes GABA at autoreceptors regulating release in rat cerebral cortex. Eur. J. Pharmacol. 154: 223-224, 1988. BORMANN, J. Electrophysiology of GABA, and GABAB receptor subtypes. Trends Neurosci. 11: 112- 118, 1988. BOWERY, N. G., HILL, D. R., HUDSON, A. L., DOBLE, A., MIDDLEMISS, D. N., SHAW, J., AND TURNBULL, M. (-)Baclofen decreases neurotransmitter release in the mammalian CNS by an action at a novel GABA receptor. Nature Lond. 283: 92-94, 1980. CONNORS, B. W., MALENKA, R. C., AND SILVA, L. R. Two inhibitory postsynaptic potentials, and GABA, and GABA, receptor-mediated responses in neocortex of rat and cat. J. Physiol. Land. 406: 443-468, 1988. CRUNELLI, V., HABY, M., JASSIK-GERSCHENFELD, D., LEREWHE, N., AND PIRCHIO, M. Cl-- and K+-dependent inhibitory postsynaptic potentials evoked by interneurones of the rat lateral geniculate nucleus. J. Physiol. Lond. CURTIS,

399:

153-176,

1988.

D. R., DUGGAN, A. W., FELIX, D., AND JOHNSTON, G. A. R. Bicuculline, an antagonist of GABA and synaptic inhibition in the spinal cord of the cat. Brain Res. 32: 69-96, 197 1. CURTIS, D. R., GYNTHER, B. D., BEATTIE, D. T., KERR, D. I. B., AND PRAGER, R. H. Baclofen antagonism by 2-hydroxy-saclofen in the cat spinal cord. Neurosci. Lett. 92: 97- 10 1, 1988. DAVIES, C. H., DAVIES, S. N., AND COLLINGRIDGE, G. L. Paired-pulse depression of monosynaptic GABA-mediated inhibitory postsynaptic responses in rat hippocampus. J. Physiol. Lond. 424: 5 13-53 1, 1990. DE MOLINA, A. F., YAJEYA, J., COLINO, A., AND MA VELASCO, J. Cyclic hyperpolarising activity in amygdaloid neurons and limbic epilepsy. In: The Amygdaloid Complex, edited by Y. Ben-Ari. Amsterdam: Elsevier/ North-Holland, 198 1, p. 465-474.

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.

1008

RAINNIE

A. C. AND HUSTON, E. The pharmacology of the GABA,-mediated inhibition of calcium channel currents and transmitter release in cultured rat dorsal root ganglion and cerebellar neurones (Abstract). J. Physiol. Lond. 422: 18P, 1990. DUNLAP, K. Two types of y-aminobutyric acid receptor on embryonic sensory neurones. Br. J. Pharmacol. 74: 579-585, 198 1. DUTAR, P. AND NICOLL, R. A. A physiological role for GABAB receptors in the central nervous system. Nature Lond. 332: 156- 158, 1988a. DUTAR, P. AND NICOLL, R. A. Pre- and postsynaptic GABA, receptors in the hippocampus have different pharmacological properties. Neuron 1: 585-591, 1988b. FRERE, R. C., MACDONALD, R. L., AND YOUNG, A. B. GABA binding and bicuculline in spinal cord and cortical membranes from adult rat and from mouse neurons in cell culture. Brain Res. 244: 145- 153, 1982. GAHWILER, B. H. AND BROWN, D. A. GABA,-receptor-activated K+ current in voltage-clamped CA3 pyramidal cells in hippocampal cultures. Proc. Natl. Acad. Sci. USA 82: 1558-1562, 1985. GALLAGER, D. W. AND TALLMAN, J. F. Relationship of GABA, receptor heterogeneity to regional differences in drug response. Neurochem. Res. 15: 113-118, 1990. GEAN, P.-W., SHINNICK-GALLAGHER, P., AND ANDERSON, A. C. Spontaneous epileptiform activity and alteration of GABA- and NMDA-mediated neurotransmission in amygdala neurons kindled in vivo. Brain Res. 494: 177-181, 1989. GROSS, R. A. AND MACDONALD, R. L. Barbiturates and nifedipine have different and selective effects on calcium currents of mouse DRG neurons in culture: a possible basis for differing clinical actions. The 1987 S. Wier Mitchell Award. Neurology 38: 443-45 1, 1988. HABLITZ, J. J. AND THALMANN, R. H. Conductance changes underlying a late synaptic hyperpolarization in hippocampal CA3 neurons. J. Neurophysiol. 58: 160-179, 1987. HASUO, H. AND GALLAGHER, J. P. Comparison of antagonism by phaclofen and baclofen induced hyperpolarizations and synaptically mediated late hyperpolarizing potentials recorded intracellularly from rat dorsolatera1 septal neurons. Neurosci. Lett. 86: 77-8 1, 1988. HILL, D. R. AND BOWERY, N. G. 3H-baclofen and 3H-GABA bind to bicuculline-insensitive GABAB sites in rat brain. Nature Land. 290: 149-152, 1981. HOWE, J. R., SUTOR, B., AND ZEIGLGANSBERGER, W. Baclofen reduces post-synaptic potentials of rat cortical neurones by an action other than its hyperpolarising action. J. Physiol. Lond. 384: 539-569, 1987. JOHNSTON, G. A. R., HAILSTONE, M. H., AND FREEMAN, C. G. Baclofen: stereoselective inhibition of excitant amino acid release. J. Pharm. Pharmacol. 32: 230-231, 1980. KATO, K., GOTO, M., AND FUKUDA, H. Baclofen: inhibition of the release of L-[3H]glutamate and L-[3H]aspartate from rat whole brain synaptosomes. Gen. Pharmacol. 13: 445-447, 1982. KELLY, J. S., KRNJEVIC, K., MORRIS, M. E., AND YIM, G. K. W. Anionic permeability of cortical neurones. Exp. Brain Res. 7: 1 l-3 1, 1969. KERR, D. I. B., ONG, J., JOHNSTON, G. A. R., ABBENANTE, J., AND PRAGER, R. H. 2-Hydroxy-saclofen: an improved antagonist at central and peripheral GABAB receptors. Neurosci. Lett. 92: 92-96, 1988. KERR, D. I. B., ONG, J., PRAGER, R. H., GYNTHER, B. D., AND CURTIS, D. R. Phaclofen: a peripheral and central baclofen antagonist. Brain Res. 405: 150-154, 1987. KRIEGSTEIN, A. R. AND CONNORS, B. W. Cellular physiology of the turtle visual cortex: synaptic properties and intrinsic circuitry. J. Neurosci. 6: 178-191, 1986. KRNJEVIC, K. Neurotransmitters in the cerebral cortex: a general account. In: Cerebral Cortex. Functional Properties of Cortical Cells, edited by E. G. Jones and A. Peters. New York: Plenum, 1984, vol. 2, p. 39-62. KRNJEVIC, K. AND SCHWARTZ, S. The action of y-aminobutyric acid on cortical neurones. Exp. Brain Res. 3: 320-336, 1967. KROGSGAARD-LARSEN, P. GABA synaptic mechanisms: stereochemical and conformational requirements. Med. Res. Rev. 8: 27-56, 1988. LAMBERT, N. A., HARRISON, N. L., KERR, D. I. B., ONG, J., PRAGER, R. H., AND TEYLER, T. J. Blockade of the late IPSP in rat CA1 hippocampal neurons by 2-hydroxy-saclofen. Neurosci. Lett. 107: 125- 128, DOLPHIN,

1989. LANTHORN,

T. H. AND COTMAN, C. W. Baclofen selectively inhibits excitatory synaptic transmission in the hippocampus. Brain Res. 225: 17 l-

178,1981. LE GAL LA SALLE,

G. Unitary responses in the amygdaloid complex follow-

ET AL. ing stimulation of various diencephalic structures. Brain Res. 118: 475478, 1976. LE GAL LA SALLE,

G. AND BEN-ARI, Y. Unit activity in the amygdaloid complex: a review. In: The Amygdaloid Complex, edited by Y. Ben-AI-i. Amsterdam: Elsevier/North-Holland, 198 1, p. 227-237. MACDONALD, R. L. AND BARKER, J. L. Different actions of anticonvulsant and anaesthetic barbiturates revealed by use of cultured mammalian neurons. Science Wash. DC200: 775-777, 1978. MCCARREN, M. AND ALGER, B. E. Use dependent depression of IPSPs in rat hippocampal pyramidal cells in vitro. J. Neurophysiol. 53: 557-57 1, 1985. MCCORMICK, D. A. GABA as an inhibitory neurotransmitter in human cerebral cortex. J. Neurophysiol. 62: 10 18- 1027, 1989. MCDONALD, A. J. Neurons of the lateral and basolateral amygdaloid nuclei: a Golgi study in the rat. J. Comp. Neurol. 2 12: 293-3 12, 1982. MCDONALD, A. J. Immunohistochemical identification of y-aminobutyric acid-containing neurons in the rat basolateral amygdala. Neurosci. Lett. 53: 203-207, 1985. MIWA, A., UI, M., AND KAWAI, N. G protein is coupled to presynaptic glutamate and GABA receptors in lobster neuromuscular synapse. J. Neurophysiol. 63: 173- 180, 1990. NAKAHIRO, M., YEH, J. Z., BRUNNER, E., AND NARAHASHI, T. General anaesthetics modulate GABA receptor channel complex in rat dorsal root ganglion neurons. FASEB J. 3: 1850-l 854, 1989. NEAL, M. J. AND SHAH, M. A. Baclofen and phaclofen modulate GABA release from slices of rat cerebral cortex and spinal cord but not from retina. Br. J. Pharmacol. 98: 105-l 12, 1989. NEWBERRY, N. R. AND NICOLL, R. A. Direct hyperpolarizing action of baclofen on hippocampal pyramidal cells. Nature Land. 308: 450-452, 1984. NEWBERRY,

N. R. AND NICOLL, R. A. Comparison of the action of baclofen with y-aminobutyric acid on rat hippocampal pyramidal cells in vitro. J. Physiol. Land. 360: 16 l-l 85, 1985. NICOLL, R. A. Pentobarbital: action on frog motoneurons. Brain Res. 96: 119-123,1975. NICOLL, R. A., ECCLES,

J. C., OSHIMA, T., AND RUBIA, F. Prolongation of hippocampal inhibitory postsynaptic potentials by barbiturates. Nature Land. 258: 625-627, 1975. NOWACK, L. M., YOUNG, A. B., AND MACDONALD, R. L. GABA and bicuculline actions on mouse spinal cord and cortical neurons in cell culture. Brain Res. 244: 155-164, 1982. O’BEIRNE, M., GUREVICH, N., AND CARLEN, P. L. Pentobarbital inhibits hippocampal neurons by increasing potassium conductance. Can. J. Physiol. Pharmacol. 65: 36-4 1, 1987. ONG, J., KERR, D. I. B., JOHNSTON, G. A. R., AND HALL, R. G. Differing actions of baclofen and 3-aminopropylphosphinic acid in rat neocortical slices. Neurosci. Lett. 109: 169-173, 1990. OTTERSEN, 0. P., FISCHER, B. O., RINVIK, E., AND STORM-MATHISEN, J. Putative amino acid transmitters in the amygdala (Abstract). Adv. Exp. Med. Biol. 203: 53-66, 1986. PEET, M. J. AND MCLENNAN, H. Pre- and postsynaptic actions of baclofen: blockade of the late synaptically-evoked hyperpolarisation of CA 1 hippocampal neurones. Exp. Brain Res. 6 1: 567-574, 1986. PENG, Y.-Y. AND FRANK, E. Activation of GABAB receptors causes presynaptic inhibition at synapses between muscle spindle afferents and motoneurons in the spinal cord of bullfrogs. J. Neurosci. 9: 1502- 15 15, 1989a. PENG, Y.-Y. AND FRANK, E. Activation of GABA, receptors causes presynaptic and postsynaptic inhibition at synapses between muscle spindle afferents and motoneurons in the spinal cord of bullfrogs. J. Neurosci. 9: 1516-1522, 1989b. PITTALUGA, A., ASARO, D., PELLEGRINI, G., AND RAITERI, M. Studies on ‘3H:GABA and endogenous GABA release in rat cerebral cortex suggest the presence of autoreceptors of the GABAB type. Eur. J. Pharmacol. 144:45-52, PRELEVIC,

1987.

S., BURNHAM, W. M., AND GLOOR, P. A microelectrode study of amygdaloid afferents: temporal neocortical inputs. Brain Res. 105:

437-457,1976. RAINNIE, D. G., ASPRODINI,

E. K., AND SHINNICK-GALLAGHER, P. Excitatory transmission in the basolateral amydala. J. Neurophysiol. 66: 986-

998,1991. RANSOM,

B. R. AND BARKER,

J. L. Pentobarbital modulates transmitter

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.

INHIBITORY

TRANSMISSION

effects on mouse spinal neurones grown in tissue culture. Nature Land. 254: 703-705, 1975. SEGAL, M. A subset of local interneurons generate slow inhibitory postsynaptic potentials in hippocampal neurons. Brain Rex 5 11: 163-164, 1990. SOLTESZ,I., LIGHTOWLER,S.,LERESCHE,N., ANDCRUNELLI, V.Onthe properties and origin of the GABAB inhibitory postsynaptic potential recorded in morphologically identified projection cells of the cat dorsal lateral geniculate nucleus. Neuroscience 33: 23-33, 1989. STEVENS, D. R., GALLAGHER, J. P., AND SHINNICK-GALLAGHER, P. In vitro studies of the role of y-aminobutyric acid in inhibition in the lateral septum of the rat. Synapse NY 1: 184-190, 1987. TAKAGI, M. AND YAMAMOTO, C. The long-lasting inhibition recorded in vitro from the lateral nucleus of the amygdala. Brain Res. 206: 474-478, 1981. THALMANN, R. H. Blockade of late inhibitory postsynaptic potential in hippocampal CA3 neurons in vitro reveals a late depolarising potential that is augmented by pentobarbital. Neurosci. Lett. 95: 155- 160, 1988.

IN THE BLA

1009

R. D., MILES, R., AND WONG, R. K. S. Models of synchronized hippocampal bursts in the presence on inhibition. I. Single population events. J. Neurophysiol. 58: 739-764, 1987. TSENG, G.-F. AND HABERLY, L. B. Characterisation of synaptically mediated fast and slow inhibitory processes in piriform cortex in an in vitro slice preparation. J. Neurophysiol. 59: 1352-l 376, 1988. TURNER, D. A. Feed-forward inhibitory potentials and excitatory interactions in guinea-pig hippocampal pyramidal cells. J. Physiol. Lond. 422: 333-350, 1990. TWYMAN, R. E., ROGERS, C. J., AND MACDONALD, R. L. Pentobarbital and picrotoxin have reciprocal actions on single GABA,,, receptor channels. Neurosci. Lett. 96: 89-95, 1989. WALDMEIER, P.C., WICIU, P., FELDTRGUER, J.J., AND BAUMAN-N, P. A. Potential involvement of a baclofen-sensitive autoreceptor in the modulation of the release of endogenous GABA from rat brain slices in vitro. Naunyn-SchmiedebergS Arch. Pharmakol. 337: 289-295, 1988. WONG, R. K. S. AND WATKINS, D. J. Cellular factors influencing GABA response in hippocampal pyramidal cells. J. Neurophysiol. 48: $38-95 1, 1982. TRAUB,

Downloaded from www.physiology.org/journal/jn by ${individualUser.givenNames} ${individualUser.surname} (129.186.138.035) on January 16, 2019.