Brain Research, 595 (1992) 220-227 Elsevier Science Publishers B.V.

220

BRES 18229

Ethanol enhances synaptically evoked G A B A A receptor-mediated responses in cerebral cortical neurons in rat brain slices W i l l i a m R. P r o c t o r

a,b, B r a n d i

L. Soldo a, A n d r e a M. A l l a n c a n d T h o m a s V. D u n w i d d i e a,b

a Department of Pharmacology, Unirersity of Colorado Heal,t1 Sciences Center, Denrer, CO 80262 (USA), t, Dem'er Veterans Administration Medical J'~.~-

~l|lll~F,

i'~ ....

~.

f~f~

Oll~l"~/i

lJrlP

A ~ --..,J

i.~l;/llt gJ'p ~ - ~ ' OVK~z-V ~ U J t - I /

fill"

(" lr~ . . . .

• ......

,

~dr'lr'J_.~l.:_._.

L J C ~ I I I | F l l U I I | OJ • ~ ' L / | I W | I

~,

lil~_l.:..-~-..

IF..:

FFU311111~|UII U l i l t

.....

:,~..

~l,'lly

p.I._~I

~I"

iB~J:_:.._

(n.

I~ . . . . : _

i,f,¢-~ ,~,~lllfb

OI~IlUUi UJ IVI~t~I~III~¢ O l . I . ~ U t l I ~ I V i ~

UJI

il/

/lrrg'.

4k

~ IL./dl,/~ll]

(Accepted 9 June 1992)

key words: Ethanol; v-Aminobutyric acid; Chloride flux; Synaptic modulation; Brain slice; Hippocampus; Cerebral cortex; Electrophysiology

Previous intracellular electrophysiological studies on rat hippocampal brain slices have shown very little effect of acute ethanol application on synaptically evoked GABA A receptor-mediated responses recorded in CA1 pyramidal neurons. The present study was designed to compare the effects of ethanol on pyramidal neurons in the hippocampus and cerebral cortex. Using conventional intracellular microelectrodes (60-80 M.O) to impale cortical neurons in brain slices, 80 mM ethanol application did not affect the membrane input impedance nor evoked EPSPs, but significantly affected the resting membrane potential (usually a 2-5 mV hyperpolarization). When stimulus-evoked GABAA-mediated IPSCs were studied using whole-cell recordings from cortical neurons voltage-clamped at depolarizing potentials, monophasic IPSCs were evoked that were blocked by bicuculline, increased by pentobarbital, and enhanced by ethanol superfusion in a dose dependent manner over the range of 20-160 raM. Hippocampal IPSCs recorded under identical conditions were not enhanced by ethanol. Parallel studies of GABA-stimulated 3 6 0 flux measurements in microsacs prepared from hippocampal, cerebral cortical and cerebellar tissue demonstrated that ethanol significantly enhanced (30-50%) "~CI- flux in microsacs derived from the cerebral cortex and cerebellum, but not in microsacs prepared from the hippocampus. These results demonstrate that there are clear brain region-dependent differences in the way that GABA A receptor function is altered by acute ethanol, and that these differences are apparent not only as an enhancement of responses to exogenous GABA, but also as a facilitation of the responses to endogenous GABA released from inhibitory nerve termimds during synaptic activation.

INTRODUCTION

Historically, ethanol has been hypothesized to act on brain cell membranes in a relatively non-specific manner by modifying the fluidity of the plasma membrane, resulting in lipid rearrangement adjacent to membrane-bound protein ~'.a°. Such changes in effective viscosity might then affect the function of all of the proteins embedded within the membrane. This rather general, non-specific view of ethanol action in the brain has now been modified somewhat because of the demonstration that ethanol has selective, site-specific actions on some membrane proteins while others remain relatively unaffected. For example, the function of NMDA, GABA A and 5-HT a receptors (for reviews, see refs. 3, 9, 17, 36 and 48) as well as adenylyl cyclase -~'~''~'4~ seem implicated as specific targets of ethanol action.

Previous studies on the effects of ethanol on the GABA A receptor/chloride channel complex have reported variable and sometimes inconsistent results. While there are many reports that acute ethanol potentiates GABA g responses in various preparations 4'7'2s'4°'42, there are nearly as many studies that show no ethanol modulation of this response. Biochemical studies have generally shown the most consistent ethanol potentiation of GABAA-mediated responses, usually in the form of GABA- or muscimol-induced 360- influx. Eiectrophysiological studies have reported enhancement of GABAergic inhibition in cat cerebral cortex 2s, chick spinal cord 7, rat cerebellum 2m, rat dorsal root ganglion24'26 and rat medial septum and inferior colliculus ~m,4°. However, other electrophysi0logical studies, particularly in the hippocampus and lateral septum, have failed to observe such potent/ation 6'~2'~s'mp'zL22'27'2~'39"49.Several factors might be pro-

Correspondence: W.R. Proctor, Department of Pharmacology (C-236), University of Colorado Health Sciences Center, 4200 E. 9th Avenue, Denver, CO 80262, USA. Fax: (l) (303) 270-7097,

221 posed as responsible for the disparities between these studies concerning ethanol modulation. First, ethanol modulation of GABA responses could be brain region dependent. This hypothesis gains some support from the studies of Allan and Harris 2, in which they observed modulation of 36C1- flux in microsacs prepared from cerebral cortex and cerebellum, but not from hippocampus. Other investigators using electrophysiological techniques have also reported differences between brain regions in their responsiveness to ethanol and GABA 11'4°'49. A second possibility is that ethanol might modulate responses to exogenous GABA, but not to synaptically released endogenous GABA. The GABA responses examined in nearly all studies in which potentiation by ethanol has beee ?eported were evoked by exogenous GABA or muscimol, while most of the negative reports have focused on evoked synaptic responses. A further complication is suggested by the observation that exogenously applied GABA elicits a biphasic response, and only the initial, rapidly desensitizing component of the response is potentiated by ethanol 24. In order to resolve these issues, we investigated the ethanol sensitivity of GABAA receptor-mediated responses measured electrophysiologically, using an experimental protocol involving synaptic release of GABA, in a brain region where 36C!- flux studies have reported robust potentiation by ethanol. Thus, we recorded from pyramidal neurons in the cerebral cortex in rat brain slices, using local stimulation of inhibitory interneurons to evoke GABA A receptor-mediated responses. For comparison purposes, we conducted parallel studies of GABA A receptor-activated responses in the hippocampus, and related all of these results to a 6 0 - flux in microsac preparations from cerebral cortical and hippocampai tissues. MATERIALS AND METHODS Slice preparation Cerebral cortical brain slices were prepared from male SpragueDawley rats (110-150 g; Sasco, Omaha, NE). The brain was quickly removed from decapitated animals and chilled for 30-60 s in ice-cold oxygenated (by bubbling with 95% 0 2-5% CO 2 gas) artificial cerebral spinal fluid (aCSF (in mM) NaCI 124, KC! 3-5, MgCi 2 1.3, CaCI 2 2.5, KH2PO 4 1, NaHCO 3 25.7, D-glucose 11). The brain was blocked in a coronal plane (i) on the posterior end approximately 1-2 mm anterior to the cortex/cerebellum boundary and (ii) on the anterior side just at the cortex/striatum border. The anterior end and the posterior side was dried and glued using cyan~acrylate adhesive onto a stainless-steel pedestal mounted inside a plastic chamber, which was secured to the stage of a vibratome (Technical Products International, St. Louis, MO). The chamber was quickly filled with ice-cold aCSF and 400-v.m-thick slices were made and transferred to a glass petri dish containing cold aCSF. The slice was trimmed (part of the temporal cortex was removed and some of the slices were split down the midline) and transferred to a plastic-mesh basket submerged in a 250-ml beaker filled with continuously gassed

( 0 2 / C O 2) and warmed (32"C) aCSF. Following a recovery period of approximately 1 h, slices were transferred as needed onto nylon netting in a recording chamber (approximate volume: 1.0 ml) and continuously superfused with warm (32.5°C) oxygenated aCSF at 2.0 ml/min. Electrophysioiogical recording: cont'entionai intracellular and whole-cell patch Cortical pyramidal neurons located in layer V in the frontal cortex, area 1 and forelimb area were impaled with conventional intraceilular microelectrodes that were pulled (Flaming/Brown, model P-87, Suner Instrument Co., Novato, CA) from 1.2-mm capillary glass (SuRer Instrument Co.) containing a small diameter filament and filled with 2.5 M potassium acetate (resistances were 60-80 M,O). Whole-cell patch electrodes were pulled (using a 3-cycle program) from 1.5-ram capillary, filament glass (Glass Company of America, Millville, N J) and filled with a solution containing (in raM) KOH 130, gluconic acid 130, EGTA 1.0, MgC! 2 2.0, HEPES (free acid) 10.0, CaC! 2 0.5, ATP (di Na + ) 2.5 which was adjusted to pH 7.25 with KOH. Positive pressure was applied to the internal fluid in the electrode as it was lowered into the submerged slice. When an increase in resistance (monitored by observing a voltage change while passing a 0. l-nA, 10-ms duration current pulse through the recording electrode) was encountered (usually indicating close contact with a cell), negative pressure was applied. The formation of the seal was continuously monitored on the oscillascope. When a suitable seal was attained ( > 1 G~), the negative pressure was sharply increased briefly to rupture the membrane. An immediate drop in the potential ( - 6 5 to - 7 5 mV) was the initial requirement for a successful whole-cell recording. Excitatory postsynaptic potentials (EPSPs) and inhibitory postsynaptic potentials or currents (IPSPs or IPSCs) were elicited by local stimulation (monophasic, 0.1-0.2-ms pulses) either from the deeper layer VI or from the adjacent layer V area using bipolar twisted tungsten wire electrodes. Data analysis Responses from intracellular and whole-cell patch electrodes were digitized (16-channel A / D board, RC Electronics, Goleta, CA) and analyzed on-line using an 80286 computer and NeuroPro software (R.C. Electronics) developed in this laboratory. Results are presented as mean + standard error of the mean (S.E.M.) and statistical evaluations ",,ere made using the two-tailed Student's t-test. •~6C! - flltr determination The procedure used to measure GABA-induced chloride flux from cerebral cortex and hippocampus was similar to that described by Proctor et ai. 2'~. Briefly, rat brain tissue was homogenized in assay buffer (in mM: NaCI 145, KCI 5, MgCI 2 1, CaCI 2 1, HEPES 10, glucose 10) and washed twice by centrifugation (900x g, 15 rain). Chloride uptake was initiated by the addition of a solution containing 36C1- (2 /zCi/ml of assay buffer) and GABA (10 v.M), and terminated by dilution with cold assay buffer containing 100 /zM picrotoxin after 3 s. When present, the desired ethanol concentration was added to the tissue in the radioactive chloride solution.

RESULTS

Conventional intracellular recording In initial experiments on cortical neurons, we examined the effects of ethanol application on a number of cell membrane characteristics and on responses evoked by synaptic stimulation and intracellular current injection (Fig. 1, Table I). During superfusion with 80 mM ethanol, a consistent hyperpolarization (2-5 mV) of the resting membrane potential (RMP) was observed in cortical pyramidal neurons from layer V (Fig. 1). The

222 Control

80 mM EIOH

B

--J~

Washout

,

Fig. 1. Synaptic responses recorded with conventional intracellular electrodes. A: local stimulation (from the deeper layer V|) evokes a depolarizing EPSP (Control) in a cortical cell in layer V. Superfusion with 80 mM ethanol hyperpolarized the cell and reduced the EPSP amplitude (80 mM EtOH), and both effects returned to control values following a 30-40-min washout period (Washout). B: passing a depolarizing current pulse through the recording electrode elicits a train of action potentials followed by an after-hyperpolarizing potential (AHP). No significant effects of ethanol were observed on the amplitude of the AHP (action potentials appear truncated in these records). C: input resistance was measured by passing a hyperpolarizing current pulse through the recording microelectrode and measuring the resulting hyperpolarizing potential change. In some cells, a synaptic response was also evoked during the hyperpolarizing current pulse (illustrated). Ethanol did not sigmficantly alter the input resistance (C), but the amplitude of the EPSP was substantially reduced. Each record is from an individual response. Bars = 5 mV and 50 ms.

membrane input resistance (Ri.), determined by passing a constant hyperpolarizing pulse through the recording electrode, showed a small, but non-significant reduction during ethanol application (Fig. 1C, Table I). (It is unclear whether this small change in Rin is enough to explain the hyperpolarization of the RMP, or if other mechanisms such as alterations of electrogenic ion pumps are involved.) Injecting depolarizing current into cortical neurons produced a train of action potentials followed by a long lasting after-hyperpolarization (AHP; Fig. 1B). This after-hyperpolarization reflects the activation of a calcium-activated potassium channel by the depolarization 34,as,43. Superfusion of 80

mM ethanol did not significantly change this response (Table I). Electrical stimulation in adjacent layer V, the deeper layer VI or in the corpus callosum, usually elicited an EPSP and sometimes a very small IPSP when neurons were recorded at or near their resting membrane potential 4~ (Fig. 1A). Although we often observed a small reduction in the EPSP amplitude during ethanol superfusion (Table I, Fig. 1), this decrease was not statistically significant.

Whole-cell voltage-clamp recordings with patch electrodes We observed a small increase in outward current activated by ethanol superfusion (Table I), consistent with the hyperpolarizing effect on the cells seen with the conventional intraceUular recordings. Using either conventional intracellular electrodes or patch electrodes, it was usually difficult to demonstrate evoked IPSPs in response to local stimulation. A possible reason for this might be that the chloride equilibrium potential is close to the resting membrane potential in cortical cells. However, when cells were depolarized by constant current injection, a synaptically evoked hyperpolarizing potential following the EPSP was usually evident (Fig. 2A). Superfusion with 10 /zM 6,7-dinitroquinoxaline-2,3-dione (DNQX; Tocris Neuramin, Essex, UK) and 40/zM o-(-)-2-amino-5-phosphonovaleric acid (APV; Sigma), left a clear uniphasic response as shown in Fig. 2B. Under voltage clamp conditions, evoked synaptic currents were observed that were similar in the time course to the voltage responses measured under current clamp mode (Fig. 2C). As with the voltage responses, DNQX and APV superfusion changed the current response from a primarily biphasic one to a uniphasic response that was

TABLE !

Control calues and ethanol effects on conical neurons Concentional recording

Whole.cell recording

RMP

R~n

EPSP

AHP

lhoid

Ri,

GABA A

Control Mean S.E,M. n

-63.0 mV + 1.7 9

22.1 Mr/ -+ 3.2 9

11.1 mV + !.6 9

1.9 mV +0.5 7

229 pA _+41 7

105 M~(~ .~.6.2 '15

68 pA _+.16 7

80 mM EtOH Mean S.E,M. n

- 66.6 mV + 1.4 9

20.7 MD -+ 2.9 9

9.5 mV _+ ! .6 9

1.6 mV 4- 0.4 7

252 pA +_45 7

Not tested

86 pA -+ 17 7

94% n.s.

86% n.s.

84% n.s.

% Control P *