Ž .Brain Research 798 1998 330–332

Short communication

GABA-evoked depolarisations in the rat cortical wedge: involvement ofGABA receptors and HCOy ionsA 3

Ian Phillips, Keith F. Martin 1, Kevin S.J. Thompson ), David J. HealCNS Biology, Knoll Pharmaceuticals Research & DeÕelopment, Nottingham NG1 1GF, UK

Accepted 28 April 1998

Abstract

Ž .The effect of g-aminobutyric acid GABA was investigated on cortical wedges prepared from male Sprague–Dawley rats. GABAŽ .evoked concentration-dependent depolarisations EC : 0.8 mM , which were attenuated by up to 60% when bicarbonate-buffered aCSF50

Ž .was replaced with HEPES-buffered aCSF. Responses to 1 mM GABA were attenuated by y -bicuculline and picrotoxin and werepotentiated by chlordiazepoxide and pentobarbitone. Ionotropic glutamate receptor antagonists had no effect. We conclude thatGABA-evoked depolarisations are mediated via GABA receptors, arising in part from HCOy efflux from cells. q 1998 ElsevierA 3

Science B.V. All rights reserved.

Keywords: g-Aminobutyric acid; Bicarbonate; Electrophysiology; Grease-gap

Ž .g-Aminobutyric acid GABA is the major inhibitoryneurotransmitter in the mammalian CNS, acting predomi-nantly via two of three known receptor subtypes, GABAA

w xand GABA 6 . GABA receptors form a ligand-gatedB A

Cly ionophore with modulatory binding sites for benzodi-w xazepines, barbiturates, neurosteroids and alcohol 4,10,12 .

Activation of postsynaptic GABA receptors was classi-A

cally shown to mediate an inward Cly conductance, hyper-w xpolarising cells and inhibiting neuronal activity 9 . More

recently however, GABA has been reported to exert excita-tory actions on neurones in brain slices in vitrow x1,3,5,7,13,15 . Conditions when GABA is excitatory mayhave as a common feature an altered transmembrane Cly

gradient, such that the GABA receptor gated Cly fluxA

will be outward rather than inward, and therefore depolar-ising. However, because the GABA ionophore is alsoA

selectively permeable to HCOy, albeit five times less than3y w x yfor Cl 2 , the involvement of HCO ions in GABA3

depolarisations cannot be excluded. We have thereforeused the rat cortical wedge preparation to investigate thecontribution of bicarbonate ions to GABA depolarisations.

Cortical wedges were prepared from coronal sectionsŽ .500 mm thick of Sprague–Dawley rat brain, essentially

) Corresponding author. Fax: q44-115-912-4169.1 New address: BTG plc, 10 Fleet Place, London EC4M 7SB, UK.

w xas described in Phillips et al. 11 . Wedges were perfusedŽ . Žwith gassed 95% O r5% CO aCSF composition in2 2

mM: NaCl 124, KCl 3.3, KH PO 1.2, NaHCO 25.5,2 4 3.MgSO 1.0, D-glucose 10.0, CaCl 2.5 at a rate of 24 2

mlrmin. For bicarbonate-free aCSF, NaHCO was re-3

placed with 10 mM HEPES and was gassed with 100%O . Solutions were maintained at room temperature. The2

D.C. potential between the two compartments of therecording chamber was measured with AgrAgCl elec-trodes using AC–DC amplifiers, filters and a MacLab dataacquisition system on a Power Macintosh computer.

All drugs were dissolved at stock concentrations of1=10y2 M or higher in appropriate solvents, with dilu-tions being made in gassed aCSF. All the drugs wereapplied to the cortical side of the preparation only. GABAwas applied for 1 min with 15–20 min between applica-tions. An increasing series of GABA concentrations wasused to determine the concentration giving 50% of maxi-

Ž .mum response EC . This EC concentration of GABA50 50

was then used to determine the effects on the agonistresponse of an increasing concentration series of GABAantagonists, GABA modulators or ionotropic glutamatereceptor antagonists. These were applied 10 min beforeGABA applications and continued for 10 min after with-drawal of GABA.

Depolarisation amplitude was used to quantitate agonistresponses. The treatment-induced response was converted

0006-8993r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved.Ž .PII S0006-8993 98 00479-X

( )I. Phillips et al.rBrain Research 798 1998 330–332 331

Ž .to a percent of control mean pre-intervention response .Data were log transformed then analysed by two-wayanalysis of variance. Comparisons against control weremade using Williams’ test. Data are reported as mean %

Ž .control"standard error S.E.M. .Ž .GABA application 0.1–30 mM caused concentration-

Ž .dependent depolarisations Fig. 1A . From the concentra-Ž . Žtion–response curve generated Fig. 1B , an EC 95%50

. Ž .confidence limits of 0.8 mM 0.5, 1.1, ns3–4 wasŽ . Ž .calculated. Application of " -baclofen 0.1–30 mM did

not evoke similar depolarisation responses.Replacing bicarbonate-buffered aCSF with HEPES-

buffered aCSF, resulted in up to 60% attenuation of theŽ . ŽGABA-evoked depolarisations Fig. 1B . The EC 95%50

.confidence limits for GABA in HEPES-buffered aCSFŽ .was 0.9 mM 0.5, 1.6, ns4 , not significantly different to

the value in HCOy-buffered aCSF.3Ž .The selective GABA receptor antagonists y -bi-A

cuculline and picrotoxin concentration-dependently attenu-ated responses evoked by 1 mM GABA application. IC50

Ž . Žvalues were: 95% confidence limits 13.0 mM 5.4, 31.4,. Ž . Ž .ns4 for y -bicuculline and 28.3 mM 21.5, 37.3, ns4

Ž .for picrotoxin Fig. 2A . However, application of theŽ .GABA receptor antagonist, CGP 35348 0.1–30 mMB

Fig. 1. Effect of exogenous applications of GABA in the rat corticalŽ .wedge preparation. A Traces showing GABA-induced depolarisations.

Ž . Ž .B Depolarisation responses recorded in bicarbonate- –v – andŽ . wHEPES-buffered –B– aCSF. ) Indicates significant difference be-

tween bicarbonate and HEPES responses at the 5% level, )) indicatesxsignificance at the 1% level.

Ž . Ž . Ž . Ž .Fig. 2. A Effects of y -bicuculline –v – and picrotoxin –B– onŽ . wresponses to GABA 1 mM . Picrotoxin: † significant at 5%, ††† at

x Ž .0.1%. Bicuculline: )) significant at 1%, ))) at 0.1%. B Effects ofŽ . Ž .chlordiazepoxide –v – and pentobarbitone –B– on responses to

Ž . wGABA 1 mM . Chlordiazepoxide: ) significant at 5%, )) at 1%.xPentobarbitone: † significant at 5%, ††† at 0.1%.

Žwas without effect on responses to 1 mM GABA data not.shown .

Responses to 1 mM GABA were significantly potenti-Ž .ated by the benzodiazepine, chlordiazepoxide 33% and

Ž . Žby the barbiturate, pentobarbitone 42% at 100 mM Fig..2B . Chlordiazepoxide showed a concentration-dependent

effect at 0.3–3 mM, with the effect plateauing at higherconcentrations, while pentobarbitone showed a concentra-tion dependency at 10 mM or higher. This reflects therelative affinities of these compounds at their respectivesites of action.

The selective NMDA receptor antagonists MK-801 andAP-7; the selective AMPA receptor antagonists NBQXand CNQX; and MNQX, an antagonist at both NMDA andAMPA receptors, were all without effect on responses to

Ž .GABA at concentrations up to 100 mM data not shown .This study demonstrates that GABA evokes concentra-

tion-dependent depolarisations in rat cortical wedges, whichare inhibited by GABA receptor antagonists and potenti-A

ated by GABA receptor modulators, confirming that theyA

are mediated by GABA receptors. The selective GABAA BŽ .receptor agonist " -baclofen did not evoke similar depo-

larisations and the GABA receptor antagonist, CGPB

35348, had no effect on GABA depolarisations, showingthat GABA receptors are not involved in this response.B

( )I. Phillips et al.rBrain Research 798 1998 330–332332

Although the more usual action of GABA is hyperpolar-isation, mediated by an inward flux of Cly ions through

w xthe GABA receptor channel 9 , a number of studies haveA

shown that under various conditions, GABA can exertw xexcitatory actions on neurones in vitro 1,3,5,7,13,15 .

Because the GABA ionophore is also permeable to HCOyA 3

w xions 2 , we investigated the ionic mechanisms underlyingGABA depolarisations, by replacing HCOy-buffered aCSF3

with HEPES-buffered aCSF. This resulted in attenuation ofthe GABA-evoked depolarisations by up to 60%, a finding

w xin agreement with Grover et al. 5 . Therefore, bicarbonateefflux comprises a significant proportion of the depolaris-ing current.

Ž .It has been suggested that the intracellular pH pH ini

excitable cells is higher than would be predicted fromq w xpassive distribution of H ions 14 , elevating intracellular

HCOy. It is also likely that partial anoxia during prepara-3

tion of slices will increase CO concentration in the brain2

tissue. This will tend to increase further intracellular pH byformation of HCOy. Thus, activation of GABA receptors3 A

in the presence of increased intracellular HCOy would3

lead to a significant HCOy efflux. Therefore, GABA-gated3

HCOy conductance has a direct effect on E , as3 GABAw xproposed by Kaila 8 . The GABA-induced depolarisation

effect would be sustained because, the efflux of HCOy3

through GABA receptors would cause a decrease in theA

concentration of intracellular HCOy, and would therefore3

drive the diffusion of CO across the membrane with the2

resulting regeneration of intracellular HCOy by carbonic3

anhydrase.The incomplete attenuation of the depolarisation re-

sponse in HEPES buffer suggests that GABA-evoked de-polarisations are not solely HCOy efflux dependent. How-3

ever, the similarity of EC values in the presence and50

absence of bicarbonate confirm that a single receptor type,GABA , is responsible. The remaining component of theA

GABA-evoked depolarisation seems likely to be due toCly efflux as a consequence of increased intracellular Cly.This may be attributable to Cly entry into the cell duringthe preparation of wedges, or a retarded Cly transporterpassing Cly back out of cells. Indeed, the high intra-cellular HCOy would tend to resist exchange of intra-3

cellular Cly for extracellular HCOy, leading to an in-3

creased intracellular Cly, and subsequently a change inE .GABA

In conclusion, we have shown that GABA producesdepolarising responses in rat cortical wedges, which aremediated via GABA receptors. The GABA-evoked depo-A

larisation arises from HCOy efflux from cells, and may3y y Žalso include Cl efflux. High intracellular HCO and3

y.possibly Cl are therefore a requirement for the depolari-

sation effect. It is not clear whether these conditions occurŽonly under conditions of stress e.g., intense GABAA

.activation, heat shock, trauma , or will be present innormal brain tissue.

Acknowledgements

ŽWe would like to thank Dr. J.A. Davies University of.Wales, Cardiff for helpful discussion in the course of this

work.

References

w x1 M. Avoli, P. Perreault, A GABAergic depolarizing potential in thehippocampus disclosed by the convulsant 4-aminopyridine, Brain

Ž .Res. 400 1987 191–195.w x2 J. Bormann, O.P. Hamill, B. Sakmann, Mechanism of anion perme-

ation through channels gated by glycine and g-aminobutyric acid inŽ .mouse cultured spinal neurones, J. Physiol. 385 1987 243–286.

w x3 E. Cherubini, J.L. Gaiarsa, Y. Ben-Ari, GABA: an excitatory trans-Ž .mitter in early postnatal life, Trends Pharmacol. Sci. 14 1991

515–519.w x4 K.J. Child, J.P. Currie, B. Davis, M.G. Dodds, D.R. Pearce, D.J.

Twissell, The pharmacological properties in animals of CT1341—aŽ .new steroid anaesthetic agent, Br. J. Anaesth. 43 1971 2–13.

w x5 L.M. Grover, N.A. Lambert, P.A. Schwartzkroin, T.J. Teyler, Roleof HCOy ions in depolarising GABA receptor-mediated responses3 A

Ž .in pyramidal cells of rat hippocampus, J. Neurophysiol. 65 19931541–1555.

w x 3 36 D.R. Hill, N.G. Bowery, H-baclofen and H-GABA bind to bicu-Ž .culline-insensitive GABA sites in rat brain, Nature 290 1981B

149–152.w x7 R.Q. Hu, J.A. Davies, Tiagabine hydrochloride, an inhibitor of

Ž .g-aminobutyric acid GABA uptake, induces cortical depolarisa-Ž .tions in vitro, Brain Res. 753 1997 260–268.

w x8 K. Kaila, Ionic basis of GABA receptor channel function in theAŽ .nervous system, Prog. Neurobiol. 42 1994 489–537.

w x9 K. Krnjevic, S. Schwartz, The action of g-aminobutyric acid onŽ .cortical neurones, Exp. Brain Res. 3 1967 320–336.

w x10 R.W. Olsen, GABA–benzodiazepine–barbiturate receptor interac-Ž .tions, J. Neurochem. 37 1981 1–13.

w x11 I. Phillips, K.F. Martin, K.S.J. Thompson, D.J. Heal, Weak blockadeof AMPA receptor-mediated depolarisations in the rat cortical wedgeby phenytoin but not lamotrigine or carbamazepine, Eur. J. Pharma-

Ž .col. 337 1997 189–195.w x12 R.F. Schmidt, E. Vogel, M. Zimmermann, Die Wirkung von Di-

azepam auf die prasynaptische Hemmung und andere Ruckenmark-Ž .sreflexe, Naunyn-Schmiedeberg’s Arch. Pharmacol. 258 1967 69–

82.w x13 A.C. Shakesby, J.A. Davies, The GABA uptake inhibitor, NO-711,

induces depolarisations in mouse cortical wedges, Br. J. Pharmacol.Ž .120 1997 366.

w x14 R.C. Thomas, The role of bicarbonate, chloride and sodium ions inthe regulation of intracellular pH in snail neurones, J. Physiol.Ž . Ž .London 273 1977 317–338.

w x15 A.N. van den Pol, K. Obrietan, G. Chen, Excitatory actions ofŽ .GABA after neuronal trauma, J. Neurosci. 16 1996 4283–4292.