Brain Research, 516 (1990) 147-150 Elsevier

147

BRES 24072

Physiological modulation of the GABA receptor by convulsant and anticonvuIsant barbiturates in cultured rat hippocampal neurons Katherine D. Holland, Daniel J. Canney, Steven M. Rothman, J.A. Ferrendelli and

Douglas F. Covey Departments of Pharmacology and Anatomy and Neurobiology, Washington University School of Medicine, St. Louis, MO 63110 (U.S.A.)

(Accepted 23 January 1990) Key words: Barbiturate; y-Aminobutyric acid; Convulsant; Voltage-clamp; Inhibitory synapse

The actions of convulsant and sedative barbiturates on responses to y-aminobutyric acid (GABA) application and on inhibitory postsynaptic currents were compared using voltage-clamp techniques in cultured rat hippocampal neurons. The convulsant barbiturates, 5-ethyl5-(3-methylbut-2-enyl) barbituric acid (3M2B), and (+)-5-ethyl-5-(1,3-dimethylbutyl) barbituric acid ((+)-DMBB), and the sedative barbiturate, 5-ethyl-5-(3-methylbutyl)barbituric acid (3MB), all potentiated GABA-mediated chloride currents. In addition, these compounds prolonged the duration of GABAergic inhibitory postsynaptic currents. The similarity between the action of convuisant and sedative barbiturates suggests that the convulsant activity of 3M2B and (+)-DMBB are not mediated by their actions at GABAergic synapses. The y-aminobutyric acid ( G A B A ) receptor/chloride ionophore complex ( G A B A A receptor) possesses binding sites for diverse modulatory agents including G A B A , benzodiazepines/fl-carbolines, picrotoxin, barbiturates and steroids 7. Structurally related compounds binding to the benzodiazepine site on the G A B A A receptor have been shown to have the unique ability to elicit opposing neurological and electrophysiological effects 7. Recently, the picrotoxin site on the G A B A A receptor was also reported to be capable of producing opposite effects on GABA-mediated chloride currents 3'5'8'1°. For example, convulsant agents such as picrotoxin and fl-alkyl-ybutyrolactones were shown to antagonize G A B A responses while anticonvulsant a-alkyl-y-thiobutyrolactones potentiated GABA-induced currents. Based on these observations, other modulatory sites on the GABA A receptor might also be anticipated to have similar properties. The action of sedative barbiturates has been attributed to interactions with the barbiturate site located on the G A B A A receptor 7'12'15. Minor structural alterations convert sedative barbiturates into potent convulsant agents. For example, the anesthetic barbiturate 5-eth¢l-5-(3methylbutyl) barbituric acid (amobarbital; 3MB) differs from the convulsant barbiturate 5-ethyl-5-(3-methylbut2-enyl) barbituric acid (3M2B) by the presence of a double bond in the side chain (Fig. 1) 2. Introduction of a methyl group on the first carbon atom in the side chain

of amobarbital (3MB) produces an optically active compound, 5-ethyl-5-(1,3-dimethylbutyl) barbituric acid (DMBB), whose isomers have opposing neurological effects 6. Thus, ( + ) - D M B B is a convulsant while (-)DMBB is an anesthetic. In order to investigate the possibility that structurally related convulsant and anticonvulsant barbiturates act at the same site on the G A B A A receptor, their effects on G A B A responses and inhibitory postsynaptic currents were compared in voltage-clamped rat hippocampal neurons grown in culture. Hippocampal cells were cultured from 1-day-old female Sprague-Dawley rat pups by previously described methods 8'21. Electrophysiological experiments were carried out on the stage of an inverted microscope (Leitz) using whole-cell patch clamp techniques on cells that had been in culture for 10-21 days. The growth medium was removed and the cells were placed in an extracellular recording solution which contained (in raM): 140 NaCI, 10 NaHEPES (pH 7.3), 3 KCI, 4 CaCI2, 4 MgC12, and 5.5 glucose. Electrodes (4-10 Mr2 resistance) were filled with an intracellular recording solution that contained (in mM): 138 Klsethionate (Kodak), 2 KCI, 10 NaHEPES (pH 7.3), 2 MgATP, and 1.1 EGTA. Neurons were voltage-clamped at -50 mV using a standard patch amplifier (Dagan) and exposed to the test compounds by applying 100 ms air pressure pulses (10 psi) to micropipettes filled with either G A B A or a mixture of G A B A and drug dissolved in extraceUular solution. To examine

Correspondence: D.E Covey, Department of Pharmacology, Washington University School of Medicine, 4566 Scott Ave., St. Louis, MO 63110, U.S.A.

0006-8993/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)

148 CONVULSANTS:

ANTICONVULSANT:

o

N.-% /

NH'--~O

GABA+ 30/zM (+)DMBB

/

O o

~

A

O

I

5-Ethyl-5-(3-Methylbut-2-enyl) Barblturlc Acid (3M2B)

I

O

5-Ethyl-5-(3-Methylbutyl) Barblturlc Acid (3MB)

/

CH2CH3

(+)-5-Ethyl-5-(1,3-Dlmethylbutyl) Barbiturlc Acid ((+)-OMBB)

Fig. 1. Chemical structures of 3MB, 3M2B and (+)-DMBB.

the effects of various barbiturates on GABAergic synapses, pairs of neurons with inhibitory connections were also studied. This was accomplished by recording from neighboring neurons with 2 whole-cell patch electrodes which were connected to our voltage-clamp circuit and a standard bridge amplifier. The postsynaptic current was considered inhibitory if stimulation of the presynaptic cell (through the bridge circuit) produced an outward current which was diminished by bicuculline and was insensitive to the application of 6-cyano-7-nitroquinoxaline-2,3dione (CNQX), an antagonist at glutamate receptors 9'21. Barbiturates were dissolved in extracellular recording solution and applied by bath perfusion of the culture dish. The data were digitized at 0.5 KHz and stored on

Fig. 2. (+)-DMBB potentiates GABA responses and prolongs GABAergic synaptic currents. A: outward current produced by the application of 30/aM G A B A is increased in the presence of 30/aM (+)-DMBB in a neuron voltage-clamped at -50 inV. Arrow indicates the beginning of the 100 ms drug application. Calibrations: 65 pA, 250 ms. B: IPSCs elicited before, during, and after bath perfusion with 30/aM (+)-DMBB. Arrow indicates the peak of the presynaptic action potential. Calibrations: 20 pA, 25 ms.

disc for off-line analysis. Inhibitory postsynaptic current (IPSC) decay time constant, peak amplitude, and total charge transfer were determined using a commercially available analysis program (pCLAMP). 3M2B and 3MB were synthesized according to previously described methods 4. The compounds exhibited satisfactory spectroscopic data 19 and produced the previously reported behavioral effects2,6. ( + ) - D M B B was

A

- 600

O Control [] 30/~M(+)OMBB

- 400

/~ 300/zM(+)DMBB

- 200 -90

-80

-70 ~ -50

-40

0 •

-

A comparison of the effects of 3MB, 3M2B, and (+)-DMBB on GABA currents and inhibitory postsynaptic currents (IPSCs)

~t. _q

'E

- -200 •

TABLE I

r~ 5

-400

L_ j (3

- -600

Membrane Potential (mV)

Data are mean + S.E.M.; the number of experiments is given in parentheses.

Compound

Response to 30/aM GABA

3 MB (30/aM)

99 + 3

3 MB (100/aM)

165 + 14 (6)**

B

% of control Peakwsc

Tzesc

Qresc

198+14

182+17

(3) 3M2B (30/aM)

94 + 5 (3)

3M2B (100/aM)

150 + 9

(6)**

(+)-DMBB (30 /aM)

136 + 10 (6)*

93+8 (4)

104+5 (6) 101 + 9 (5)

(4)**

(4)*

171+_12 167+-6 (6)** (6)** 303 + 36 (5)**

*P < 0.02; **P < 0.01 compared to control, Student's t-test.

364 + 58 (5)*

~ Control

t 600

30#M (+)DMBB

300#M(+)DMBB --90 ~

--80

--70

1400 ~

2000 -5o

'

_

-E

-4o

-200

-400 MembranePotential (rnV)

I_ o

-600

Fig. 3. High concentrations of (+)-DMBB are OABAmimetic. A: the I-V relationship is altered by 300/aM but not 30/aM (+)-DMBB. B: the effect of 300/aM (+)-DMBB on the I-V curve is blocked in the presence of 100/aM bicuculline.

149 provided by Eli Lilly (Indianapolis, IN), and C N Q X was purchased from Tocris Neuramin (Essex, U.K.). Stock solutions of the barbiturates were diluted in 0.1 N N a O H and diluted 1:100 for the electrophysiological experiments. Control solutions also contained 0.001 N NaOH. Unless noted, all other reagents were purchased from Sigma Chemical (St. Louis, MO). Both the convulsant barbiturates (3M2B and (+)DMBB), and the sedative barbiturate (3MB) potentiated currents produced by the application of 30/~M G A B A (Table I). Of the compounds examined, (+)-DMBB was the most potent. It augmented G A B A responses at concentrations as low as 30/~M (Fig. 2A). The other two compounds were inactive at 30/~M but did potentiate G A B A currents at 100/~M. The direct membrane effects of ( + ) - D M B B were also investigated by studying the current-voltage (I-V) relationships of these cells near the resting membrane potential alone and in the presence of ( + ) - D M B B . Thirty micromolar ( + ) - D M B B had no effect on the I-V relationship (n = 3), but 300 p M ( + ) - D M B B applied to cells held at a membrane potential o f - 5 0 mV produced an outward current and increased the slope of the I-V curve (n = 3; Fig. 3A). The direct action of ( + ) - D M B B on the I-V curve was blocked by the addition of 100 /~M bicuculline (a G A B A receptor antagonist) to the extracellular solution (n = 3; Fig. 3B). The effects of the barbiturates on GABAergic synapses were also examined by testing the actions of the barbiturates on the amplitude, decay time constant (riesc) and charge transfer (QiPsc) of IPSCs. Bath perfusion of 3MB (100 pM), 3M2B (100 pM), and ( + ) - D M B B (30 pM) all prolonged the exponential decay of IPSCs and increased Q~Psc without altering IPSC peak amplitude (Fig. 2B). The effects of ( + ) - D M B B on excitatory postsynaptic currents (EPSCs) was also studied. In contrast to its effects on IPSCs, 3 0 p M (+)-DMBB had no significant effect on EPSCs (n = 3). It is important to note that not all convulsant barbiturates share a common site of action ~6. The convulsant 1 Allan, A.M. and Harris, R.A., Anesthetic and convulsant barbiturates alter ~'-aminobutyric acid-stimulated chloride flux across brain membranes, J. Pharmacol. Exp. Ther., 238 (1986) 763-768. 2 Andrews, P.R., Jones, G.P. and Lodge, D., Convulsant, anticonvulsant and anaesthetic barbiturates. 5-Ethyl-5-(3"-methylbut-2"-enyl)-barbituric acid and related compounds, Eur. J. Pharmacol., 55 (1979) 115-120. 3 Baker, K., Yang, J., Covey, D.E, Clifford, D.B. and Zorumski, C.F., a-Substituted thiobutyrolactones potentiate GABA currents in voltage-clamped chick spinal cord neurons, Neurosci. Len., 87 (1988) 133-138. 4 Cain, C.K. and Kleis, J., 5-Ethyl-(1,3-dimethyi-2-butenyl)barbituric acid: a potent central nervous system stimulant, J. Med. Pharm. Chem., 1 (1959)31-36. 5 Clifford, D.B., Baker, K., Yang, J., Covey, D.E and Zorumski, C.E, Convulsant gamma-butyrolactones block GABA currents

compound S-(+)-l-methyl-5-phenyl-5-propyl barbituric acid (S-(+)-MPPB) is thought to produce convulsions by diminishing G A B A responses, not by binding at the barbiturate receptor, but by modulation of the picrotoxin site on the G A B A A receptor complexl'17. Because of this, we have restricted our studies to barbiturates thought to interact with the barbiturate site and not the picrotoxin site TM. Previous reports concerning the action of convulsant barbiturates on GABA-mediated processes produced conflicting results L2,1L~3A4,2°. However, none of these experiments examined physiological modulation of G A B A responses in voltage-clamped neurons. The convulsant barbiturates studied here potentiate the GABAA complex in a manner identical to those of the anesthetic barbiturates. Anesthetic barbiturates (e.g. amobarbital and pentobarbital) are known to increase the magnitude of responses to exogenous application of G A B A , prolong the duration of inhibitory synaptic events, and, at high concentrations, produce GABA-Iike openings of the chloride ionophore 12'~5. Here we have shown that 3M2B and ( + ) - D M B B also increase G A B A response and prolong the duration of IPSCs. In addition, high concentrations of ( + ) - D M B B also open the chloride ionophore in the absence of G A B A . The similarity between the action of convulsant and anesthetic barbiturates suggests that the convulsant activity of 3M2B and ( + ) - D M B B are not mediated by their actions at GABAergic synapses. This similarity may explain why convulsant barbiturates have anticonvulsant actions against pentylenetetrazol seizures 2. Our results suggest that barbiturates are not capable of modulating G A B A responses in opposing ways via a common receptor site. We would like to thank Nancy Lancaster for culture preparation and Kelvin Yamada for help in preparation of the manuscript. Assistance was also provided by the Washington University High Resolution NMR facility (supported in part by NIH 1 S10 RR00204 and a gift from the Monsanto Co.). This work was supported by NIH Grant GM 07805, NS 07129, NS 14834 and NS 19988.

6 7

8

9

in cultured chick spinal cord neurons, Brain Research, 484 (1989) 102-110. Dowries, H., Perry, R.S., Ostland, R.E. and Karler, R., A study of the excitatory effects of barbiturates, J. Pharmacol. Exp. Ther., 175 (1970) 692-699. Haefely, W. and Polc, P., Physiology of GABA enhancement by benzodiazepines and barbiturates. In R.W. Olsen and C.J. Venter (Eds.), Benzodiazepine/GABA Receptor and Chloride Channels: Structural and Functional Properties, Alan Liss, New York, 1986, pp. 97-133. Holland, K.D., Ferrendelli, J.A., Covey, D.F. and Rothman, S.M., Physiological regulation of the picrotoxin receptor by y-butyrolactones and 7-thiobutyrolactones in cultured hippocampal neurons, J. Neurosci., in press. Honor6, T., Davies, S.N., Drejer, J., Fletcher, E.J., Jacobsen, P., Lodge, D. and Nielsen, EE., Quinoxalinediones: potent competitive non-NMDA glutamate receptor antagonists, Sci-

150 ence, 241 (1988) 701-703. 10 Klunk, W.E., Kalman, B.L., Ferrendelli, J.A. and Covey, D.E, Computer-assisted modeling of the picrotoxinin and ~,-butyrolactone receptor site, Mol. Pharmacol., 23 (1983) 511-517. 11 Leeb-Lundberg, E and Oisen, R.W., Interactions of barbiturates of various pharmacological categories with benzodiazepine receptors, Mol. Pharmacol., 21 (1982) 320-328. 12 Macdonald, R.L. and Barker, J.L., Anticonvulsant and anesthetic barbiturates: differential postsynaptic actions in cultured mammalian neurons, Neurology, 29 (1979) 432-447. 13 Nicholson, G.M., Spence, I. and Johnston, G.A.R., Differentiating the actions of convuisant and nonconvulsant barbiturates: an electrophysiological study in the isolated spinal cord of the rat, Neuropharmacology, 27 (1988) 459-465. 14 Nicholson, G.M., Spence, I. and Johnston, G.A.R., Depolarizing actions of convulsant barbiturates on isolated rat dorsal root ganglion cells, Neurosci. Lett., 93 (1988) 330-335. 15 Shultz, D.W. and Macdonald, R.L., Barbiturate enhancement of GABA-mediated inhibition and activation of chloride ion conductance: correlation with anticonvulsant and anaesthetic actions, Brain Research, 209 (1981) 177-188.

16 Skerritt, J.H. and Macdonald, R.L., Multiple actions of convulsant barbiturates on mouse neurons in cell culture, J. Pharmacol. Exp. Ther., 230 (1984) 82-88. 17 Ticku, M.K., Rastogi, S.K. and Thyagarajan, R., Separate site(s) of action of optical isomers of 1-methyl-5-phenyl-propylbarbituric acid with opposite pharmacological activities at the GABA receptor complex, Eur. J. Pharmacol., 112 (1985) 1-9. 18 Triffiletti, R.R., Snowman, A.M. and Snyder, S.H., Barbiturate recognition site on the GABA/benzodiazepine receptor complex is distinct from the picrotoxinin/TBPS recognition site, Eur. J. Pharrnacol., 106 (1985) 441-447. 19 Walton, H., Doezi, J. and King, J.A., Barbiturates containing the 3-methyl-2-butenyl group, J. Am. Chem. Soc., 72 (1950) 4319-4320. 20 Willow, M. and Johnston, G.A.R., Enhancement by anesthetic and convulsant barbiturates of GABA binding to rat synaptosomal membranes, J. Neurosci., 1 (1981) 364-367. 21 Yamada, K.A., Dubinsky, J.M. and Rothman, S.M., Quantitative physiological characterization of a quinoxalinedione nonNMDA receptor antagonist, J. Neurosci., 9 (1989) 3230-3236.