Modulation of the GABAA-like autoreceptor by barbiturates but not by steroids

Neuropharmacology Vol. 32, No. 4, pp. 355-357, 1993 Printed in Great Britain. All rights reserved

0028-3908/93 $6.00+ 0.00 Copyright © 1993 Pergamon Press Ltd

MODULATION OF THE GABAA-LIKE AUTORECEPTOR BY BARBITURATES BUT NOT BY STEROIDS CHRISTINE ENNIS and M. C. W. M1NCHIN Wyeth Research (UK) Ltd, Huntercombe Lane South, Taplow, Maidenhead, Berkshire SL6 0PH, U.K. (Accepted 26 November 1992)

Summary--In slices of cerebral cortex from rat preloaded with [3H]GABA, muscimol produced a concentration-related inhibition of K +-evoked release of tritium with a pIC25 value of 7.80 + 0.39. Dimethylbarbituric acid (I0 and 100 #M) and pentobarbitone (100 #M) significantly increased this value to 8.31 + 0.09, 9.91 + 0.21 and 8.50 + 0.21, respectively, whereas the steroid ligands alphaxalone (1 t~M) and 5/Lpregnane-3ct-ol-20-one (10nM) had no significant effect. The 5fl-pregnane-3~t-ol-20-one and 5fl-pregnane-3,20-dione produced a concentration-related increase in K ÷ -evoked release of tritium alone. These data suggest that the GABAA-Iike autoreceptor may be modulated by barbiturates but not by steroids and thus may be different from the postsynaptic GABAA receptor. Key words--GABA autoreceptor, barbiturate, steroid, modulation.

The GABAA receptor is now known to belong to the growing family of ligand-gated ion channels and is composed of a number of different subunits, the exact stochiometry of which is still under investigation (for review see Olsen and Tobin, 1990). Contained within the subunits are the allosteric modulatory sites for benzodiazepines, barbiturates and steroids which, together with the G A B A recognition site, control the opening of a chloride ion channel. Although there remains some controversy about the nature of the G A B A autoreceptor (Waldmeier and Baumann, 1990), it was previously reported that the autoreceptor which modulates the release of [aH]GABA from slices of cerebral cortex from the rat is sensitive to the G A B A receptor ligands muscimol and bicuculline, under the conditions described. Furthermore, the response of this receptor to muscimol was enhanced by benzodiazepine receptor agonists and attenuated by inverse agonists (Ennis and Minchin, 1988). These data suggest that the autoreceptor may be similar to a classical G A B A A receptor, although agonist potencies and antagonist affinities indicate pharmacological differences (Minchin, Ennis, Lattimer, White, White and Lloyd, 1992). The present study was undertaken in order to investigate if the allosteric sites for barbiturates and steroids are also present in this G A B A autoreceptor. A preliminary account of this work has previously been presented (Ennis and Minchin, 1989).

METHODS

For each experiment, a male rat (Sprague-Dawley, 180-200g) was killed by stunning followed by cervical dislocation and decapitation. The brain was

removed and the cerebral cortex was dissected and chopped into 0.25 x 0.25 x 2 m m slices using a Mcllwain tissue chopper. The slices of cerebral cortex were incubated at 37°C in Krebs'-Henseleit buffer, gassed with 5% carbon dioxide in oxygen and containing amino-oxyacetic acid (10#M) and [3H]GABA (0.1 #M, 3/~Ci ml-l). After 20min, the slices were washed three times and equal aliquots, equivalent to approximately 40 mg of tissue, were placed in each of 20 superfusion chambers (volume 300 #1). The tissue slices were superfused with previously oxygenated Krebs'-Henseleit solution containing amino-oxyacetic acid (10#M), at a rate of 0.4 ml min -t and were maintained at 37°C by means of a heated water jacket. Fractions of the superfusate were collected every 4 min and the radioactivity in each fraction, together with that remaining in the tissue at the end of the experiment, was determined by liquid scintillation counting. Two pulses of Krebs'-Henseleit buffer containing 25 mM KC1, obtained by iso-osmotic replacement of NaCI, were administered for 4 min at 68 (St) and 92 ($2) min after the start of the superfusion. Test agents were added to the superfusing medium immediately after St. The percentage eiflux of tritium was calculated as the fractional release. Basal or spontaneous efflux of [3H]GABA was taken as the percentage efflux occurring immediately prior to St and to $2. The total percentage of radioactivity, released above the basal values by the raised potassium ion concentration, was expressed as the ratio SJSI for both control and drug-treated slices. The 52/81 ratio was then expressed as a percentage of the control S2/St ratio from the same experiment. Statistical analysis of the results was performed using the Mann-Whitney U-test, 2-tailed. 355

356

CHRISTINE ENNIS

and M. C. W. MINCHIN B 50 -

A

I.

< <

70

--

60

--

40 -

50 30 -

40 ~z 30 o

20

-

10

-

d~

0

-12

I

I

~

I

I

I

I

-11

-10

-9

-8

-7

-6

-5

20

-

10

-

1)

-10

I

I

I

I

-9

-8

-7

-6

I

-5

Log concentration muscimol (M) Fig. I. Effect of (A) pentobarbitone 100#M (O) or DMBB 10#M (O) and 100#M (0) and (B) alphaxalone 1 # M (&) on the concentration-effect curve to muscimol (11, control) to inhibit K ÷-evoked release of [3H]GABA from slices of cerebral cortex of the rat. RESULTS The basal release of tritium from slices of rat cerebral cortex was initially rapid but declined exponentially and after 60 min of superfusion became relatively constant at a rate of 0.11 + 0.01% min -1 (n = 14) of total tissue radioactivity. Raising the potassium ion concentration in the superfusing medium to 25 mM produced a total fractional release of tritium above basal of 1.70 +0.24 (S,) and 1.74+0.26 ($2). The control SJSI ratio was 1.03-t-0.04 (n = 14). The K+-evoked, but not the basal release, was found to be largely calcium-dependent. The $2/S] ratio in the absence of Ca 2+ (obtained by iso-osmotic replacement with Mg 2+ in the presence of EGTA (10#M) was 0.33 + 0.06 (n = 3). Muscimol (1 nM to 1/~M) produced a concentration related inhibition of K+-evoked efltux of tritium with a plC25 value (negative logarithm of the molar concentration required to produce a 25% inhibition) of 7.34+ 0.10 (n = 13). The maximum inhibition (Emax) produced by muscimol ( I # M ) was 35 ___5.6%. Larger concentrations of muscimol increased the basal effiux of tritium. The effect of muscimol was antagonised in a competitive manner by bicuculline with a pK Bvalue of 7.52 +__0.11 (n = 3). Pentobarbitone (100/~M) and dimethylbarbituric acid (DMBB) (10 and 100/tM) had no significant effect on K+-evoked [3H]GABA efflux (103 + 5.8%, 97 ___1.7% and 100_ 6.2% of control, respectively). Both compounds, however, shifted the concentrationeffect curve for muscimol to the left [Fig. I(A)]. The plC25 and Emax values are shown in Table 1. Alphaxalone (1 #M) had no significant effect on either K ÷-evoked release of tritium or on the concentration-effect curve to muscimol [Fig. I(B) and Table 1]. 5fl-Pregnane-3~t-ol-20-one and 5flpregnane-3,20-dione both significantly enhanced the K ÷-evoked release of tritium in a dose-related manner, although 5fl-pregnane-3,20-dione produced

the greater increase (Table 2). At a concentration of 5fl-pregnane-3~t-ol-20-one (10 nM) which had no significant effect alone, there was no effect on the response to muscimol (100 nM) (28 + 1.6% and 25 + 3.0% inhibition for steroid and control, respectively). DISCUSSION The results of the present study demonstrate a modulation of the GABA autoreceptor by barbiturates but not by steroids of the pregnane series. Both the barbiturates tested, pentobarbitone and DMBB, produced a shift to the left in the concentration-effect curve to muscimol and an increase in the apparent maximum effect. A true estimate of the maximum inhibition could not be obtained, since at concentrations greater than 1 #M, muscimol increased the basal efflux of tritium, therefore the percentage inhibition produced by 1/~ M muscimol in the absence and presence of the modulators was compared. The effect of DMBB appeared to be concentration-dependent and the degree of shift in the concentration-effect curve to muscimol, produced by DMBB, was greater than that produced by pentobarbitone, the effect of 10/~M DMBB being approximately equivalent to that of 100/~M pentobarbitone. Although it is not valid to compare potencies using a single concentration, this observation is in agreement with the reported greater potency of DMBB in potentiating the response to muscimol in the cuneate nucleus of the rat and in the enhancement of [3H]GABA binding (Simmonds and Turner, 1987). Table 1. Effect of barbiturates and alphaxalone on the concentration-effect curve to muscimol

Control +Pentobarbitone 100/aM + D M B B 10uM + D M B B 100/aM +Alphaxalone l/aM

plC2~ _+_SEM

Emax + SEM

7.86 + 0.39 8.50 + 0.21"* 8.31 +0.09** 9.91 +0.21"* 7.91 _ 0.10

35.16 +_ 5.55 51.99 _+ 3.75* 51.42+ 1.59" 58.64+3.11"* 33.58 + 1.52

*P ~<0.05, **P ~<0.01 compared to muscimol alone.

Modulation of the GABAA-Iike autoreceptor Table 2. Effect of pregnane steroids on K ÷-evoked release of [3H]GABA Steroid Concentration (#M) % Control 5fl-Pregnane-3~t-ol-20-one 0.01 101 _+_2.7 0.1 119+3.2" 1.0

10.0 0.01 5fl-Pregnane-3,20-dione 0.1 1.0 10.0 *P ~<0.05 compared to potassium alone.

126 -4- 6.8*

90 + 9.4 111 + 11 145 + 10" 157 _+17" 165 + 9*

At the concentrations tested, neither D M B B nor pentobarbitone affected K +-evoked release of [3H]GABA per se, although both compounds have been reported to have a direct effect on the chloride channel at concentrations above 5 0 # M (Peters, Kirkness, Callachan, Lambert and Turner, 1988). The steroid anaesthetic alphaxalone (3ct-hydroxy5ct-pregnane-ll,20-dione) has been reported to potentiate the depolarizing effect of muscimol at the postsynaptic G A B A A receptor in the cuneate nucleus of the rat in vitro with a maximum effect of a shift of 0.3-0.4 log units in the dose-response curve at 1 # M (Turner and Simmonds, 1989). At this concentration, alphaxalone had no effect on muscimol-induced inhibition of K + -evoked release of [3H]GABA. The maximum effect of alphaxalone in the cuneate nucleus preparation is less than that reported previously for the barbiturates in the same preparation (Harrison and Simmonds, 1983), suggesting a lower efficacy for the former compound. However, a concentration of 1 0 p M D M B B produced a 0.3log unit shift in the concentration-effect curve to muscimol in the cuneate nucleus and in the present study produced a shift of 0.36 log units. Therefore, a concentration of alphaxalone which was equi-effective with a small concentration of barbiturate at the postsynaptic G A B A A receptor failed to affect the presynaptic G A B A autoreceptor whereas this concentration of barbiturate was effective in potentiating the effect of muscimol at the autoreceptor. This finding is difficult to explain by differences of efficacy of the two compounds. Furthermore, two other steroids of the pregnane series, 5fl-pregnane-3ct-ol-20-one and 5fl-pregnane-

357

3,20-dione both enhanced the K +-evoked release of [3H]GABA. However, at a concentration of 5fl-pregnane-3~t-ol-20-one ( 1 0 n M ) which had little effect alone, there was no change in the response to muscimol. The increase in release of [3H]GABA produced by these two steroids may result from a physical perturbation of the neuronal membrane, as suggested by Simmonds and Turner (1987). These data suggest that the G A B A autoreceptor is modulated by barbiturates in addition to benzodiazepines but unlike the classical G A B A A receptor, it is not modulated by steroids. This suggests that there is a distinction between the G A B A autoreceptor described in this study and other G A B A A receptors and raises the possibility that this may reflect an unusual receptor subunit composition. REFERENCES

Ennis C. and Minchin M. C. W. (1988) Modulation of the GABA autoreceptor by benzodiazepine ligands. Neuropharmacology 27: 1003-1006. Ennis C. and Minchin M. C. W. (1989) Barbiturates but not steroids modulate the GABA autoreceptor. Br. J. Pharmac. 97: 498P. Harrison N. L. and Simmonds M. A. (1983) Two distinct interactions of barbiturates and chlormethiazole with the GABA a receptor complex in rat cuneate nucleus in-vitro. Br. J. Pharmac. 80: 387-394. Minchin M. C. W., Ennis C., Lattimer N., White J. F., White A. C. and Lloyd G. K. (1992) A novel GABAA-like autoreceptor modulates GABA release. Molec. Neuropharmac. 2: 137-139. Olsen R. W. and Tobin A. J. (1990) Molecular biology of GABAa receptors. FASEB J. 4: 1469-1480. Peters J. A., Kirkness E. F., Callachan H., Lambert J. J. and Turner A. J. (1988) Modulation of the GABA A receptor by depressant barbiturates and pregnane steroids. Br. J. Pharmac. 94: 1257-1269. Simmonds M. A. and Turner J. P. (1987) Potentiators of responses to activation of y-aminobutryric acid (GABAA) receptors. Neuropharmacology 26: 923-930. Turner J. P. and Simmonds M. A. (1989) Modulation of the GABA A receptor complex by steroids in slices of rat cuneate nucleus. Br. J. Pharmac. 96: 409-417. Waldmeier P. C. and Baumann P. A. (1990) Presynaptic GABA receptors. In: Presynaptic Receptors and the Question of Autoregulation of Neurotransmitter Release (Kalsner S. and Westfall T. C., Eds). Ann. N. Y. Acad. Sci. 604: 136-151.