- Email: [email protected]
Benzodiazepines both enhance γ-aminobutyrate responses and decrease calcium action potentials in guinea-pig myenteric neurones
0306-4522/S $3.00 + 0,OO Pergamon Press Ltd IBRO
N~~rosc~e~eeVol. 14, No. 1, pp. 309-315, 1985 Printed in Great Britain
BENZODIAZEPINES BOTH ENHANCE ~-AMINOBUTY~TE RESPONSES AND DECREASE CALCIUM ACTION POTENTIALS IN GUINEA-PIG MYENTERIC NEURONES Neuropha~acology
E. CHERUBINI and R. A. NORTH* Laboratory, Massachusetts Institute of Technology, 56245, Cambridge, U.S.A.
MA 02139,
Abstract-The effect of two benzodiazepines, midazolam and diazepam, was studied in guinea-pig myenteric neurones, using intracellular recording techniques. Both these benzodiazepines (lO@-300PM) potentiated the rapidly desensitizing, bicuculline-sensitive depolarization, induced by a-aminobutyrate ionophoresis. Concentrations of midazolam and diazepam higher than 100nM depressed the ~-aminobutyrate-indu~d depolarization. The potentiating effect of the benzodiazepines was reversibly abolished by Ro 15-1788 (l-100 nM) and by ~ntylenetetr~ol (100 FM). A second effect of midazolam and diazepam (1~3~pM) was a reversible depression of the amplitude and duration of the directly evoked action potential in 29% of neurones, without affecting membrane potential or conductance. The effect was very marked when electrodes were filled with CsCl, and was also seen in the presence of tetrodotoxin. In some but not all of these neurones, the amplitude and duration of the action potentials was reduced also by y-aminobutyrate (l-10pM). Ro 15-1788 and pentylenetetrazol reversibly abolished the effect of benzodiazepines on the action potential, but not that of y-aminobutyrate. Thus, benzodiazepines have two effects on myenteric neurones. The first is an enhan~ment of the ~-aminobutyrate response (activation of Cl ~onductance~; the second is a depression of the calcium action potential, which appears to be independent of y-aminobutyrate.
It is well established that benzodiazepines act on central neurones by potentiating the action of y-aminobutyrate (GABA). This potentiation may be due to an increase in the affinity of GABA for its receptor (for a review see Ref. 19); it could account at least in part for the anxiolytic, antiepileptic and sedative-hypnotic effects of these drugs.)’ In peripheral tissues, benzodiazepine binding sites have been described which are pha~acologi~ally distinguishable from those in the central nervous system, 3o but their functional role is still unclear. Benzodiazepine binding sites of this peripheral type have been recently identified in the longitudinal muscle/myenteric plexus preparation of the guineapig ileum.” No correlation was found between the binding affinity of various benzodiazepine ligands and their potency in inhibiting electrically induced contractions.“*‘3 y-Aminobutyrate is present in nerves of myenteric plexus,‘4 and it may play a physiological role,.I5 therefore, we tested the action of two ~~odiazepines on the effect of GABA on single neurones of guinea-pig myenteric plexus. EXPERIMENTAL
PROCEDURES
Adult guinea-pigs were stunned and bled. The ileum was removed and placed in a solution of the following composition (mM): NaCl 117, KC1 4.7, NaH,PG, 1.2, CaCl,, 2.5, MgCl, 1.2, NaHCO, 25, glucose 11, gassed with OZ95% and CO, 5%. Intracellular recordings were made from neurones lying within ganglia of the myenteric plexus.*OThe *To whom all correspondence should be addressed. Abbreviutiort: GABA, y-aminobutyrate.
superfusing solution was pumped at l-2 ml/min and heated so that its temperature over the tissue was 37°C. Microelectrodes filled with KC1 (2 M) were used (d.c. resistance 60-100 MD). The cells were impaled under direct visual control as they lay in a shallow bath (volume l-2 ml) on the stage of a microscope (Zeiss Nomarski, total magnification x 500). Data were stored on a digitizing oscilloscope and then played on to a chart recorder. y-Aminobutyrate and muscimol were applied by ionophoresis from ~cropi~ttes containing 100 mM-1 M solution. The pipette tips were positioned 5-10 pm from the cell soma under visual control. in some experiments GABA was administered by applying a brief (IO-50 ms) pressure (SO-100 kPa) pulse to a micropipette containing GABA (100 PM-10 mM in physiological saline, pH 7.4) which was positioned 1&20~m from the impaled cell. Other drugs were apphed by changing the superfusing solution to one which differed only in its content of drug. There was a delav of 45-60 s between turning the tap and the first arrival at the tissue of the changed solution. The ratio of flow rate to bath volume ensured complete exchange within 1 min. Drugs used were: GABA (Sigma), pentylenetetrazol (Sigma), diazepam, midazolam maleate, ko i5-1788 (Dr. Mohler) and diltiazem (Dr. Muaelli). Diazeoam and Ro 1S- 1788 were dissolved in one to rwo drops of’ glacial acetic acid and diluted to a stock solution of 1 pM. These and other drugs were then diluted in physiological saline solution. Values given in this paper are means with the SEM for the number of observations in parenthesis. RESULTS
Myenteric neurones were ciassified as S type or AH type according to the presence (S type) or the absence (AH type) of a fast nicotinic excitatory postsynaptic potential, along with the presence (AH) or absence (S) of a prominent afterhyperpolarization follow309
E. Cherubini and R. A. North
310
ing a single soma action potential.‘0~20 Since we previously found that only AH type neurones responded to GABA,’ we studied the effect of benzodiazepines mainly on AH type neurones (106 AH and 12 S). The resting membrane potential of the neurones was between -55 and -65 mV. Potentiation of y aminobutyrate
response
y-Aminobutyrate has two actions on myenteric neurones. When applied by ionophoresis or pressure ejection it causes a rapidly desensitizing, bicucullinesensitive depolarization; when applied by superfusion, it causes a rn~nt~n~ depo1a~~tion.l The bicuculline-sensitive depolarization results from an
increase of chloride conductance which carries the membrane toward the chloride equilibrium potential (about -18 mV with KC1 electrode).’ The maintained depoiari~tion observed with GABA superfusion may result from blockade of a constant inward calcium current that contributes to the resting potassium conductance.‘,’ We tested the action of diazepam and midazolam on both these response to GABA. Midazolam and diazepam, in very low doses, potentiated the response to GABA iontophoresis. This potentiation occurred in 51 of 83 AH neurones (61%) that responded to GABA (Figs 1 and 2). The enhancement of GABA response by diazepam
A
A
A
GABA
Wash
Midarolam
Control
-I 500
20
mV
ms
Fig. 1. Potentiation of GABA depolarization by midazolam. y-Aminobutyrate was applied by ionophoresis (at the triangles) before (left), during (center) and after (right) superfusion of midazolam (300 PM). The recoveries were 20 min after washing out midazolam. The amount of GABA charge ejected was 0.6 nC in the upper trace and 1nC in the lower trace. Downward deflections are electrotonic potentials resulting from repeated passage of a fixed current pulse. Resting membrane potential was - 58 mV.
A,..,..,h
Oiazepamh
Rh
hh
!L.
h
GABA
h
(nC)
h
120
mV
Fig. 2. Potentiation of GABA response by diazepam. (A} Dose-response curve to GABA applied by ionophoresis (the amount of charge ejected is indicated in nC below the triangles) before (upper trace), during (central trace) and after (lower trace) super&on of diazepam (300pM). (B) The amplitude of GABA depolarization (ordinate) is plotted vs the amount of charge ejected (abscissa), before (e), during (Of and after ( x ) superfusion of diazepam (300 PM). The flattening of the response with high doses of GABA presumably reflects the approach to the chioride equilibrium potential.
Benz~i~pines
311
on my~te~c neurons
(100pM) was 4.5.3fS.l% (3), diazepam (300pM) 49.6 f 8.3% (6), midazolam (100 PM) 51 It 5.3% (4), and midazolam (300pM) 45.8 + 3.9% (7). A similar potentiation was observed by higher doses of midazolam and diazepam [l nM gave an increase of 43.4 + 6.3% (IQ, 1OnM gave an increase of 44.2 + 6.7% (15)]. The potentiation of the GABA response occurred also when the tissue was superfused with a solution containing tetrodotoxin (300 nM-1 PM) (n = 4), implying that it was a direct effect on the impaled neurone. The effect started 1-3 min after the be~odi~epines reached the tissue and lasted for IO-30min after washing out. The increase in amplitude of the GABA depolarization was associated with an increase in the conductance change. If the peak input conductance during the GABA depolarization was 100% prior to benzodiazepines, then it was 120 $- 1.8% (8) in the presence of benzodiazepines. This increase in conductance was not due to membrane rectification, because it was still present during recording with CsCl-filled electrodes which made linear the v/Z relation between - 60 and -20 mV. No change in the GABA reversal potential (- 18 mV with KC1 electrodes’) was observed in the presence of midazolam (300 pM-1 nM). This suggests that midazolam does not directly alter the conductance activated by GABA.
Midazolam
Higher concentrations (> 100 nM) of midazolam depressed GABA depolarizations. Such a biphasic action, potentiation at lower concentrations and depression at higher concentrations, is similar to that observed in central neurones.1’~‘,22 The potentiating effect of the benzodiazepines on the GABA response was not observed in the presence of Ro 15-1788 (n = 5). In four cells, the effect was blocked by 1 nM Ro 15-1788, and in another cell by 100 nM Ro 15-1788 (Fig. 3). Ro 15-1788 itself had no effect on membrane potential or conductance, or on the GABAinduced depola~zation. The enhancement of the GABA response was also reversibly abolished by pentylenetetrazol in a dose (IOOpM) that only slightly depressed the GABA response (Fig. 4). Bicuculline (10 PM) depressed the GABA response, and the facilitatory effect of midazolam and diazepam were no longer observed. When the GABA depolarization gave rise to action potentials, the potentiation of the GABA response by midazolam or diazepam was associated with a reduction in the number or complete blockade of the action potentials (Figs 3 and 4). It is possible that this effect might be partially explained by the “shunt” of the membrane due to the increase in conductance. This implies that benzodiazepines enhance the inhibitory effect of GABA (ie. chloride conductance increase)
Wash
250ms Fig. 3. The effect of midazoiam is blocked by Ro 15-l788. (A) Superfusion of midazolam ( 1nM) enhanced GABA response (GABA was ejected by ionophoresis at the triangles; amount of charge: 1.2 nC). The effect of midazolam was completely blocked by superfusion of Ro 1.5-1788(1 nM) (B), and washed out in 45 min (C). In this neurone GABA depolarization was followed by a hyperpolarization that lasted 8 s. Note that midazolam reduced the number of action potentials evoked by the GABA depolarization.
E. Che~bini and R. A. North
312 Control
B
Diazepam
diazepines was studied after the impalement had been continued for 30min, by which time the action potential duration reached a steady value (100-500 ms). These prolonged action potentials were reversibly blocked by tetrodotoxin and cobalt (l-2 mM) or by tetrodotoxin in a calcium-free, high (10 mM) magnesium solution. Superfusion with midazolam and diazepam reversibly depressed the duration and the amplitude of the action potential in 11 neurones (29%) (Fig. 5). Diazepam (100 PM) reduced the duration of the action potential to 45.3 + 5.8% (6) of control. Equivalent values for diazepam (I nM) were 41% (1), midazolam (300pM) 21 and 38%,
Wash
’
PTZ I J A
C
A
I
I
I
I \ I
&“+roI
L -I 15 mV 200
ms
Fig. 4. The enhancement of GABA response by diazepam is blocked by pentylenetetrazol (PTZ). The potentiating effect of diazepam (100 PM) on the depolarizing response to GABA ionophoresis (2.5 nC at the triangles) (A), was blocked by superfusion of pentylenetetrazol (1OOpM) (B), and washed out in 25min (C). Pentylenetetrazol had no effect on the resting membrane potential or on GABA response. In this cell GABA depolarization was followed by a hype~olarization lasting 12 s. even in conditions when GABA itself is excitatory. Midazolam and diazepam did not change the amplitude and duration of the afterhyperpolarization that sometimes followed GABA response,’ or the afterhyperpolarizat~on that followed the direct evoked action potential. The effect of midazoiam and diazepam was not due to the interference with the GABA uptake system, since midazolam (1 nM) also enhanced muscimol-induced depolarization and the associated conductance increases [increase in amplitude was 37 * 2.17; (4)]. The maintained depolarization caused by perfusion of GABA (100 PM), which seems not to result from activation of GABA receptors,’ was not changed by midazolam 300 pM (n = 2) or 1 nM (n = 1). Midazolam (I nM) also did not change the degree or the time course of the desensitization of the response to GABA ionophoresis which occurs during GABA perfusion.’
Shortening
qf’the calcium action potentials
Midazolam (300pM, n = 1) and diazepam (100 pM, n = 2) reversibly decreased the amplitude and the duration of the action potentiai evoked by passing a brief depolarizing current. The effect of these benzodiazepines was particularly marked when the action potential was prolonged by inhibiting spike repolarization. Neurones (n = 37) were impaled with CsCl (1-2 M) electrodes and the effect of benzo-
B
I
GJASA
:“:i
i
Control
. Wash
jL
j
Midazolam
Wash
L-
‘-
-I 25ms mV
200
Fig. 5. Depression of the calcium component of the action potential by GABA and mid~oIam. Single oscilloscope traces of direct evoked action potentials recorded with CsCl-filled electrodes. (A) Effeci of GABA (1 PM). (B) Effect of midazolam (300 PM) on the same cell. No changes in membrane potential or conductance were observed during satisfaction of GABA and midazolam.
Control
JControl i
GABA
’
$ABA
(10
PM)
‘L_
(100 p4)
V---i
Wash
‘WashL
l---Y
1 L_ J-LControl --
Midazolam
iash :I
-I
200
50 mV
ms
Fig. 6. Depression of the calcium component of the action potentia1 by midazolam but not by GABA. Single oscilloscope traces of the direct evoked action potentials recorded with CsCl-filled electrodes. The traces on the left are controls and on the right recoveries after wash out of GABA or midazolam. (A), (B) y-Aminobutyrate (10, 1OOyM) had no effect on the action potential. (C) Midazolam (I nM) much reduced the amplitude and duration of the calcium component of the spike, without affecting membrane potential or conductance.
~nzodiazepin~
313
on myenteric neurons
I
50 mV
200 ms Fig. 7. Pentylenetetrazol blocks the effect of diazepam on the action potential, but does not affect the action of GABA. Oscilloscope traces of the direct evoked action potentials recorded with CsCl-filled electrodes. (A) Superfusion of diazepam (100 PM) (DZ) depressed the amplitude and the duration of the direct evoked action potential. (B) The effect of diazepam did not occur after pentylenetetrazol(lO0 PM) (PTZ). Pentylenetetrazol had no effect on membrane potential and resistance, and did not change the shape of the action potential. (C) Superfusion of GABA (100 PM) caused 8 mV depolarization of the membrane potential and reduce the amplitude and duration of the calcium action potential (evoked after repolarization to the resting membrane potential). Pentylenetetrazol (100 PM) did not affect the action of GABA. (D) After washing out the PTZ, DZ had its initial effect. midazolam (1 nM) 13 and 45%. In five of these cells tetrodotoxin (300 nM-1 PM) was present also in the superfusing solution. This effect occurred in absence of any change in membrane potential and conductance, and washed out in 3-6min. Since diazepam is structurally similar to diltiazem, a benzothiazepine known to block calcium channels in smooth muscles,” the action of diltiazem was tested in six neurones that responded to diazepam. None were affected by diltiazem in concentrations of 100 nM-1 pM. y-Aminobutyrate is known to reduce the tetrodotoxin resistant spikes in AH neurones.’ y-Aminobutyrate was tested in the same neurones affected by benzodiazepines. In eight of them the action potentials were also reduced in duration by GABA (l-10pM) (Fig. 5). In the other three the action potentials were not affected, even by higher concentrations of GABA (100 PM) (Fig. 6). Ro 15-1788 (100nM) and pentylenetetrazol (100pM) reversibly abolished the action of midazolam on the action potential, but they did not affect the action of GABA (Figs 7 and 8). Bicuculline (10 @h/l) did not block this action of midazolam. Midazolam (100 PM) also depressed the duration of the action potentials of two S neurones recorded with CsCl electrodes that were also not affected by GABA. Membrane potential and input resistance were genera& unaffected by the benzodiazepines
Six AH neurones were depolarized 2-4 mV from the resting membrane potential by low concentrations of midazolam and diazepam. The small depolarization was associated with a surprisingly large increase in membrane conductance. Conductance was changed by (control = 100%): midazolam
(3Or)PM) = 55 and 38%; midazolam (10 nM) 32 and 62%; diazepam (1 nM) = 70% and diazepam (10 nM) = 5 1%. The increase in conductance was still present after restoring the membrane potential to the resting level. The depolarization and conductance
A
j\‘l C0ntr0l
6
A
Wash
r---l r) ,~~
Control
52 8
Ro 15-1788
c \ Control
i
_J
Wash
4
GABA< 8 Ro 15-1788
‘ashJ2O 500
mV ms
Fig. 8. Ro 15-1788 blocks the effect of diazepam (DZ) on the calcium component of the action potential, but does not block the effect of GABA. Oscilloscope traces of action potentials recorded with CsCl-filled electrodes. On the left are controls, on the right are recoveries. {A) Superfusion with diazepam (1 nM) (DZ) reduced the duration of the action potential. (B) This effect of diazepam was blocked by concomitant superfusion of Ro 15-1788 (LOOnM). (C) y-Aminobutyrate (10 h M) depressed the amplitude and duration of the action potential even in presence of Ro U-1788. y-Aminobutyrate and diazepam had no effect on the resting membrane potential of this neurone.
314
E. Cherubim and R. A. North
increase started a few seconds after the drug reached the bath and lasted for a few minutes after washing out midazolam or diazepam. This effect was reversibly blocked by Ro 15-1788 (100 nM) against midazolam (10 nM, n = 2) and diazepam (1 nM, n = 1) and partially blocked by bicuculline (10 /IM, n = 3). The effect of midazolam on the membrane potential and conductance was also reversibly abolished by cobalt (2 mM) and high magnesium (10 mM) solution (n = 4), implying that it resulted from release of other substances from surrounding neurones. Higher doses of midazolam and diazepam (100 nM-1 PM) occasionally caused a hyperpolarization of the membrane, associated with a small decrease in conductance. No effects on the membrane potential and conductance of S neurones were observed. DISCUSSION
Two main actions of benzodiazepines on myenteric plexus neurones appear in the present study. The first is a potentiation of the bicuculline-sensitive depolarizing response to GABA; the second is the inhibition of the calcium action potential. The potentiation of the bicuculline-sensitive GABA depolarization by low doses of diazepam and midazolam is similar in all respects to that observed in central neurones,3~‘b’8~2’~22~26~29 and in cat superior cervical ganglia*’ (for a review see also Ref. 32). This effect results from a direct postsynaptic action on the impaled cell, because it is still present in tetrodotoxin. It is a specific receptor-mediated event, since it is blocked by Ro 15-1788, a selective antagonist of benzodiazepines.‘2,28 The benzodiazepines probably bind to sites associated with GABA receptors, activation of which leads to an increase in chloride conductance. In central neurones, where bicuculline blocks the enhancement of benzodiazepine binding by GABA agonists,33 a functional interaction occurs between the benzodiazepine binding site and the GABA receptor complex. In the guinea-pig ileum longitudinal muscle-myenteric plexus preparation, the benzodiazepine binding sites are of relatively low affinity (“peripheral type ““.30). The present observations suggest that the myenteric neurones themselves bear receptors more similar to the “central effective concentrations are type” because 100 pM-1 nM whereas the IC,, value for displacing [3H]diazepam bound to guinea-pig ileum longitudinal muscle membrane is 14nM.” The sites observed by Hullihan et al.” may be on the smooth muscle. The second and previously undescribed effect is a
reduction of the calcium spike. This is similar to that
recently reported for GABA.’ Both the benzodiazepine and the GABA effect are insensitive to bicuculline. Could this effect of diazepam and midazolam result from an endogenous release of GABA, which then acts on GABA, receptors? Release of GABA by benzodiazepines has been reported.4,23,3’In a few neurones midazolam and diazepam induced a small depolarization of the membrane potential associated with conductance increase; this was partially blocked by bicuculline and reversibly abolished by cobalt and high magnesium suggesting that release of GABA might be involved. However, the blockade of calcium spike by low doses of benzodiazepines was never associated with changes in membrane potential or conductance and several other findings suggest that this action of diazepam is independent of GABA. Some cells (including two S type cells) responded to diazepam and midazolam but not to exogenously applied GABA. In those neurones that respond to benzodiazepines and GABA, the potency of diazepam and midazolam in inhibiting the calcium spike was at least 100-1000 times higher than that of GABA. Ro 15-1788 and pentylenetetrazol blocked the effect of diazepam but not that of GABA. This action of benzodiazepines to reduce calcium spikes might underlie the inhibition of potassium-evoked noradrenaline release in synaptosomes of rat hippocampus.’ Since benzodiazepines inhibit adenosine uptake in central neurones,24.25 the possibility exists that the block of the calcium spike is mediated through adenosine receptors. Adenosine is known to reduce acetylcholine release in the guinea-pig ileumS and to shorten the calcium spike in rat superior cervical ganglion neurones.’ Adenosine antagonists or selective GABA, antagonists could be used to confirm this conclusion. The clinical effects of benzodiazepines in the treatment of anxiety, seizures and insomnia have been related to the enhancement of GABA-induced inhibition and associated chloride conductance increase. The benzodiazepine receptor has been considered as a part of a supramolecular complex of functionally linked membrane proteins including the GABA receptor and its associated ionophore.‘9 The present findings lend support to this notion, yet while strengthening this hypothesis, further suggest that some clinical effect of benzodiazepines may be also independent of GABA. They might result from a reduction of transmitter release. Acknowledgemenf-Supported
by U.S. Department Health and Human Services grant AM/NS 32979.
of
REFERENCES
1. Cherubini E. and North R. A. (1984) y-Aminobutyric acid actions on neurones of guinea-pig myenteric plexus. Br. J. Phurmac. 82, 93-100. 2. Cherubini E. and North R. A. (1984) Inhibition of calcium spikes and transmitter release by y-aminobutyric acid in the guinea-pig myenteric plexus. Br. J. Pharmac. 82, 101-106.
~~~ia~pines
on myenteric neurons
315
3. Choi D. W., Farb D. H. and Fischbach G. D. (1981) Chlordiazepoxide selectively potentiates GABA conductance of spinal cord and sensory neurons in cell culture. J. Neurophysiol. 45, 621-631. 4. Curtis D. R., Lodge D., Johnston G. A. R. and Brand S. J. (1976) Central actions of benzodiazepines. Brain Res. 118, 344-347. 5. Dowdle E. B. and Maske R. (1980) The effects of calcium concentration
on the inhibition of cholinergic neurotransmission in the myenteric plexus of guinea-pig ileum by adenosine nucleotides. iir. J. Pharmac. 71,245-252. 6. Fleckenstein A. (1977) Specific pharmacology of calcium in myocardium, cardiac pacemakers, and vascular smooth muscle. A. Rev. Pharmac. Toxicol. 17, 149-166. 7. Fung S. C. and Fillenz M. (1983) The role of presynaptic GABA and benzodiazepine receptors in the control of noradrenaline release in rat hippocampus. Neur&ci. -Lett. 42, 61-66. 8. Grafe P.. Maver C. J. and Wood J. D. (1980) Svnantic modulation of calcium-denendent notassium conductance in myenteric neurones in the guinea-pig. J.‘ Physioi, Land. 305, 235-248. 9. Henon B. K. and McAfee D. A. (1983) The ionic basis of adenosine receptor actions on post-ganglionic neurones in the rat. J. Physiol., Lond. 336, 607620. 10. Hirst G. D. S., Holman M. E. and Spence I. (1974) Two types of neurones in the myenteric plexus of duodenum in the guinea-pig. J. Physiof., Lond. 236, 303-326. 11. Hullihan J. P., Spector S., Taniguchi T. and Wang J. K. T. (1983) The binding of [)H]-diazepam to guinea-pig ileal longitudinal muscle and the in vitro inhibition of contraction by benzodiazepines. Br. J. Phartnac. 78, 321-327. 12. Hunkeler W., Mohler H., Pieri L., Pole P., Bonetti E. P., Cumin R., Schaffner R. and Halfely W. (1981) Selective antagonists of benzodiazepines. Nature 290, 514516. 13. Ishii K., Kano T., Akutagawa M., Makino M., Tanaka T. and Ando J. (1982) Effects of flurazepam and diazepam in isolated guinea-pig taenia coli and lon~tu~nal muscle. Eur. J. P~r~c. 83, 329-333. 14. Jessen K. R., Mirsky R., Dennison M. E. and Burnstock G. (1979) GABA may be a neurotransmitter in the vertebrate peripheral nervous system. Nature 281, 71-74. 1.5. Krantis A. and Kerr D. I. B. (1981) Autoradiographic localization of [3H]gamma-aminobutyric acid in the myenteric plexus of the guinea-pig small intestine. Neurosci. Lett. 23, 263-268. 16. Lee H. K., Dunwiddie T. W. and Hoffer B. J. (1979) Interaction of diazepam with synaptic transmission in the in vitro rat hippocampus. NQunyn-Schmeidebergs Arch. Pharmac. 309, 131-136. 17. Macdonald J. F. and Barker J. L. (1982) Multiple actions of picomolar concentrations of thtrazepam on the excitability of cultured mouse spinal neurons. Brain Res. 246, 257-264. 18. Macdonald R. and Barker J. L. (1978) Benzodiazepines specifically modulate GABA-mediated postsynaptic inhibition in cultured mammalian neurones. Nature 271, 563-564. 19. Mohler H. and Richards J. G. (1983) Receptors for anxiolytic drugs. In Anxiolytics: Neurochemical, Behavioral and Clinical Perspectives (eds Malick J. B., Enne S. J. and Yamamura H. I.), pp. 15-39. Raven Press, New York. 20. Nishi S. and North R. A. (1973) Intracellutar recording from the myenteric plexus of the guinea-pig ileum. J. Physiol., Lond. 231, 471-491. 21. Nistri A. and Berti C. (1983) Potentiating action of midazolam on GABA-rn~iat~ responses and its ~tagonism by Ro 14-7437 in the frog spinal cord. Neurosei. Lett. 39, 199-204. 22. Nistri A. and Constanti A. (1978) Effects of flurazepam on amino acid-evoked responses recorded from the lobster muscle and the frog spinal cord. Neuropharmacology 17, 127-135. 23. Olsen R. W., Lamar E. E. and Bayliss J. D. (1977) Calcium-induced release of y-aminobutyric acid from synaptosomes: effect of tranquilizer drugs. J. Neurochem. 28, 299305. 24. Phillis J. W. (1979) Diazepam potentiation of purinergic depression of central neurones. Can. J. Physiol. Pharmac. 57, 432-435. 25. Phillis J. W., Bender A. S. and Wu P. H. (1980) Benzodiazepines inhibit adenosine uptake into rat brain synaptosomes. Brain Res. 195, 494-498. 26. Riley M. and Scholfield C. N. (1983) Diazepam increases GABA mediated inhibition in the olfactory cortex slice. Pfiigers Arch. ges. Physiol. 397, 312-318. 27. Schlosser W. and Franc0 S. (1979) Modification of GABA-mediated depolarization of the cat ganglion by pentobarbitol and two benzodiazepines. Neuropharmacology 18, 377-381. 28. Scholheld C. N. (1983) Ro 1S-1788 is a potent antagonist of benzodiazepines in the olfactory cortex slice. Ppiigers Arch. ges. Physiol. 396, 292-296. 29. Simmons M. A. (1981) Distinction between the effects of barbiturates, benzodiazepines and phenytoin on responses to y-aminobutyric acid receptor activation and antagonism by bicuculline and picrotoxin. Br. J. Phurmac. 73, 739-747. 30. Squires R. F. and Braestrup C. (1977) Benzodiazepine receptors in rat brain. Nature 266, 732-734. 31. Suria A. and Costa E. (1975) Action of diazepam, dibutyryl cGMP and GABA on presynaptic nerve terminals in bullfrog sympathetic ganglion. Brain Res. 87, 102-106. 32. Tallman J. F.. Paul S. M.. Skolnick P. and Gallaeer D. W. (1980) Receptors for the age of anxiety: -_ _ - pharmacology _
of the benzodiazepines Science, N. Y. 207, 274-281. 33. White W. F., Dichtcr M. A. and Snodgrass S. R. (1981) Benzodiazepine binding and interactions with the GABA receptor complex in living cultures of rat cerebral cortex. Brain Res. 215, 162-176. (Accepted 25 July 1984)
N~~rosc~e~eeVol. 14, No. 1, pp. 309-315, 1985 Printed in Great Britain
BENZODIAZEPINES BOTH ENHANCE ~-AMINOBUTY~TE RESPONSES AND DECREASE CALCIUM ACTION POTENTIALS IN GUINEA-PIG MYENTERIC NEURONES Neuropha~acology
E. CHERUBINI and R. A. NORTH* Laboratory, Massachusetts Institute of Technology, 56245, Cambridge, U.S.A.
MA 02139,
Abstract-The effect of two benzodiazepines, midazolam and diazepam, was studied in guinea-pig myenteric neurones, using intracellular recording techniques. Both these benzodiazepines (lO@-300PM) potentiated the rapidly desensitizing, bicuculline-sensitive depolarization, induced by a-aminobutyrate ionophoresis. Concentrations of midazolam and diazepam higher than 100nM depressed the ~-aminobutyrate-indu~d depolarization. The potentiating effect of the benzodiazepines was reversibly abolished by Ro 15-1788 (l-100 nM) and by ~ntylenetetr~ol (100 FM). A second effect of midazolam and diazepam (1~3~pM) was a reversible depression of the amplitude and duration of the directly evoked action potential in 29% of neurones, without affecting membrane potential or conductance. The effect was very marked when electrodes were filled with CsCl, and was also seen in the presence of tetrodotoxin. In some but not all of these neurones, the amplitude and duration of the action potentials was reduced also by y-aminobutyrate (l-10pM). Ro 15-1788 and pentylenetetrazol reversibly abolished the effect of benzodiazepines on the action potential, but not that of y-aminobutyrate. Thus, benzodiazepines have two effects on myenteric neurones. The first is an enhan~ment of the ~-aminobutyrate response (activation of Cl ~onductance~; the second is a depression of the calcium action potential, which appears to be independent of y-aminobutyrate.
It is well established that benzodiazepines act on central neurones by potentiating the action of y-aminobutyrate (GABA). This potentiation may be due to an increase in the affinity of GABA for its receptor (for a review see Ref. 19); it could account at least in part for the anxiolytic, antiepileptic and sedative-hypnotic effects of these drugs.)’ In peripheral tissues, benzodiazepine binding sites have been described which are pha~acologi~ally distinguishable from those in the central nervous system, 3o but their functional role is still unclear. Benzodiazepine binding sites of this peripheral type have been recently identified in the longitudinal muscle/myenteric plexus preparation of the guineapig ileum.” No correlation was found between the binding affinity of various benzodiazepine ligands and their potency in inhibiting electrically induced contractions.“*‘3 y-Aminobutyrate is present in nerves of myenteric plexus,‘4 and it may play a physiological role,.I5 therefore, we tested the action of two ~~odiazepines on the effect of GABA on single neurones of guinea-pig myenteric plexus. EXPERIMENTAL
PROCEDURES
Adult guinea-pigs were stunned and bled. The ileum was removed and placed in a solution of the following composition (mM): NaCl 117, KC1 4.7, NaH,PG, 1.2, CaCl,, 2.5, MgCl, 1.2, NaHCO, 25, glucose 11, gassed with OZ95% and CO, 5%. Intracellular recordings were made from neurones lying within ganglia of the myenteric plexus.*OThe *To whom all correspondence should be addressed. Abbreviutiort: GABA, y-aminobutyrate.
superfusing solution was pumped at l-2 ml/min and heated so that its temperature over the tissue was 37°C. Microelectrodes filled with KC1 (2 M) were used (d.c. resistance 60-100 MD). The cells were impaled under direct visual control as they lay in a shallow bath (volume l-2 ml) on the stage of a microscope (Zeiss Nomarski, total magnification x 500). Data were stored on a digitizing oscilloscope and then played on to a chart recorder. y-Aminobutyrate and muscimol were applied by ionophoresis from ~cropi~ttes containing 100 mM-1 M solution. The pipette tips were positioned 5-10 pm from the cell soma under visual control. in some experiments GABA was administered by applying a brief (IO-50 ms) pressure (SO-100 kPa) pulse to a micropipette containing GABA (100 PM-10 mM in physiological saline, pH 7.4) which was positioned 1&20~m from the impaled cell. Other drugs were apphed by changing the superfusing solution to one which differed only in its content of drug. There was a delav of 45-60 s between turning the tap and the first arrival at the tissue of the changed solution. The ratio of flow rate to bath volume ensured complete exchange within 1 min. Drugs used were: GABA (Sigma), pentylenetetrazol (Sigma), diazepam, midazolam maleate, ko i5-1788 (Dr. Mohler) and diltiazem (Dr. Muaelli). Diazeoam and Ro 1S- 1788 were dissolved in one to rwo drops of’ glacial acetic acid and diluted to a stock solution of 1 pM. These and other drugs were then diluted in physiological saline solution. Values given in this paper are means with the SEM for the number of observations in parenthesis. RESULTS
Myenteric neurones were ciassified as S type or AH type according to the presence (S type) or the absence (AH type) of a fast nicotinic excitatory postsynaptic potential, along with the presence (AH) or absence (S) of a prominent afterhyperpolarization follow309
E. Cherubini and R. A. North
310
ing a single soma action potential.‘0~20 Since we previously found that only AH type neurones responded to GABA,’ we studied the effect of benzodiazepines mainly on AH type neurones (106 AH and 12 S). The resting membrane potential of the neurones was between -55 and -65 mV. Potentiation of y aminobutyrate
response
y-Aminobutyrate has two actions on myenteric neurones. When applied by ionophoresis or pressure ejection it causes a rapidly desensitizing, bicucullinesensitive depolarization; when applied by superfusion, it causes a rn~nt~n~ depo1a~~tion.l The bicuculline-sensitive depolarization results from an
increase of chloride conductance which carries the membrane toward the chloride equilibrium potential (about -18 mV with KC1 electrode).’ The maintained depoiari~tion observed with GABA superfusion may result from blockade of a constant inward calcium current that contributes to the resting potassium conductance.‘,’ We tested the action of diazepam and midazolam on both these response to GABA. Midazolam and diazepam, in very low doses, potentiated the response to GABA iontophoresis. This potentiation occurred in 51 of 83 AH neurones (61%) that responded to GABA (Figs 1 and 2). The enhancement of GABA response by diazepam
A
A
A
GABA
Wash
Midarolam
Control
-I 500
20
mV
ms
Fig. 1. Potentiation of GABA depolarization by midazolam. y-Aminobutyrate was applied by ionophoresis (at the triangles) before (left), during (center) and after (right) superfusion of midazolam (300 PM). The recoveries were 20 min after washing out midazolam. The amount of GABA charge ejected was 0.6 nC in the upper trace and 1nC in the lower trace. Downward deflections are electrotonic potentials resulting from repeated passage of a fixed current pulse. Resting membrane potential was - 58 mV.
A,..,..,h
Oiazepamh
Rh
hh
!L.
h
GABA
h
(nC)
h
120
mV
Fig. 2. Potentiation of GABA response by diazepam. (A} Dose-response curve to GABA applied by ionophoresis (the amount of charge ejected is indicated in nC below the triangles) before (upper trace), during (central trace) and after (lower trace) super&on of diazepam (300pM). (B) The amplitude of GABA depolarization (ordinate) is plotted vs the amount of charge ejected (abscissa), before (e), during (Of and after ( x ) superfusion of diazepam (300 PM). The flattening of the response with high doses of GABA presumably reflects the approach to the chioride equilibrium potential.
Benz~i~pines
311
on my~te~c neurons
(100pM) was 4.5.3fS.l% (3), diazepam (300pM) 49.6 f 8.3% (6), midazolam (100 PM) 51 It 5.3% (4), and midazolam (300pM) 45.8 + 3.9% (7). A similar potentiation was observed by higher doses of midazolam and diazepam [l nM gave an increase of 43.4 + 6.3% (IQ, 1OnM gave an increase of 44.2 + 6.7% (15)]. The potentiation of the GABA response occurred also when the tissue was superfused with a solution containing tetrodotoxin (300 nM-1 PM) (n = 4), implying that it was a direct effect on the impaled neurone. The effect started 1-3 min after the be~odi~epines reached the tissue and lasted for IO-30min after washing out. The increase in amplitude of the GABA depolarization was associated with an increase in the conductance change. If the peak input conductance during the GABA depolarization was 100% prior to benzodiazepines, then it was 120 $- 1.8% (8) in the presence of benzodiazepines. This increase in conductance was not due to membrane rectification, because it was still present during recording with CsCl-filled electrodes which made linear the v/Z relation between - 60 and -20 mV. No change in the GABA reversal potential (- 18 mV with KC1 electrodes’) was observed in the presence of midazolam (300 pM-1 nM). This suggests that midazolam does not directly alter the conductance activated by GABA.
Midazolam
Higher concentrations (> 100 nM) of midazolam depressed GABA depolarizations. Such a biphasic action, potentiation at lower concentrations and depression at higher concentrations, is similar to that observed in central neurones.1’~‘,22 The potentiating effect of the benzodiazepines on the GABA response was not observed in the presence of Ro 15-1788 (n = 5). In four cells, the effect was blocked by 1 nM Ro 15-1788, and in another cell by 100 nM Ro 15-1788 (Fig. 3). Ro 15-1788 itself had no effect on membrane potential or conductance, or on the GABAinduced depola~zation. The enhancement of the GABA response was also reversibly abolished by pentylenetetrazol in a dose (IOOpM) that only slightly depressed the GABA response (Fig. 4). Bicuculline (10 PM) depressed the GABA response, and the facilitatory effect of midazolam and diazepam were no longer observed. When the GABA depolarization gave rise to action potentials, the potentiation of the GABA response by midazolam or diazepam was associated with a reduction in the number or complete blockade of the action potentials (Figs 3 and 4). It is possible that this effect might be partially explained by the “shunt” of the membrane due to the increase in conductance. This implies that benzodiazepines enhance the inhibitory effect of GABA (ie. chloride conductance increase)
Wash
250ms Fig. 3. The effect of midazoiam is blocked by Ro 15-l788. (A) Superfusion of midazolam ( 1nM) enhanced GABA response (GABA was ejected by ionophoresis at the triangles; amount of charge: 1.2 nC). The effect of midazolam was completely blocked by superfusion of Ro 1.5-1788(1 nM) (B), and washed out in 45 min (C). In this neurone GABA depolarization was followed by a hyperpolarization that lasted 8 s. Note that midazolam reduced the number of action potentials evoked by the GABA depolarization.
E. Che~bini and R. A. North
312 Control
B
Diazepam
diazepines was studied after the impalement had been continued for 30min, by which time the action potential duration reached a steady value (100-500 ms). These prolonged action potentials were reversibly blocked by tetrodotoxin and cobalt (l-2 mM) or by tetrodotoxin in a calcium-free, high (10 mM) magnesium solution. Superfusion with midazolam and diazepam reversibly depressed the duration and the amplitude of the action potential in 11 neurones (29%) (Fig. 5). Diazepam (100 PM) reduced the duration of the action potential to 45.3 + 5.8% (6) of control. Equivalent values for diazepam (I nM) were 41% (1), midazolam (300pM) 21 and 38%,
Wash
’
PTZ I J A
C
A
I
I
I
I \ I
&“+roI
L -I 15 mV 200
ms
Fig. 4. The enhancement of GABA response by diazepam is blocked by pentylenetetrazol (PTZ). The potentiating effect of diazepam (100 PM) on the depolarizing response to GABA ionophoresis (2.5 nC at the triangles) (A), was blocked by superfusion of pentylenetetrazol (1OOpM) (B), and washed out in 25min (C). Pentylenetetrazol had no effect on the resting membrane potential or on GABA response. In this cell GABA depolarization was followed by a hype~olarization lasting 12 s. even in conditions when GABA itself is excitatory. Midazolam and diazepam did not change the amplitude and duration of the afterhyperpolarization that sometimes followed GABA response,’ or the afterhyperpolarizat~on that followed the direct evoked action potential. The effect of midazoiam and diazepam was not due to the interference with the GABA uptake system, since midazolam (1 nM) also enhanced muscimol-induced depolarization and the associated conductance increases [increase in amplitude was 37 * 2.17; (4)]. The maintained depolarization caused by perfusion of GABA (100 PM), which seems not to result from activation of GABA receptors,’ was not changed by midazolam 300 pM (n = 2) or 1 nM (n = 1). Midazolam (I nM) also did not change the degree or the time course of the desensitization of the response to GABA ionophoresis which occurs during GABA perfusion.’
Shortening
qf’the calcium action potentials
Midazolam (300pM, n = 1) and diazepam (100 pM, n = 2) reversibly decreased the amplitude and the duration of the action potentiai evoked by passing a brief depolarizing current. The effect of these benzodiazepines was particularly marked when the action potential was prolonged by inhibiting spike repolarization. Neurones (n = 37) were impaled with CsCl (1-2 M) electrodes and the effect of benzo-
B
I
GJASA
:“:i
i
Control
. Wash
jL
j
Midazolam
Wash
L-
‘-
-I 25ms mV
200
Fig. 5. Depression of the calcium component of the action potential by GABA and mid~oIam. Single oscilloscope traces of direct evoked action potentials recorded with CsCl-filled electrodes. (A) Effeci of GABA (1 PM). (B) Effect of midazolam (300 PM) on the same cell. No changes in membrane potential or conductance were observed during satisfaction of GABA and midazolam.
Control
JControl i
GABA
’
$ABA
(10
PM)
‘L_
(100 p4)
V---i
Wash
‘WashL
l---Y
1 L_ J-LControl --
Midazolam
iash :I
-I
200
50 mV
ms
Fig. 6. Depression of the calcium component of the action potentia1 by midazolam but not by GABA. Single oscilloscope traces of the direct evoked action potentials recorded with CsCl-filled electrodes. The traces on the left are controls and on the right recoveries after wash out of GABA or midazolam. (A), (B) y-Aminobutyrate (10, 1OOyM) had no effect on the action potential. (C) Midazolam (I nM) much reduced the amplitude and duration of the calcium component of the spike, without affecting membrane potential or conductance.
~nzodiazepin~
313
on myenteric neurons
I
50 mV
200 ms Fig. 7. Pentylenetetrazol blocks the effect of diazepam on the action potential, but does not affect the action of GABA. Oscilloscope traces of the direct evoked action potentials recorded with CsCl-filled electrodes. (A) Superfusion of diazepam (100 PM) (DZ) depressed the amplitude and the duration of the direct evoked action potential. (B) The effect of diazepam did not occur after pentylenetetrazol(lO0 PM) (PTZ). Pentylenetetrazol had no effect on membrane potential and resistance, and did not change the shape of the action potential. (C) Superfusion of GABA (100 PM) caused 8 mV depolarization of the membrane potential and reduce the amplitude and duration of the calcium action potential (evoked after repolarization to the resting membrane potential). Pentylenetetrazol (100 PM) did not affect the action of GABA. (D) After washing out the PTZ, DZ had its initial effect. midazolam (1 nM) 13 and 45%. In five of these cells tetrodotoxin (300 nM-1 PM) was present also in the superfusing solution. This effect occurred in absence of any change in membrane potential and conductance, and washed out in 3-6min. Since diazepam is structurally similar to diltiazem, a benzothiazepine known to block calcium channels in smooth muscles,” the action of diltiazem was tested in six neurones that responded to diazepam. None were affected by diltiazem in concentrations of 100 nM-1 pM. y-Aminobutyrate is known to reduce the tetrodotoxin resistant spikes in AH neurones.’ y-Aminobutyrate was tested in the same neurones affected by benzodiazepines. In eight of them the action potentials were also reduced in duration by GABA (l-10pM) (Fig. 5). In the other three the action potentials were not affected, even by higher concentrations of GABA (100 PM) (Fig. 6). Ro 15-1788 (100nM) and pentylenetetrazol (100pM) reversibly abolished the action of midazolam on the action potential, but they did not affect the action of GABA (Figs 7 and 8). Bicuculline (10 @h/l) did not block this action of midazolam. Midazolam (100 PM) also depressed the duration of the action potentials of two S neurones recorded with CsCl electrodes that were also not affected by GABA. Membrane potential and input resistance were genera& unaffected by the benzodiazepines
Six AH neurones were depolarized 2-4 mV from the resting membrane potential by low concentrations of midazolam and diazepam. The small depolarization was associated with a surprisingly large increase in membrane conductance. Conductance was changed by (control = 100%): midazolam
(3Or)PM) = 55 and 38%; midazolam (10 nM) 32 and 62%; diazepam (1 nM) = 70% and diazepam (10 nM) = 5 1%. The increase in conductance was still present after restoring the membrane potential to the resting level. The depolarization and conductance
A
j\‘l C0ntr0l
6
A
Wash
r---l r) ,~~
Control
52 8
Ro 15-1788
c \ Control
i
_J
Wash
4
GABA< 8 Ro 15-1788
‘ashJ2O 500
mV ms
Fig. 8. Ro 15-1788 blocks the effect of diazepam (DZ) on the calcium component of the action potential, but does not block the effect of GABA. Oscilloscope traces of action potentials recorded with CsCl-filled electrodes. On the left are controls, on the right are recoveries. {A) Superfusion with diazepam (1 nM) (DZ) reduced the duration of the action potential. (B) This effect of diazepam was blocked by concomitant superfusion of Ro 15-1788 (LOOnM). (C) y-Aminobutyrate (10 h M) depressed the amplitude and duration of the action potential even in presence of Ro U-1788. y-Aminobutyrate and diazepam had no effect on the resting membrane potential of this neurone.
314
E. Cherubim and R. A. North
increase started a few seconds after the drug reached the bath and lasted for a few minutes after washing out midazolam or diazepam. This effect was reversibly blocked by Ro 15-1788 (100 nM) against midazolam (10 nM, n = 2) and diazepam (1 nM, n = 1) and partially blocked by bicuculline (10 /IM, n = 3). The effect of midazolam on the membrane potential and conductance was also reversibly abolished by cobalt (2 mM) and high magnesium (10 mM) solution (n = 4), implying that it resulted from release of other substances from surrounding neurones. Higher doses of midazolam and diazepam (100 nM-1 PM) occasionally caused a hyperpolarization of the membrane, associated with a small decrease in conductance. No effects on the membrane potential and conductance of S neurones were observed. DISCUSSION
Two main actions of benzodiazepines on myenteric plexus neurones appear in the present study. The first is a potentiation of the bicuculline-sensitive depolarizing response to GABA; the second is the inhibition of the calcium action potential. The potentiation of the bicuculline-sensitive GABA depolarization by low doses of diazepam and midazolam is similar in all respects to that observed in central neurones,3~‘b’8~2’~22~26~29 and in cat superior cervical ganglia*’ (for a review see also Ref. 32). This effect results from a direct postsynaptic action on the impaled cell, because it is still present in tetrodotoxin. It is a specific receptor-mediated event, since it is blocked by Ro 15-1788, a selective antagonist of benzodiazepines.‘2,28 The benzodiazepines probably bind to sites associated with GABA receptors, activation of which leads to an increase in chloride conductance. In central neurones, where bicuculline blocks the enhancement of benzodiazepine binding by GABA agonists,33 a functional interaction occurs between the benzodiazepine binding site and the GABA receptor complex. In the guinea-pig ileum longitudinal muscle-myenteric plexus preparation, the benzodiazepine binding sites are of relatively low affinity (“peripheral type ““.30). The present observations suggest that the myenteric neurones themselves bear receptors more similar to the “central effective concentrations are type” because 100 pM-1 nM whereas the IC,, value for displacing [3H]diazepam bound to guinea-pig ileum longitudinal muscle membrane is 14nM.” The sites observed by Hullihan et al.” may be on the smooth muscle. The second and previously undescribed effect is a
reduction of the calcium spike. This is similar to that
recently reported for GABA.’ Both the benzodiazepine and the GABA effect are insensitive to bicuculline. Could this effect of diazepam and midazolam result from an endogenous release of GABA, which then acts on GABA, receptors? Release of GABA by benzodiazepines has been reported.4,23,3’In a few neurones midazolam and diazepam induced a small depolarization of the membrane potential associated with conductance increase; this was partially blocked by bicuculline and reversibly abolished by cobalt and high magnesium suggesting that release of GABA might be involved. However, the blockade of calcium spike by low doses of benzodiazepines was never associated with changes in membrane potential or conductance and several other findings suggest that this action of diazepam is independent of GABA. Some cells (including two S type cells) responded to diazepam and midazolam but not to exogenously applied GABA. In those neurones that respond to benzodiazepines and GABA, the potency of diazepam and midazolam in inhibiting the calcium spike was at least 100-1000 times higher than that of GABA. Ro 15-1788 and pentylenetetrazol blocked the effect of diazepam but not that of GABA. This action of benzodiazepines to reduce calcium spikes might underlie the inhibition of potassium-evoked noradrenaline release in synaptosomes of rat hippocampus.’ Since benzodiazepines inhibit adenosine uptake in central neurones,24.25 the possibility exists that the block of the calcium spike is mediated through adenosine receptors. Adenosine is known to reduce acetylcholine release in the guinea-pig ileumS and to shorten the calcium spike in rat superior cervical ganglion neurones.’ Adenosine antagonists or selective GABA, antagonists could be used to confirm this conclusion. The clinical effects of benzodiazepines in the treatment of anxiety, seizures and insomnia have been related to the enhancement of GABA-induced inhibition and associated chloride conductance increase. The benzodiazepine receptor has been considered as a part of a supramolecular complex of functionally linked membrane proteins including the GABA receptor and its associated ionophore.‘9 The present findings lend support to this notion, yet while strengthening this hypothesis, further suggest that some clinical effect of benzodiazepines may be also independent of GABA. They might result from a reduction of transmitter release. Acknowledgemenf-Supported
by U.S. Department Health and Human Services grant AM/NS 32979.
of
REFERENCES
1. Cherubini E. and North R. A. (1984) y-Aminobutyric acid actions on neurones of guinea-pig myenteric plexus. Br. J. Phurmac. 82, 93-100. 2. Cherubini E. and North R. A. (1984) Inhibition of calcium spikes and transmitter release by y-aminobutyric acid in the guinea-pig myenteric plexus. Br. J. Pharmac. 82, 101-106.
~~~ia~pines
on myenteric neurons
315
3. Choi D. W., Farb D. H. and Fischbach G. D. (1981) Chlordiazepoxide selectively potentiates GABA conductance of spinal cord and sensory neurons in cell culture. J. Neurophysiol. 45, 621-631. 4. Curtis D. R., Lodge D., Johnston G. A. R. and Brand S. J. (1976) Central actions of benzodiazepines. Brain Res. 118, 344-347. 5. Dowdle E. B. and Maske R. (1980) The effects of calcium concentration
on the inhibition of cholinergic neurotransmission in the myenteric plexus of guinea-pig ileum by adenosine nucleotides. iir. J. Pharmac. 71,245-252. 6. Fleckenstein A. (1977) Specific pharmacology of calcium in myocardium, cardiac pacemakers, and vascular smooth muscle. A. Rev. Pharmac. Toxicol. 17, 149-166. 7. Fung S. C. and Fillenz M. (1983) The role of presynaptic GABA and benzodiazepine receptors in the control of noradrenaline release in rat hippocampus. Neur&ci. -Lett. 42, 61-66. 8. Grafe P.. Maver C. J. and Wood J. D. (1980) Svnantic modulation of calcium-denendent notassium conductance in myenteric neurones in the guinea-pig. J.‘ Physioi, Land. 305, 235-248. 9. Henon B. K. and McAfee D. A. (1983) The ionic basis of adenosine receptor actions on post-ganglionic neurones in the rat. J. Physiol., Lond. 336, 607620. 10. Hirst G. D. S., Holman M. E. and Spence I. (1974) Two types of neurones in the myenteric plexus of duodenum in the guinea-pig. J. Physiof., Lond. 236, 303-326. 11. Hullihan J. P., Spector S., Taniguchi T. and Wang J. K. T. (1983) The binding of [)H]-diazepam to guinea-pig ileal longitudinal muscle and the in vitro inhibition of contraction by benzodiazepines. Br. J. Phartnac. 78, 321-327. 12. Hunkeler W., Mohler H., Pieri L., Pole P., Bonetti E. P., Cumin R., Schaffner R. and Halfely W. (1981) Selective antagonists of benzodiazepines. Nature 290, 514516. 13. Ishii K., Kano T., Akutagawa M., Makino M., Tanaka T. and Ando J. (1982) Effects of flurazepam and diazepam in isolated guinea-pig taenia coli and lon~tu~nal muscle. Eur. J. P~r~c. 83, 329-333. 14. Jessen K. R., Mirsky R., Dennison M. E. and Burnstock G. (1979) GABA may be a neurotransmitter in the vertebrate peripheral nervous system. Nature 281, 71-74. 1.5. Krantis A. and Kerr D. I. B. (1981) Autoradiographic localization of [3H]gamma-aminobutyric acid in the myenteric plexus of the guinea-pig small intestine. Neurosci. Lett. 23, 263-268. 16. Lee H. K., Dunwiddie T. W. and Hoffer B. J. (1979) Interaction of diazepam with synaptic transmission in the in vitro rat hippocampus. NQunyn-Schmeidebergs Arch. Pharmac. 309, 131-136. 17. Macdonald J. F. and Barker J. L. (1982) Multiple actions of picomolar concentrations of thtrazepam on the excitability of cultured mouse spinal neurons. Brain Res. 246, 257-264. 18. Macdonald R. and Barker J. L. (1978) Benzodiazepines specifically modulate GABA-mediated postsynaptic inhibition in cultured mammalian neurones. Nature 271, 563-564. 19. Mohler H. and Richards J. G. (1983) Receptors for anxiolytic drugs. In Anxiolytics: Neurochemical, Behavioral and Clinical Perspectives (eds Malick J. B., Enne S. J. and Yamamura H. I.), pp. 15-39. Raven Press, New York. 20. Nishi S. and North R. A. (1973) Intracellutar recording from the myenteric plexus of the guinea-pig ileum. J. Physiol., Lond. 231, 471-491. 21. Nistri A. and Berti C. (1983) Potentiating action of midazolam on GABA-rn~iat~ responses and its ~tagonism by Ro 14-7437 in the frog spinal cord. Neurosei. Lett. 39, 199-204. 22. Nistri A. and Constanti A. (1978) Effects of flurazepam on amino acid-evoked responses recorded from the lobster muscle and the frog spinal cord. Neuropharmacology 17, 127-135. 23. Olsen R. W., Lamar E. E. and Bayliss J. D. (1977) Calcium-induced release of y-aminobutyric acid from synaptosomes: effect of tranquilizer drugs. J. Neurochem. 28, 299305. 24. Phillis J. W. (1979) Diazepam potentiation of purinergic depression of central neurones. Can. J. Physiol. Pharmac. 57, 432-435. 25. Phillis J. W., Bender A. S. and Wu P. H. (1980) Benzodiazepines inhibit adenosine uptake into rat brain synaptosomes. Brain Res. 195, 494-498. 26. Riley M. and Scholfield C. N. (1983) Diazepam increases GABA mediated inhibition in the olfactory cortex slice. Pfiigers Arch. ges. Physiol. 397, 312-318. 27. Schlosser W. and Franc0 S. (1979) Modification of GABA-mediated depolarization of the cat ganglion by pentobarbitol and two benzodiazepines. Neuropharmacology 18, 377-381. 28. Scholheld C. N. (1983) Ro 1S-1788 is a potent antagonist of benzodiazepines in the olfactory cortex slice. Ppiigers Arch. ges. Physiol. 396, 292-296. 29. Simmons M. A. (1981) Distinction between the effects of barbiturates, benzodiazepines and phenytoin on responses to y-aminobutyric acid receptor activation and antagonism by bicuculline and picrotoxin. Br. J. Phurmac. 73, 739-747. 30. Squires R. F. and Braestrup C. (1977) Benzodiazepine receptors in rat brain. Nature 266, 732-734. 31. Suria A. and Costa E. (1975) Action of diazepam, dibutyryl cGMP and GABA on presynaptic nerve terminals in bullfrog sympathetic ganglion. Brain Res. 87, 102-106. 32. Tallman J. F.. Paul S. M.. Skolnick P. and Gallaeer D. W. (1980) Receptors for the age of anxiety: -_ _ - pharmacology _
of the benzodiazepines Science, N. Y. 207, 274-281. 33. White W. F., Dichtcr M. A. and Snodgrass S. R. (1981) Benzodiazepine binding and interactions with the GABA receptor complex in living cultures of rat cerebral cortex. Brain Res. 215, 162-176. (Accepted 25 July 1984)