Proc. Nat. Acad. Sci. USA Vol. 72, No. 4, pp. 1460-1463, April 1975
Presynaptic Action of Barbiturates in the Frog Spinal Cord (y-aminobutyric acid receptors/presynaptic inhibition/primary afferents)
ROGER A. NICOLL Department of Physiology, State University of New York at Buffalo, Buffalo, N.Y. 14214
Communicated by John C. Ecclem, February 3, 1975 The action of pentobarbital on primary ABSTRACT afferents of the isolated frog spinal cord was analyzed with sucrose gap and intracellular recording techniques. Pentobarbital in concentrations generally considered to be in the anesthetic range greatly prolonged presynaptic inhibition and also depolarized primary afferents. The depolarization was accompanied by an increase in excitability and resulted from activation of y-aminobutyric acid receptors, possibly by a direct action on these receptors, since the depolarization was reversibly blocked by ,y-aminobutyric acid, but not by glycine, antagonists, and magnesium ions. Furthermore, dorsal root ganglion cells exhibited a reduced sensitivity to both y-aminobutyric acid and pentobarbital after a "desensitizing" dose of 'yaminobutyric acid. The prolongation of presynaptic inhibition and the activation of y-aminobutyric acid receptors on primary afferents by pentobarbital should act to reduce the amount of transmitter released from the first synapse in sensory pathways.
Although the barbiturates are among the most widely used drugs, their mechanism of action in the central nervous system has remained largely enigmatic. The available evidence suggests that the barbiturates act mainly at synapses, suppressing excitatory synaptic transmission (1-8) and usually either preserving or augmenting inhibitory synaptic transmission (6, 9-19). The action of barbiturates on the spinal monosynaptic pathway has received the most rigorous analysis, and it has been shown that, in low doses, they reduce the amount of transmitter released from the primary afferents (4). The present study attempts to analyze the presynaptic action of barbiturates. In conjunction with sucrose gap recording, the isolated frog spinal cord preparation offers the advantages of precise control over the media and drug concentrations, as well as of stable recording of the membrane potential of primary afferents. METHODS Frogs (Rana pipiens) were chilled on ice and the spinal cord, with attached 8th and 9th roots, was removed, placed in a dissecting dish, and hemisected sagittally. The roots were freed of connective tissue, and the preparation was transferred to a sucrose gap chamber (20) and perfused with frog Ringer's solution. The Ringer's and drug solutions were passed through a thermoelectric cooling unit (Cambion), maintained at 8-14°, before entering the sucrose gap chamber. The Ringer's solution consisted of 115 mM NaCl, 2 mM CaCl2, 2 mM KCl, and 10 mM tris(hydroxymethyl)aminomethane buffered to pH 7.3 with HCl. To exclude indirect synaptic effects, we added 1 Abbreviation: GABA,y-aminobutyric acid. The abbreviation recommended by the International Union of Biochemistry is ly-Abu.
,uM tetrodotoxin or 20 mM MgSO4 to the Ringer's solution. The pH of the drug solutions was adjusted when necessary with NaOH or HCl. Low resistance microelectrodes (1-3 MO) filled with 2 M NaCl were used for focal recording of synaptic potentials. The same electrodes were also used to test the excitability of primary afferent terminals (21). The terminals were stimulated with submaximal pulses, and the antidromically conducted volley was recorded in mineral oil with bipolar platinum electrodes. The distal electrode was placed on the crushed end of the dorsal root. Since the action of pentobarbital required several minutes to develop, it was important to ensure that the position of the microelectrode did not shift during the experiment. This was accomplished by firmly pinning down the cord in the immediate vicinity of the microelectrode and obtaining control observations for at least 30 min. To relate the excitability changes to the depolarization of the dorsal root, we monitored potential changes in the neighboring dorsal root with sucrose gap recording. Rana catesbiana were used for intracellular recording from dorsal root ganglion cells. Microelectrodes (10-30 MO) filled with 2 M K acetate were used for intracellular recording. A bridge circuit was used to pass current through the microelectrode. RESULTS Addition of pentobarbital to the perfusate slightly increased (114% i 7 SD, n = 5) and greatly prolonged the dorsal root potentials, as shown in Fig. 1, Al and Bi for the dorsal root potentials elicited by ventral root stimulation (15-19). The prolongation was the most sensitive index of pentobarbital action, being demonstrable at concentrations as low as 5 AM and requiring approximately 10-15 min to reach a maximum effect. With concentrations of pentobarbital in the range found during surgical anesthesia (0.2 mM) (7), the initial increase in amplitude of dorsal root potential was followed by a decrease (62% 4- 12.4 SD, n = 5), although the duration continued to lengthen, reaching half-decay values of over four times normal. The prolongation of the dorsal root potential was associated with a lengthening of the synaptic current generated across the primary afferents, as reflected by the slow field potential recorded by a microelectrode in the vicinity of the primary afferents (Fig. 1, A2 and B2) (22). As might be expected, the presynaptic inhibition associated with the dorsal root potential was similarly prolonged by pentobarbital. This effect is evident from the diminution of the focal synaptic potential during the dorsal root potential elicited by ventral root stimulation (Fig. 1, A3 and B3) (22). This reduction of the synaptic potential is presumed to occur presynaptically since the inhibition has the same time course and sensitivity 1460
Proc. Nat. Acad. Sci. USA 72
Barbiturates and Primary Afferents
(1975)
to picrotoxin as does the dorsal root potential (unpublished observations). Amylobarbital, thiopental, and barbital also markedly prolonged dorsal root potentials, while barbituric acid and the convulsant barbiturate, 5-(2-cyclohexylideneethyl)-5-ethyl barbituric acid, had little effect on the duration. In addition to lengthening presynaptic inhibition, pentobarbital (>20 gM) also depolarized the primary afferents (Figs. 2 and 3). The response had a very slow time course, requiring 10-15 min to plateau and was not reduced by addition of magnesium ions (20 mM) or tetrodotoxin (1 1AM) to the Ringer's solution, which suppressed synaptic transmission. The depolarizing potency of pentobarbital was approximately equal to that of y-aminobutyric acid (GABA) and was associated with an increase in excitability of the primary afferents (Fig. 2). However, for concentrations of 1 mM or more of pentobarbital (Fig. 2B), the increase in excitability was followed by a fall in excitability below control levels, despite the maintained depolarization of the primary afferents. Amylobarbital, thiopental, barbital, and 5-(2-cyclohexylideneethyl)-5-ethyl barbituric acid also elicited a depolarization, while barbituric acid had no effect. It is well documented that a number of amino acids have a strong depolarizing action on primary afferents (23-27), and the action of amino-acid antagonists suggest that GABA and possibly taurine or g3-alanine are transmitters mediating dorsal root potentials. An interesting observation was that the GABA antagonists, picrotoxin and bicuculline, also just as effectively and reversibly blocked the depolarizing action of pentobarbital (Fig. 3A). This finding suggests that pentobarbital might be eliciting its effects by activating GABA receptors. This notion is strengthened by the observation that strychnine, which blocked fl-alanine responses, failed to block the pentobarbital (or GABA) response (Fig. 3B). The depolarizing response has been further characterized by intracellular recording from dorsal root ganglion cells. In
A1
6
EXCITABILITY OF DR TERMINALS
j
A1
1B1
CONTROL
PENTOBARBITAL
-
2 ^.
VR-SFP
3
200 msec 14 mV
2 / 2f
DR ALONE VREDR
1461
80 msec.4 mV
3
.^| PRE
FIG. 1. Effect of pentobarbital on presynaptic inhibition. (A) Top trace (Al) is the dorsal root potential elicited by ventral root stimulation (VR-DRP), the middle trace (A2) is the accompanying field potential recorded in the dorsal horn with an extracellular microelectrode (SFP, slow field potential), and lower trace (A3) shows the focally recorded pre- and postsynaptic response obtained from the same site, in response to stimulating the dorsal root. Superimposed on the control response ("DR alone") is the response obtained 100 msec after a conditioning ventral root stimulus. The calibration pulse is 0.5 mV and 1 msec. (B) The same responses elicited 20 min after addition of 50 ,uM pentobarbital. In all traces negativity is up, signifying depolarization.
general, dorsal root ganglion cells were less sensitive to pentobarbital than the primary afferent terminals, requiring approximately a 10-fold increase for the minimally effective concentration. The depolarization of ganglion cells by pentobarbital exhibited properties similar to those of the GABA response. Applications of GABA in concentrations greater than 2 X 10-4 M "desensitized" the ganglion cell membrane to a subsequent application of GABA and recovery required approximately 5-15 min, depending on the concentration of GABA used. Interestingly, the membrane was also less sensitive to pentobarbital after a "desensitizing" concentration
B1
POTENTIAL OF DR FIBERS
4
/
2
N
/ PENTOARITALI
o
i~ ~ o ~io
io TIME(min)
4'O
5
TIME(min)
FIG. 2. Effect of pentobarbital on the membrane potential and excitability of primary afferents. (A) Addition of 4 X 10-4 M pentobarbital reversibly increases the excitability of dorsal root terminals (Al), the excitability increase having the same time course as the depolarization of the neighboring dorsal root recorded simultaneously across the sucrose gap (A2). The inserts show samples of the antidromic volley at the times indicated. (B) Addition of 4 X 10-3 M pentobarbital causes an initial increase in excitability (Bi), followed by a decrease below the control values. B2 shows a pen recording of the depolarization recorded from the neighboring dorsal root, which is maintained throughout the application. Preparation in B was superfused with a Ringer's solution containing 20 mM magnesium sulfate.
1462
Proc. Nat. Acad. Sci. USA 72 (1976)
Physiology: Nicoll A1
GABA
BARB
B1 GLU
BALA
BARB
D
BARB
1
2
min
10
mv
STRYCHNINE ~~~~~~~~~~~~~~~~~~~~~~~~~~~~60
--
3 !l WASH (30 min)
A 90 100 110
4
120
~~~(90 mnim) C
-
_-
p~~~I
BARB
10m
~5 mV GABA
BARB
BARB
FIG. 3. Pharmacological characterization of the pentobarbital depolarization of primary afferents. Al shows the control responses to glutamate (GLU) (1 mM), GABA (1 mM), and pentobarbital (BARB) (4 mM). In A2 addition of picrotoxin (0.5 mM) for 7 min markedly reduces the GABA and pentobarbital response but not the glutamate response. Washing the preparation (A3 and A4) progressively restores the antagonized responses. In B, the addition of strychnine (0.2 mM) for 12 min abolishes the f-alanine response (BALA) but not the glutamate or pentobarbital response. The calibration is the same as in A. The responses in C were recorded intracellularly from a dorsal root ganglion cell; they show that immediately after a desensitizing application of GABA (1 mM), pentobarbital (1 mM) no longer elicits a response. The final response was obtained 16 min after the GABA application. The resting membrane potential was -70 mV. D shows that the pentobarbital (0.5 mM) depolarization of dorsal root ganglion cells is associated with a fall in membrane resistance, as indicated by the decreased voltage deflection induced by constant current (-1.5 nA) hyperpolarizing pulses. The resting membrane potential was -56 mV. In E, curves of current against voltage were obtained before (0) and during (0) the pentobarbital (1 mM) depolarization in another cell.
of GABA (Fig. 3C), and the time course for recovery was similar to that seen with GABA. The depolarization was associated with a large fall in membrane resistance (Fig. 3D), and a comparison of curves of current against potential in the presence and absence of pentobarbital suggests that the reversal potential for the pentobarbital depolarization is similar to that for GABA (27); in Fig. 2E the current-voltage curves intersect at -33 mV. DISCUSSION
The prolongation of synaptically induced presynaptic inhibition by a number of barbiturates, excluding the convulsant barbiturate 5-(2-cyclohexyldieneethyl)-5-ethyl barbituric acid, confirms previous results (15-19, 33) and would be expected to contribute to the depression of sensory input. This enhancement in the presence of barbiturates could result from the prolonged release of GABA (and possibly taurine), substances proposed as transmitters mediating presynaptic inhibition (23-27), an inhibition in the uptake or catabolism of these substances from the synaptic cleft (29), or an enhancement of the action of the transmitter. It remains to be determined how much the prolongation of presynaptic inhibition contributes to the depressant properties of barbiturates. The question arises as to whether the depolarizing action of barbiturates on primary afferents and the increase in terminal
excitability occur by a direct action on GABA receptors or indirectly by the release of GABA. The resistance of the response to magnesium ions does not entirely exclude an indirect action, since anesthetics have been proposed to release transmitter from the neuromuscular junction by a calcium-independent mechanism (28). However, pentobarbital has been shown to decrease, not increase, the efflux of [3H]GABA from brain slices (29); furthermore, the dorsal root ganglion, which contains no synapses and only small amounts of GABA (30), is also depolarized by pentobarbital. The observation that prolonged applications of 1 mM or more of pentobarbital eventually resulted in a fall in excitability, is presumably due to the local anesthetic action of pentobarbital, which occurs with these concentrations (31). The fact that investigators in previous studies have not seen an increase in excitability of primary afferents (2, 11) may be related to technical problems associated with a preparation in vivo, including slow time course of drug action and instability of the position of the stimulating microelectrode. At any rate, the depolarizing action of barbiturates on primary afferents could provide an explanation for a number of the presynaptic effects previously reported. Since pentobarbital activates the same receptors as the presynaptic inhibitory transmitter (i.e., GABA), it would be expected that the amplitude of the presynaptic spike (2, 5) and the release
Proc. Nat. Acad. Sci. USA 72
(1975)
of excitatory transmitter would be reduced (4). However, the convulsant barbiturate, 5-(2-cyclohexylideneethyl)-5-ethyl barbituric acid, also depolarized the primary afferents (32), and yet was shown to augment the monosynaptic reflex (33). Possibly the depolarization of motoneurons by the convulsant barbiturate (34) counteracts a presynaptic depressant effect on the monosynaptic reflex. It is unclear whether the depolarizing action of barbiturates on primary afferents also applies to nonprimary afferent terminals, which are not under presynaptic inhibitory control, since these synapses might not be expected to possess GABA receptors. For instance, pentobarbital actually increases the quantal content of the end plate potential at the neuromuscular junction (35). The pentobarbital-induced depression of excitatory transmission in the prepiriform cortex is not associated with any change in the size of the presynaptic spike or the excitability of the presynaptic fibers (7). In summary, the present results indicate that concentrations of pentobarbital similar to those found during surgical anesthesia greatly prolong the presynaptic inhibition of primary afferents and also depolarize primary afferents by activating GABA receptors. Such- actions on the first central synapse of sensory pathways may reduce the sensory input and contribute to the depressant properties of this agent. I thank Dr. H. Downes for his generous gift of 5-(2-cyclo-
hexylideneethyl)-5-ethyl barbituric acid. This work was supported by a grant from the National Institute of Neurological Diseases and Stroke, R01NB0822101, 2, 3, 4, and by generous research support by Dr. Henry C. and Bertha H. Buswell Fund to the author.
1. Larrabee, M. G. & Posternak, J. M. (1952) J. Neurophysiol. 15, 91-114. 2. L0yning, T., Oshima, T. & Yokota, J. (1964) J. Neurophysiol. 27, 408-428. 3. Somjen, G. (1967) Anesthesiology 28, 135-143. 4. Weakly, J. N. (1969) J. Physiol. (London) 204, 63-77. 5. Richens, A. (1969) Brit. J. Phamacol. 36, 312-328. 6. Nicoll, R. A. (1972) J. Physiol. (London) 223, 803-814. 7. Richards, C. D. (1972) J. Physiol. (London) 227, 749-767. 8. Barker, J. L. & Gainer, H. (1973) Science 182, 720-722. 9. Bloedel, J. R. & Roberts, W. J. (1969) J. Neurophyiol. 32, 75-84.
Barbiturates and Primary Afferents
1463
10. Larson, M. D. & Major, M. A. (1970) Brain Res. 21, 309311. 11. Galindo, A. (1969) J. Pharmacol. Exp. Ther. 169, 185-195. 12. Bremer, F. (1970) Electroenceph. Clin. Neurophysiol. 28, 1-16. 13. Veselyunene, M. A., Gutman, A. M. & Lesene, V. A. (1971) Farmakol. Toksikol. (Moscow) 5, 520-522. 14. Eccles, J. C., Faber, D. S. & TUbof6kovi, H. (1971) Brain Res. 25, 335-356. 15. Eccles, J. C. & Malcolm, J. L. (1946) J. Neurophysiol. 9, 139-159. 16. Eccles, J. C., Schmidt, R. F. & Willis, W. D. (1963) J. Physiol. 68, 500-530. 17. Schmidt, R. F. (1963) Pftugers Arch. Ges. Physiol. 277, 325-346. 18. Miyahara, J. T., Esplin, D. W. & Zablocka, B. (1966) J. Pharmacol. Exp. Ther. 154, 118-127. 19. Banna, N. R. & Jabbur, S. J. (1969) Int. J. Neuropharmacol. 8, 299-307. 20. Koketsu, K., Karczmar, A. G. & Kitamura, R. (1969) Int. J. Neuropharmacol. 8, 329-336. 21. Wall, P. D. (1958) J. Physiol. (London) 142, 1-21. 22. Glusman, S. & Rudomin, P. (1974) Exp. Neurol. 45, 474-490. 23. Tebecis, A. K. & Phillis, J. (1969) Comp. Biochem. Physiol. 28, 1303-1315. 24. Davidoff, R. A. (1972) Science 175, 331-333. 25. Barker, J. L. & Nicoll, R. A. (1972) Science 176, 1043-1045. 26. Nicoll, R. A. & Barker, J. L. (1973) Nature New Biol. 246, 224-225. 27. Nishi, S., Minota, S. & Karczmar, A. G. (1974) Neuropharmacology 13, 215-219. 28. Quastel, D. M. J., Hackett, J. T. & Cooke, D. (1971) Science 172, 1034-1036. 29. Cutler, R. W. P., Markowitz, D. & Dudzinski, D. S. (1974) Brain Res. 81, 189-197. 30. Roberts, P. J., Keen, P. & Mitchell, J. F. (1973) J. Neurochem. 21, 199-209. 31. Stainman, A. & Seeman, P. (1974) Can. J. Physiol. Pharmacol. 52, 535-556. 32. Downes, H. & Franz, D. N. (1971) J. Pharmacol. Exp Ther. 179, 660-670. 33. Downes, H. & Williams, J. K. (1969) J. Pharmacol. Exp Ther. 168, 283-289. 34. Downes, H., as quoted by Esplin, D. N. & ZablockaEsplin, B. (1969) in Basic Mechanisms of the Epilepsies, eds. Jasper, H. H., Ward, A. A. & Pope, A. (Little, Brown and Co., Boston), pp. 167-183. 35. Thomson, T. D. & Turkanis, S. E. (1973) Brit. J. Pharmacol. 48, 48-58.
Presynaptic Action of Barbiturates in the Frog Spinal Cord (y-aminobutyric acid receptors/presynaptic inhibition/primary afferents)
ROGER A. NICOLL Department of Physiology, State University of New York at Buffalo, Buffalo, N.Y. 14214
Communicated by John C. Ecclem, February 3, 1975 The action of pentobarbital on primary ABSTRACT afferents of the isolated frog spinal cord was analyzed with sucrose gap and intracellular recording techniques. Pentobarbital in concentrations generally considered to be in the anesthetic range greatly prolonged presynaptic inhibition and also depolarized primary afferents. The depolarization was accompanied by an increase in excitability and resulted from activation of y-aminobutyric acid receptors, possibly by a direct action on these receptors, since the depolarization was reversibly blocked by ,y-aminobutyric acid, but not by glycine, antagonists, and magnesium ions. Furthermore, dorsal root ganglion cells exhibited a reduced sensitivity to both y-aminobutyric acid and pentobarbital after a "desensitizing" dose of 'yaminobutyric acid. The prolongation of presynaptic inhibition and the activation of y-aminobutyric acid receptors on primary afferents by pentobarbital should act to reduce the amount of transmitter released from the first synapse in sensory pathways.
Although the barbiturates are among the most widely used drugs, their mechanism of action in the central nervous system has remained largely enigmatic. The available evidence suggests that the barbiturates act mainly at synapses, suppressing excitatory synaptic transmission (1-8) and usually either preserving or augmenting inhibitory synaptic transmission (6, 9-19). The action of barbiturates on the spinal monosynaptic pathway has received the most rigorous analysis, and it has been shown that, in low doses, they reduce the amount of transmitter released from the primary afferents (4). The present study attempts to analyze the presynaptic action of barbiturates. In conjunction with sucrose gap recording, the isolated frog spinal cord preparation offers the advantages of precise control over the media and drug concentrations, as well as of stable recording of the membrane potential of primary afferents. METHODS Frogs (Rana pipiens) were chilled on ice and the spinal cord, with attached 8th and 9th roots, was removed, placed in a dissecting dish, and hemisected sagittally. The roots were freed of connective tissue, and the preparation was transferred to a sucrose gap chamber (20) and perfused with frog Ringer's solution. The Ringer's and drug solutions were passed through a thermoelectric cooling unit (Cambion), maintained at 8-14°, before entering the sucrose gap chamber. The Ringer's solution consisted of 115 mM NaCl, 2 mM CaCl2, 2 mM KCl, and 10 mM tris(hydroxymethyl)aminomethane buffered to pH 7.3 with HCl. To exclude indirect synaptic effects, we added 1 Abbreviation: GABA,y-aminobutyric acid. The abbreviation recommended by the International Union of Biochemistry is ly-Abu.
,uM tetrodotoxin or 20 mM MgSO4 to the Ringer's solution. The pH of the drug solutions was adjusted when necessary with NaOH or HCl. Low resistance microelectrodes (1-3 MO) filled with 2 M NaCl were used for focal recording of synaptic potentials. The same electrodes were also used to test the excitability of primary afferent terminals (21). The terminals were stimulated with submaximal pulses, and the antidromically conducted volley was recorded in mineral oil with bipolar platinum electrodes. The distal electrode was placed on the crushed end of the dorsal root. Since the action of pentobarbital required several minutes to develop, it was important to ensure that the position of the microelectrode did not shift during the experiment. This was accomplished by firmly pinning down the cord in the immediate vicinity of the microelectrode and obtaining control observations for at least 30 min. To relate the excitability changes to the depolarization of the dorsal root, we monitored potential changes in the neighboring dorsal root with sucrose gap recording. Rana catesbiana were used for intracellular recording from dorsal root ganglion cells. Microelectrodes (10-30 MO) filled with 2 M K acetate were used for intracellular recording. A bridge circuit was used to pass current through the microelectrode. RESULTS Addition of pentobarbital to the perfusate slightly increased (114% i 7 SD, n = 5) and greatly prolonged the dorsal root potentials, as shown in Fig. 1, Al and Bi for the dorsal root potentials elicited by ventral root stimulation (15-19). The prolongation was the most sensitive index of pentobarbital action, being demonstrable at concentrations as low as 5 AM and requiring approximately 10-15 min to reach a maximum effect. With concentrations of pentobarbital in the range found during surgical anesthesia (0.2 mM) (7), the initial increase in amplitude of dorsal root potential was followed by a decrease (62% 4- 12.4 SD, n = 5), although the duration continued to lengthen, reaching half-decay values of over four times normal. The prolongation of the dorsal root potential was associated with a lengthening of the synaptic current generated across the primary afferents, as reflected by the slow field potential recorded by a microelectrode in the vicinity of the primary afferents (Fig. 1, A2 and B2) (22). As might be expected, the presynaptic inhibition associated with the dorsal root potential was similarly prolonged by pentobarbital. This effect is evident from the diminution of the focal synaptic potential during the dorsal root potential elicited by ventral root stimulation (Fig. 1, A3 and B3) (22). This reduction of the synaptic potential is presumed to occur presynaptically since the inhibition has the same time course and sensitivity 1460
Proc. Nat. Acad. Sci. USA 72
Barbiturates and Primary Afferents
(1975)
to picrotoxin as does the dorsal root potential (unpublished observations). Amylobarbital, thiopental, and barbital also markedly prolonged dorsal root potentials, while barbituric acid and the convulsant barbiturate, 5-(2-cyclohexylideneethyl)-5-ethyl barbituric acid, had little effect on the duration. In addition to lengthening presynaptic inhibition, pentobarbital (>20 gM) also depolarized the primary afferents (Figs. 2 and 3). The response had a very slow time course, requiring 10-15 min to plateau and was not reduced by addition of magnesium ions (20 mM) or tetrodotoxin (1 1AM) to the Ringer's solution, which suppressed synaptic transmission. The depolarizing potency of pentobarbital was approximately equal to that of y-aminobutyric acid (GABA) and was associated with an increase in excitability of the primary afferents (Fig. 2). However, for concentrations of 1 mM or more of pentobarbital (Fig. 2B), the increase in excitability was followed by a fall in excitability below control levels, despite the maintained depolarization of the primary afferents. Amylobarbital, thiopental, barbital, and 5-(2-cyclohexylideneethyl)-5-ethyl barbituric acid also elicited a depolarization, while barbituric acid had no effect. It is well documented that a number of amino acids have a strong depolarizing action on primary afferents (23-27), and the action of amino-acid antagonists suggest that GABA and possibly taurine or g3-alanine are transmitters mediating dorsal root potentials. An interesting observation was that the GABA antagonists, picrotoxin and bicuculline, also just as effectively and reversibly blocked the depolarizing action of pentobarbital (Fig. 3A). This finding suggests that pentobarbital might be eliciting its effects by activating GABA receptors. This notion is strengthened by the observation that strychnine, which blocked fl-alanine responses, failed to block the pentobarbital (or GABA) response (Fig. 3B). The depolarizing response has been further characterized by intracellular recording from dorsal root ganglion cells. In
A1
6
EXCITABILITY OF DR TERMINALS
j
A1
1B1
CONTROL
PENTOBARBITAL
-
2 ^.
VR-SFP
3
200 msec 14 mV
2 / 2f
DR ALONE VREDR
1461
80 msec.4 mV
3
.^| PRE
FIG. 1. Effect of pentobarbital on presynaptic inhibition. (A) Top trace (Al) is the dorsal root potential elicited by ventral root stimulation (VR-DRP), the middle trace (A2) is the accompanying field potential recorded in the dorsal horn with an extracellular microelectrode (SFP, slow field potential), and lower trace (A3) shows the focally recorded pre- and postsynaptic response obtained from the same site, in response to stimulating the dorsal root. Superimposed on the control response ("DR alone") is the response obtained 100 msec after a conditioning ventral root stimulus. The calibration pulse is 0.5 mV and 1 msec. (B) The same responses elicited 20 min after addition of 50 ,uM pentobarbital. In all traces negativity is up, signifying depolarization.
general, dorsal root ganglion cells were less sensitive to pentobarbital than the primary afferent terminals, requiring approximately a 10-fold increase for the minimally effective concentration. The depolarization of ganglion cells by pentobarbital exhibited properties similar to those of the GABA response. Applications of GABA in concentrations greater than 2 X 10-4 M "desensitized" the ganglion cell membrane to a subsequent application of GABA and recovery required approximately 5-15 min, depending on the concentration of GABA used. Interestingly, the membrane was also less sensitive to pentobarbital after a "desensitizing" concentration
B1
POTENTIAL OF DR FIBERS
4
/
2
N
/ PENTOARITALI
o
i~ ~ o ~io
io TIME(min)
4'O
5
TIME(min)
FIG. 2. Effect of pentobarbital on the membrane potential and excitability of primary afferents. (A) Addition of 4 X 10-4 M pentobarbital reversibly increases the excitability of dorsal root terminals (Al), the excitability increase having the same time course as the depolarization of the neighboring dorsal root recorded simultaneously across the sucrose gap (A2). The inserts show samples of the antidromic volley at the times indicated. (B) Addition of 4 X 10-3 M pentobarbital causes an initial increase in excitability (Bi), followed by a decrease below the control values. B2 shows a pen recording of the depolarization recorded from the neighboring dorsal root, which is maintained throughout the application. Preparation in B was superfused with a Ringer's solution containing 20 mM magnesium sulfate.
1462
Proc. Nat. Acad. Sci. USA 72 (1976)
Physiology: Nicoll A1
GABA
BARB
B1 GLU
BALA
BARB
D
BARB
1
2
min
10
mv
STRYCHNINE ~~~~~~~~~~~~~~~~~~~~~~~~~~~~60
--
3 !l WASH (30 min)
A 90 100 110
4
120
~~~(90 mnim) C
-
_-
p~~~I
BARB
10m
~5 mV GABA
BARB
BARB
FIG. 3. Pharmacological characterization of the pentobarbital depolarization of primary afferents. Al shows the control responses to glutamate (GLU) (1 mM), GABA (1 mM), and pentobarbital (BARB) (4 mM). In A2 addition of picrotoxin (0.5 mM) for 7 min markedly reduces the GABA and pentobarbital response but not the glutamate response. Washing the preparation (A3 and A4) progressively restores the antagonized responses. In B, the addition of strychnine (0.2 mM) for 12 min abolishes the f-alanine response (BALA) but not the glutamate or pentobarbital response. The calibration is the same as in A. The responses in C were recorded intracellularly from a dorsal root ganglion cell; they show that immediately after a desensitizing application of GABA (1 mM), pentobarbital (1 mM) no longer elicits a response. The final response was obtained 16 min after the GABA application. The resting membrane potential was -70 mV. D shows that the pentobarbital (0.5 mM) depolarization of dorsal root ganglion cells is associated with a fall in membrane resistance, as indicated by the decreased voltage deflection induced by constant current (-1.5 nA) hyperpolarizing pulses. The resting membrane potential was -56 mV. In E, curves of current against voltage were obtained before (0) and during (0) the pentobarbital (1 mM) depolarization in another cell.
of GABA (Fig. 3C), and the time course for recovery was similar to that seen with GABA. The depolarization was associated with a large fall in membrane resistance (Fig. 3D), and a comparison of curves of current against potential in the presence and absence of pentobarbital suggests that the reversal potential for the pentobarbital depolarization is similar to that for GABA (27); in Fig. 2E the current-voltage curves intersect at -33 mV. DISCUSSION
The prolongation of synaptically induced presynaptic inhibition by a number of barbiturates, excluding the convulsant barbiturate 5-(2-cyclohexyldieneethyl)-5-ethyl barbituric acid, confirms previous results (15-19, 33) and would be expected to contribute to the depression of sensory input. This enhancement in the presence of barbiturates could result from the prolonged release of GABA (and possibly taurine), substances proposed as transmitters mediating presynaptic inhibition (23-27), an inhibition in the uptake or catabolism of these substances from the synaptic cleft (29), or an enhancement of the action of the transmitter. It remains to be determined how much the prolongation of presynaptic inhibition contributes to the depressant properties of barbiturates. The question arises as to whether the depolarizing action of barbiturates on primary afferents and the increase in terminal
excitability occur by a direct action on GABA receptors or indirectly by the release of GABA. The resistance of the response to magnesium ions does not entirely exclude an indirect action, since anesthetics have been proposed to release transmitter from the neuromuscular junction by a calcium-independent mechanism (28). However, pentobarbital has been shown to decrease, not increase, the efflux of [3H]GABA from brain slices (29); furthermore, the dorsal root ganglion, which contains no synapses and only small amounts of GABA (30), is also depolarized by pentobarbital. The observation that prolonged applications of 1 mM or more of pentobarbital eventually resulted in a fall in excitability, is presumably due to the local anesthetic action of pentobarbital, which occurs with these concentrations (31). The fact that investigators in previous studies have not seen an increase in excitability of primary afferents (2, 11) may be related to technical problems associated with a preparation in vivo, including slow time course of drug action and instability of the position of the stimulating microelectrode. At any rate, the depolarizing action of barbiturates on primary afferents could provide an explanation for a number of the presynaptic effects previously reported. Since pentobarbital activates the same receptors as the presynaptic inhibitory transmitter (i.e., GABA), it would be expected that the amplitude of the presynaptic spike (2, 5) and the release
Proc. Nat. Acad. Sci. USA 72
(1975)
of excitatory transmitter would be reduced (4). However, the convulsant barbiturate, 5-(2-cyclohexylideneethyl)-5-ethyl barbituric acid, also depolarized the primary afferents (32), and yet was shown to augment the monosynaptic reflex (33). Possibly the depolarization of motoneurons by the convulsant barbiturate (34) counteracts a presynaptic depressant effect on the monosynaptic reflex. It is unclear whether the depolarizing action of barbiturates on primary afferents also applies to nonprimary afferent terminals, which are not under presynaptic inhibitory control, since these synapses might not be expected to possess GABA receptors. For instance, pentobarbital actually increases the quantal content of the end plate potential at the neuromuscular junction (35). The pentobarbital-induced depression of excitatory transmission in the prepiriform cortex is not associated with any change in the size of the presynaptic spike or the excitability of the presynaptic fibers (7). In summary, the present results indicate that concentrations of pentobarbital similar to those found during surgical anesthesia greatly prolong the presynaptic inhibition of primary afferents and also depolarize primary afferents by activating GABA receptors. Such- actions on the first central synapse of sensory pathways may reduce the sensory input and contribute to the depressant properties of this agent. I thank Dr. H. Downes for his generous gift of 5-(2-cyclo-
hexylideneethyl)-5-ethyl barbituric acid. This work was supported by a grant from the National Institute of Neurological Diseases and Stroke, R01NB0822101, 2, 3, 4, and by generous research support by Dr. Henry C. and Bertha H. Buswell Fund to the author.
1. Larrabee, M. G. & Posternak, J. M. (1952) J. Neurophysiol. 15, 91-114. 2. L0yning, T., Oshima, T. & Yokota, J. (1964) J. Neurophysiol. 27, 408-428. 3. Somjen, G. (1967) Anesthesiology 28, 135-143. 4. Weakly, J. N. (1969) J. Physiol. (London) 204, 63-77. 5. Richens, A. (1969) Brit. J. Phamacol. 36, 312-328. 6. Nicoll, R. A. (1972) J. Physiol. (London) 223, 803-814. 7. Richards, C. D. (1972) J. Physiol. (London) 227, 749-767. 8. Barker, J. L. & Gainer, H. (1973) Science 182, 720-722. 9. Bloedel, J. R. & Roberts, W. J. (1969) J. Neurophyiol. 32, 75-84.
Barbiturates and Primary Afferents
1463
10. Larson, M. D. & Major, M. A. (1970) Brain Res. 21, 309311. 11. Galindo, A. (1969) J. Pharmacol. Exp. Ther. 169, 185-195. 12. Bremer, F. (1970) Electroenceph. Clin. Neurophysiol. 28, 1-16. 13. Veselyunene, M. A., Gutman, A. M. & Lesene, V. A. (1971) Farmakol. Toksikol. (Moscow) 5, 520-522. 14. Eccles, J. C., Faber, D. S. & TUbof6kovi, H. (1971) Brain Res. 25, 335-356. 15. Eccles, J. C. & Malcolm, J. L. (1946) J. Neurophysiol. 9, 139-159. 16. Eccles, J. C., Schmidt, R. F. & Willis, W. D. (1963) J. Physiol. 68, 500-530. 17. Schmidt, R. F. (1963) Pftugers Arch. Ges. Physiol. 277, 325-346. 18. Miyahara, J. T., Esplin, D. W. & Zablocka, B. (1966) J. Pharmacol. Exp. Ther. 154, 118-127. 19. Banna, N. R. & Jabbur, S. J. (1969) Int. J. Neuropharmacol. 8, 299-307. 20. Koketsu, K., Karczmar, A. G. & Kitamura, R. (1969) Int. J. Neuropharmacol. 8, 329-336. 21. Wall, P. D. (1958) J. Physiol. (London) 142, 1-21. 22. Glusman, S. & Rudomin, P. (1974) Exp. Neurol. 45, 474-490. 23. Tebecis, A. K. & Phillis, J. (1969) Comp. Biochem. Physiol. 28, 1303-1315. 24. Davidoff, R. A. (1972) Science 175, 331-333. 25. Barker, J. L. & Nicoll, R. A. (1972) Science 176, 1043-1045. 26. Nicoll, R. A. & Barker, J. L. (1973) Nature New Biol. 246, 224-225. 27. Nishi, S., Minota, S. & Karczmar, A. G. (1974) Neuropharmacology 13, 215-219. 28. Quastel, D. M. J., Hackett, J. T. & Cooke, D. (1971) Science 172, 1034-1036. 29. Cutler, R. W. P., Markowitz, D. & Dudzinski, D. S. (1974) Brain Res. 81, 189-197. 30. Roberts, P. J., Keen, P. & Mitchell, J. F. (1973) J. Neurochem. 21, 199-209. 31. Stainman, A. & Seeman, P. (1974) Can. J. Physiol. Pharmacol. 52, 535-556. 32. Downes, H. & Franz, D. N. (1971) J. Pharmacol. Exp Ther. 179, 660-670. 33. Downes, H. & Williams, J. K. (1969) J. Pharmacol. Exp Ther. 168, 283-289. 34. Downes, H., as quoted by Esplin, D. N. & ZablockaEsplin, B. (1969) in Basic Mechanisms of the Epilepsies, eds. Jasper, H. H., Ward, A. A. & Pope, A. (Little, Brown and Co., Boston), pp. 167-183. 35. Thomson, T. D. & Turkanis, S. E. (1973) Brit. J. Pharmacol. 48, 48-58.
