- Email: [email protected]
Effects of GABA and glycine on sympathetic preganglionic neurons in the upper thoracic intermediolateral nucleus of the cat
Brain Research, 277 (1983) 365--369 Elsevier
365
Effects of GABA and glycine on sympathetic preganglionic neurons in the upper thoracic intermediolateral nucleus of the cat S. B. BACKMAN and J. L. HENRY* Department of Physiology, McGill University, 3655 Drummond Street, Montreal, Quebec H3G 1 Y6 (Canada)
(Accepted June 14th, 1983) Key words: sympathetic preganglionic neurons - - GABA - - glycine- - bicuculline - - strychnine
GABA (5-137 nA) and glycine (5-75 nA) each inhibit spontaneous activity and block antidromic invasion of the soma-dendritic region of single sympathetic preganglionic neurons (SPNs) in the intermediolateral nucleus of T1-T3 in the cat. These effects are rapid in onset and recovery. They are selectively blocked by bicuculline and strychnine respectively. Thus, GABA and glycine exert pharmacologicallyspecific inhibitory effects on SPNs and this supports the possibility that they may be chemical mediators of inhibitory inputs directly onto these neurons. G A B A has been implicated in control of sympathetic output: picrotoxin (0.1-0.4 mg/kg, i.v.) abolishes the inhibitory effect of electrical stimulation of visceral and of somatic afferents on spontaneous firing of sympathetic preganglionic neurons (SPNs) 13. Glycine has been implicated in possible recurrent inhibition of these neurons: strychnine nitrate (0.2-0.3 mg/kg i.v.) abolishes the inhibitory effect of the antidromic activation of one pool of SPNs on the orthodromic activation of another pool 12. In this study we determined specific effects of G A B A and glycine on clearly identified single SPNs located within the intermediolateral nucleus of segments T1-T3. Details of the methodology are provided elsewhere (Backman and Henry, in preparation). Briefly, cats (2-4 kg) were anesthetized with a-chloralose (60 mg/kg, i.v.), paralyzed with pancuronium bromide (Pavulon, Organon, 1 mg/kg, i.v.) and ventilated artificially after bilateral pneumothorax. Endtidal C02 concentration, arterial pressure and core temperature were monitored and maintained within normal physiological limits. Spinal segments T1-T3 were exposed by laminectomy. The sympathetic chain on the right side was exposed at the level of the stellate ganglion by a retropleural approach. Preganglionic axons were activated electrically using bipolar * To whom all correspondence should be addressed. 0006-8993/83/$03.00 © 1983 Elsevier Science Publishers B.V.
silver hook electrodes placed under the sympathetic chain just caudal to the stellate ganglion. These electrodes were connected via a Grass PSIU6 photoelectric constant current stimulus isolation unit to a Grass $88 stimulator which delivered rectangular pulses of 0.1-1.0~A and 0.5 ms duration. Extracellular single unit spikes were recorded using multibarrelled micropipettes with overall tip diameters of 5-10/~m. The recording barrel, filled with 2.7 M NaC1, was connected via an AC-coupled unity gain high impedance headstage to a Tektronix RM 122 preamplifier which in turn was connected to Tektronix 5111 and 565 oscilloscopes for filming evoked responses and continuous activity. The output from the 565 oscilloscope was connected to a mean frequency counter, the output of which was displayed on a polygraph as a continuous time histogram of spikes per unit time. Remaining barrels, for iontophoresis, were filled with G A B A (1 M, pH 4.0), glycine (1 M, pH 3.5), bicuculline methiodide (10 mM in 165 mM NaCI, pH 3.2), strychnine sulfate (10 mM, pH 5.0) and sodium-L-glutamate (1 M, pH 7.4). Negative current was used to eject glutamate; positive current was used to eject the other substances. Retaining currents (12 nA) of opposite polarity were applied to each barrel to counteract spontaneous efflux by diffusion. Accurate anatomical localization of the elec-
366 trode tip was obtained by ejection of Pontamine Sky Blue (Gurr, 2% in 0.5 M sodium acetate) following the protocol described by Henry7. Classification of neurons. Strict criteria were applied to ensure that spikes included in the results were recorded from the cell bodies of SPNs in the intermediolateral nucleus. These criteria were as follows. A unit was classified as sympathetic preganglionic if it could be activated by an antidromically propagated spike generated by electrical stimulation of the sympathetic chain. This was considered to have occurred if the response consisted of an all-ornone spike of invariant latency, if the spike followed high-frequency stimulation (tested using paired stimuli at short interstimulus intervals), and if, in the case of units demonstrating spontaneous or glutamate-induced activity, the evoked spike was cancelled by an orthodromically propagated spike (the collision test). All units included in this study were in the intermediolateral nucleus, as determined by the location of the Pontamine Sky Blue deposits. An effect of G A B A or glycine was considered a genuine response if it was repeatable, reversible and not mimicked by control current applied either through the glutamate-
containing barrel, or through barrels containing antagonists. Effects of G A B A and glycine on spontaneously active SPNs. Iontophoretic application of G A B A (~ = 42 ± 37 (S.D.) nA, n = 11 neurones) and of glycine (~ = 42 ± 22 nA, n = 8) decreased or abolished spontaneous activity in all cases (see Fig. 1). They also reduced the glutamate-induced excitation of the one neuron tested. Responses to both amino acids were typically rapid in onset, with a latency of less than 1 s; recovery followed a similarly rapid time course. Responses to G A B A and to glycine were indistinguishable when similar ejecting currents and period of application were used. Effects of G A B A and glycine on silent SPNs. Silent SPNs were tested in case they differed from spontaneously active SPNs on the basis of their sensitivities to these amino acids. Testing the effects of G A B A and glycine on the antidromic spike also provided a means of determining whether the effects of G A B A and glycine were postsynaptic (i.e. acting directly on the SPN itself). Invasion of the soma-dendritic region by an antidromically propagated spike was blocked by G A B A
8
A
O~ GABA 25
~ 25
slrychnlne
50 k
Gly
GABA
B
/
GABA 125
Gry 25
Dacucullpne
50
Fig. 1. Ratemeter records obtained from a spontaneously active SPN to illustrate the inhibitory, pharmacologically specific effects of GABA and glycine. In the continuous records shown in A. iontophoretic application of GABA and glycine depressed the rate of spontaneous discharge, lontophoretic application of strychnine antagonized the glycine-induced inhibition but not the GABA-induced inhibition. In B. iontophoretic application of bicuculline antagonized the inhibitory response to GABA but not that to glycine. Ejection currents are in nA.
367
I
A
[ CONT.RESR i00 I
POST
STRYCH.
C
GABA TS~A
GABA~~f/~4~75nA
I00
GLY.t2SnA
.
GABA37hA
GLY.37nA
GABA37hA
GLY 37hA
~
I0 msec
~
~
i Fig. 2. Oscilloscope records from 2 silent sympathetic preganglionic neurons. A and B illustrate the selective effects of GABA and glycine on antidromic activation. During each oscilloscope sweep, an electrical stimulus was applied to the sympathetic chain (arrow denotes stimulus artifact). A: top panel, from left to right records show antidromic extracellular single unit spike as control response, during iontophoretic application of GABA (75 nA) and during iontophoretic application of glycine (125 hA); middle panel, from left to right records show antidromic spike during application of bicuculline (125 nA); the effect of GABA has been abolished while glycine is still effective; bottom panel, following termination of bicuculline application, the response to GABA has recovered. B: top panel, from left to right records show antidromic extracellular single unit spike as control response, during iontophoretic application of GABA (75 nA) and during iontophoretic application of glycine (75 hA); middle panel, from left to right records show antidromic spike during application of strychnine (75 hA), the effect of glycine has been abolished while GABA is still effective; bottom panel, following termination of strychnine application, the response to glycine has recovered. C illustrates that GABA and glycine prevent antidromic invasion of the soma-dendritic region but not the initial segment; upper panel, from left to right records show antidromic extracellular single unit spike as control response, during iontophoretic application of GABA (37 nA) and during iontophoretic application of glycine (37 nA); lower panel, from left to right records show antidromic spike during application of bicuculline (37 hA), the effect of GABA is antagonized while glycine is still effective. Note, however, the increased A-B delay during GABA application.
368 (x = 36 + 21 nA, n = 21) and by glycine (:~ = 40.6 + 30.6 nA, n = 20). Typical records are shown in Fig. 2. Few neurons demonstrated the inflection on the rising phase of the antidromic spike which indicates an A - B break 2. In cases where a notch was observed recording rarely lasted for more than a few minutes, usually being terminated with the sudden brief appearance of a high-frequency injury discharge. These events suggest that the electrode tip was in close proximity to or in contact with the membrane of the cell. Accordingly, these neurons were particularly difficult to test for their sensitivity to iontophoretic application of chemical substances owing to large artifactual changes in spike shape and amplitude upon application of current. Nevertheless, 3 such neurons demonstrated clear responses to iontophoretic application of G A B A and glycine. Records from one are shown in Fig. 2C. G A B A and glycine typically abolished the second component of the spike, leaving the first relatively unaffected. Effects of bicuculline and strychnine on amino acidinduced inhibition. The inhibitory effects of G A B A and glycine on the rate of spontaneous discharge were selectively antagonized by bicuculline (n = 5 neurons) and by strychnine (n = 4 neurons), respectively, as shown in Fig. 1. Some differences were apparent between the effects of bicuculline on the response to G A B A and those of strychnine on the response to glycine. Currents required for bicuculline to produce antagonism were 1-7 times (2 = 3.4 + 2.9) those of G A B A whereas strychnine produced antagonism with currents of only 0.5-2 times (2 = 1.4 _+ 0.6) those of glycine. The effects of strychnine were also typically more prolonged than those of bicuculline. Recovery of the response to glycine was observed within 120-630 s (2 = 410 + 262) while recovery of the response to G A B A within 18-70 s (~ = 45 + 25). With 4 of 5 neurons, iontophoretic application of bicuculline alone resulted in a transient increase in the rate of discharge, whereas strychnine did so with only 1 of 4 neurons. Bicuculline and strychnine also antagonized the inhibitory effects of G A B A (n = 8 neurons) and glycine (n = 6 neurons) on invasion of the soma-dendritic region by an antidromically propagated spike (see Fig. 2). As with spontaneously active neurons, greater ejecting currents were required for bicucul-
line to antagonize the effects of G A B A (ejecting currents for bicuculline were 1-5 times, 2 = 2.0 + 1.3 those of G A B A ) than were required for strychnine to antagonize the effects of glycine (ejecting currents for strychnine were 0.5-1.5 times, 2 = 1.0 + 0.3 those of glycine). Effects of strychnine were again more prolonged than those of bicuculline: recovery of the response to glycine occurred within 120-330 sec (2 = 232 + 99 s) while recovery of the response to G A B A occurred within 14-390 s (2 = 116 + 182 s). It should be noted that, with regard to recovery of the response to G A B A , while one neuron demonstrated a long recovery time of 390 s the remaining neurons demonstrated recovery times under 35 s. Neither bicuculline nor strychnine induced a silent neuron to discharge spontaneously, even when high ejecting currents (i.e greater than 100 nA) were used. With respect to the neuron showing an A - B break, depicted in Fig. 2C, iontophoretic application of bicuculline antagonized the effects of G A B A but not those of glycine. However, it appears that during the application of bicuculline, a partial response to G A B A was still evident as indicated by the increase in A - B delay. This study has demonstrated that G A B A and glycine have transient inhibitory effects on both silent and spontaneously active SPNs in the intermediolateral nucleus. These effects are selectively blocked by bicuculline and strychnine, respectively, suggesting that the responses were elicited via activation of pharmacologically specific receptors. A postsynaptic site of action of G A B A and of glycine, specifically on the soma or proximal dendrites, is suggested by the observation that they blocked the invasion of the soma-dendritic region by an antidromically propagated spike, and reduced the glutamate-induced excitation of the one neuron tested. This is consistent with an earlier report that inputs to somata of sympathetic preganglionic neurons are predominantly inhibitoryS. A postsynaptic action does not preclude the possibility that G A B A and glycine may also have acted indirectly. G A B A and glycine were roughly equipotent when similar amounts were applied, as has been reported for sacral parasympathetic neuronsS. Glycine has a significantly greater inhibitory effect than GABAll on somatic motoneurons while G A B A has a greater inhibitory effect compared to glycine on dorsal horn
369 neuronsl, and therefore SPNs may be best visualized as interneurons on the m o t o r side of spinal reflex pathways. The observation that bicuculline is not as p o t e n t an antagonist of G A B A as strychnine is ot glycine is consistent with earlier studies on somatic motoneurons8 and on neurons in the cuneate nucleus 9, and implies a relatively lower affinity of bicuculline for G A B A receptors than strychnine for glycine receptors. The increased rate of discharge of SPNs with bicuculline suggests that it may have some direct excitatory effect on sympathetic preganglionic neurons (see also refs. 6 and 8 on cerebral cortical neurons, ref. 4 on spinal neurons and ref. 10 on cuneate neurons). In summary, these data indicate that G A B A and glycine mediate powerful, pharmacologically specific
1 Backman, S. B. and Henry, J. L., Responses of thoracic dorsal horn units to visceral inputs and to amino acids in the cat, Soc. Neurosci. Abstr., 6 (1980) 490. 2 Brock, L. G., Coombs, J. S. and Eccles, J. C., Intracellular recording from antidromically activated motoneurones, J. Physiol. (Lond.), 122 (1953) 429-461. 3 Chung, K., LaVelle, F. W. and Wurster, R. D., Ultrastructure of HRP-identified sympathetic preganglionic neurons in cats, J. comp. Neurol., 190 (1980) 147-155. 4 Curtis, D. R., Duggan, A. W., Felix, D. and Johnston, G. A. R., GABA, bicuculline and central inhibition, Nature (Lond.), 226 (1970) 1222-1224. 5 De Groat, W. C., The effects of glycine, GABA and strychnine on sacral parasympathetic preganglionic neurones, Brain Research, 18 (1970) 542-544. 6 Godfraind, J. M., Krnjevic, K. and Pumain, R., Doubtful value of bicuculline as a specific antagonist of GABA, Nature (Lond.), 228 (1970) 675-676. 7 Henry, J. L., Effects of substance P on functionally identified units in cat spinal cord, Brain Research, 114 (1976) 439-451.
and direct inhibition of sympathetic preganglionic neurons and support the possibility that G A B A and glycine may be chemical m e d i a t o r s of inhibitory synaptic inputs which have a n o r m a l physiological role in the control of excitability of sympathetic preganglionic neurons. S u p p o r t e d by the Q u 6 b e c H e a r t F o u n d a t i o n . S.B.B. is a Trainee of the C a n a d i a n H e a r t F o u n d a tion. Expert photographic assistance was provided by Mr. N. Schestakowich. The authors are i n d e b t e d to Ms. A. Cairns for typing the manuscript. The following gifts are gratefully acknowledged: halothane (Somnothane) from Hoechst Pharmaceuticals, Montr6al; p a n c u r o n i u m b r o m i d e (Pavulon) from Organon, West Hill, Ontario.
8 Hill, R. G., Simmonds, M. A. and Straughan, D. W., A comparative study of some convulsant substances as 7-aminobutyric acid antagonists in the feline cerebral cortex, Brit. J. Pharmacol., 49 (1973) 37-51. 9 Hill, R. G., Simmonds, M. A. and Straughan, D. W., Antagonism of 7-aminobutyric acid and glycine by convulsants in the cuneate nucleus of cat, Brit. J. Pharmacol., 56 (1976) 9-19. 10 Kelly, J. S. and Renaud, L. P., On the pharmacology of 7-aminobutyric acid receptors on the cuneo-thalamic relay cells of the cat, Brit. J. Pharmacol., 48 (1973) 369-386. 11 Krnjevic, K., Puil, E. and Werman, R., GABA and glycine actions on spinal motoneurons, Canad. J. Physiol. Pharmacol., 55 (1977) 658-669. 12 Lebedev, V. P., Petrov, V. I. and Skobelev, V. A., Do sympathetic preganglionic neurones have a recurrent inhibitory mechanism? Pflagers Arch., 383 (1980) 91-92. 13 Wyszogrodski, I., Central Inhibition in the Sympathetic Nervous System, Ph. D. Thesis, McGill University, Montr6al, 1972.
365
Effects of GABA and glycine on sympathetic preganglionic neurons in the upper thoracic intermediolateral nucleus of the cat S. B. BACKMAN and J. L. HENRY* Department of Physiology, McGill University, 3655 Drummond Street, Montreal, Quebec H3G 1 Y6 (Canada)
(Accepted June 14th, 1983) Key words: sympathetic preganglionic neurons - - GABA - - glycine- - bicuculline - - strychnine
GABA (5-137 nA) and glycine (5-75 nA) each inhibit spontaneous activity and block antidromic invasion of the soma-dendritic region of single sympathetic preganglionic neurons (SPNs) in the intermediolateral nucleus of T1-T3 in the cat. These effects are rapid in onset and recovery. They are selectively blocked by bicuculline and strychnine respectively. Thus, GABA and glycine exert pharmacologicallyspecific inhibitory effects on SPNs and this supports the possibility that they may be chemical mediators of inhibitory inputs directly onto these neurons. G A B A has been implicated in control of sympathetic output: picrotoxin (0.1-0.4 mg/kg, i.v.) abolishes the inhibitory effect of electrical stimulation of visceral and of somatic afferents on spontaneous firing of sympathetic preganglionic neurons (SPNs) 13. Glycine has been implicated in possible recurrent inhibition of these neurons: strychnine nitrate (0.2-0.3 mg/kg i.v.) abolishes the inhibitory effect of the antidromic activation of one pool of SPNs on the orthodromic activation of another pool 12. In this study we determined specific effects of G A B A and glycine on clearly identified single SPNs located within the intermediolateral nucleus of segments T1-T3. Details of the methodology are provided elsewhere (Backman and Henry, in preparation). Briefly, cats (2-4 kg) were anesthetized with a-chloralose (60 mg/kg, i.v.), paralyzed with pancuronium bromide (Pavulon, Organon, 1 mg/kg, i.v.) and ventilated artificially after bilateral pneumothorax. Endtidal C02 concentration, arterial pressure and core temperature were monitored and maintained within normal physiological limits. Spinal segments T1-T3 were exposed by laminectomy. The sympathetic chain on the right side was exposed at the level of the stellate ganglion by a retropleural approach. Preganglionic axons were activated electrically using bipolar * To whom all correspondence should be addressed. 0006-8993/83/$03.00 © 1983 Elsevier Science Publishers B.V.
silver hook electrodes placed under the sympathetic chain just caudal to the stellate ganglion. These electrodes were connected via a Grass PSIU6 photoelectric constant current stimulus isolation unit to a Grass $88 stimulator which delivered rectangular pulses of 0.1-1.0~A and 0.5 ms duration. Extracellular single unit spikes were recorded using multibarrelled micropipettes with overall tip diameters of 5-10/~m. The recording barrel, filled with 2.7 M NaC1, was connected via an AC-coupled unity gain high impedance headstage to a Tektronix RM 122 preamplifier which in turn was connected to Tektronix 5111 and 565 oscilloscopes for filming evoked responses and continuous activity. The output from the 565 oscilloscope was connected to a mean frequency counter, the output of which was displayed on a polygraph as a continuous time histogram of spikes per unit time. Remaining barrels, for iontophoresis, were filled with G A B A (1 M, pH 4.0), glycine (1 M, pH 3.5), bicuculline methiodide (10 mM in 165 mM NaCI, pH 3.2), strychnine sulfate (10 mM, pH 5.0) and sodium-L-glutamate (1 M, pH 7.4). Negative current was used to eject glutamate; positive current was used to eject the other substances. Retaining currents (12 nA) of opposite polarity were applied to each barrel to counteract spontaneous efflux by diffusion. Accurate anatomical localization of the elec-
366 trode tip was obtained by ejection of Pontamine Sky Blue (Gurr, 2% in 0.5 M sodium acetate) following the protocol described by Henry7. Classification of neurons. Strict criteria were applied to ensure that spikes included in the results were recorded from the cell bodies of SPNs in the intermediolateral nucleus. These criteria were as follows. A unit was classified as sympathetic preganglionic if it could be activated by an antidromically propagated spike generated by electrical stimulation of the sympathetic chain. This was considered to have occurred if the response consisted of an all-ornone spike of invariant latency, if the spike followed high-frequency stimulation (tested using paired stimuli at short interstimulus intervals), and if, in the case of units demonstrating spontaneous or glutamate-induced activity, the evoked spike was cancelled by an orthodromically propagated spike (the collision test). All units included in this study were in the intermediolateral nucleus, as determined by the location of the Pontamine Sky Blue deposits. An effect of G A B A or glycine was considered a genuine response if it was repeatable, reversible and not mimicked by control current applied either through the glutamate-
containing barrel, or through barrels containing antagonists. Effects of G A B A and glycine on spontaneously active SPNs. Iontophoretic application of G A B A (~ = 42 ± 37 (S.D.) nA, n = 11 neurones) and of glycine (~ = 42 ± 22 nA, n = 8) decreased or abolished spontaneous activity in all cases (see Fig. 1). They also reduced the glutamate-induced excitation of the one neuron tested. Responses to both amino acids were typically rapid in onset, with a latency of less than 1 s; recovery followed a similarly rapid time course. Responses to G A B A and to glycine were indistinguishable when similar ejecting currents and period of application were used. Effects of G A B A and glycine on silent SPNs. Silent SPNs were tested in case they differed from spontaneously active SPNs on the basis of their sensitivities to these amino acids. Testing the effects of G A B A and glycine on the antidromic spike also provided a means of determining whether the effects of G A B A and glycine were postsynaptic (i.e. acting directly on the SPN itself). Invasion of the soma-dendritic region by an antidromically propagated spike was blocked by G A B A
8
A
O~ GABA 25
~ 25
slrychnlne
50 k
Gly
GABA
B
/
GABA 125
Gry 25
Dacucullpne
50
Fig. 1. Ratemeter records obtained from a spontaneously active SPN to illustrate the inhibitory, pharmacologically specific effects of GABA and glycine. In the continuous records shown in A. iontophoretic application of GABA and glycine depressed the rate of spontaneous discharge, lontophoretic application of strychnine antagonized the glycine-induced inhibition but not the GABA-induced inhibition. In B. iontophoretic application of bicuculline antagonized the inhibitory response to GABA but not that to glycine. Ejection currents are in nA.
367
I
A
[ CONT.RESR i00 I
POST
STRYCH.
C
GABA TS~A
GABA~~f/~4~75nA
I00
GLY.t2SnA
.
GABA37hA
GLY.37nA
GABA37hA
GLY 37hA
~
I0 msec
~
~
i Fig. 2. Oscilloscope records from 2 silent sympathetic preganglionic neurons. A and B illustrate the selective effects of GABA and glycine on antidromic activation. During each oscilloscope sweep, an electrical stimulus was applied to the sympathetic chain (arrow denotes stimulus artifact). A: top panel, from left to right records show antidromic extracellular single unit spike as control response, during iontophoretic application of GABA (75 nA) and during iontophoretic application of glycine (125 hA); middle panel, from left to right records show antidromic spike during application of bicuculline (125 nA); the effect of GABA has been abolished while glycine is still effective; bottom panel, following termination of bicuculline application, the response to GABA has recovered. B: top panel, from left to right records show antidromic extracellular single unit spike as control response, during iontophoretic application of GABA (75 nA) and during iontophoretic application of glycine (75 hA); middle panel, from left to right records show antidromic spike during application of strychnine (75 hA), the effect of glycine has been abolished while GABA is still effective; bottom panel, following termination of strychnine application, the response to glycine has recovered. C illustrates that GABA and glycine prevent antidromic invasion of the soma-dendritic region but not the initial segment; upper panel, from left to right records show antidromic extracellular single unit spike as control response, during iontophoretic application of GABA (37 nA) and during iontophoretic application of glycine (37 nA); lower panel, from left to right records show antidromic spike during application of bicuculline (37 hA), the effect of GABA is antagonized while glycine is still effective. Note, however, the increased A-B delay during GABA application.
368 (x = 36 + 21 nA, n = 21) and by glycine (:~ = 40.6 + 30.6 nA, n = 20). Typical records are shown in Fig. 2. Few neurons demonstrated the inflection on the rising phase of the antidromic spike which indicates an A - B break 2. In cases where a notch was observed recording rarely lasted for more than a few minutes, usually being terminated with the sudden brief appearance of a high-frequency injury discharge. These events suggest that the electrode tip was in close proximity to or in contact with the membrane of the cell. Accordingly, these neurons were particularly difficult to test for their sensitivity to iontophoretic application of chemical substances owing to large artifactual changes in spike shape and amplitude upon application of current. Nevertheless, 3 such neurons demonstrated clear responses to iontophoretic application of G A B A and glycine. Records from one are shown in Fig. 2C. G A B A and glycine typically abolished the second component of the spike, leaving the first relatively unaffected. Effects of bicuculline and strychnine on amino acidinduced inhibition. The inhibitory effects of G A B A and glycine on the rate of spontaneous discharge were selectively antagonized by bicuculline (n = 5 neurons) and by strychnine (n = 4 neurons), respectively, as shown in Fig. 1. Some differences were apparent between the effects of bicuculline on the response to G A B A and those of strychnine on the response to glycine. Currents required for bicuculline to produce antagonism were 1-7 times (2 = 3.4 + 2.9) those of G A B A whereas strychnine produced antagonism with currents of only 0.5-2 times (2 = 1.4 _+ 0.6) those of glycine. The effects of strychnine were also typically more prolonged than those of bicuculline. Recovery of the response to glycine was observed within 120-630 s (2 = 410 + 262) while recovery of the response to G A B A within 18-70 s (~ = 45 + 25). With 4 of 5 neurons, iontophoretic application of bicuculline alone resulted in a transient increase in the rate of discharge, whereas strychnine did so with only 1 of 4 neurons. Bicuculline and strychnine also antagonized the inhibitory effects of G A B A (n = 8 neurons) and glycine (n = 6 neurons) on invasion of the soma-dendritic region by an antidromically propagated spike (see Fig. 2). As with spontaneously active neurons, greater ejecting currents were required for bicucul-
line to antagonize the effects of G A B A (ejecting currents for bicuculline were 1-5 times, 2 = 2.0 + 1.3 those of G A B A ) than were required for strychnine to antagonize the effects of glycine (ejecting currents for strychnine were 0.5-1.5 times, 2 = 1.0 + 0.3 those of glycine). Effects of strychnine were again more prolonged than those of bicuculline: recovery of the response to glycine occurred within 120-330 sec (2 = 232 + 99 s) while recovery of the response to G A B A occurred within 14-390 s (2 = 116 + 182 s). It should be noted that, with regard to recovery of the response to G A B A , while one neuron demonstrated a long recovery time of 390 s the remaining neurons demonstrated recovery times under 35 s. Neither bicuculline nor strychnine induced a silent neuron to discharge spontaneously, even when high ejecting currents (i.e greater than 100 nA) were used. With respect to the neuron showing an A - B break, depicted in Fig. 2C, iontophoretic application of bicuculline antagonized the effects of G A B A but not those of glycine. However, it appears that during the application of bicuculline, a partial response to G A B A was still evident as indicated by the increase in A - B delay. This study has demonstrated that G A B A and glycine have transient inhibitory effects on both silent and spontaneously active SPNs in the intermediolateral nucleus. These effects are selectively blocked by bicuculline and strychnine, respectively, suggesting that the responses were elicited via activation of pharmacologically specific receptors. A postsynaptic site of action of G A B A and of glycine, specifically on the soma or proximal dendrites, is suggested by the observation that they blocked the invasion of the soma-dendritic region by an antidromically propagated spike, and reduced the glutamate-induced excitation of the one neuron tested. This is consistent with an earlier report that inputs to somata of sympathetic preganglionic neurons are predominantly inhibitoryS. A postsynaptic action does not preclude the possibility that G A B A and glycine may also have acted indirectly. G A B A and glycine were roughly equipotent when similar amounts were applied, as has been reported for sacral parasympathetic neuronsS. Glycine has a significantly greater inhibitory effect than GABAll on somatic motoneurons while G A B A has a greater inhibitory effect compared to glycine on dorsal horn
369 neuronsl, and therefore SPNs may be best visualized as interneurons on the m o t o r side of spinal reflex pathways. The observation that bicuculline is not as p o t e n t an antagonist of G A B A as strychnine is ot glycine is consistent with earlier studies on somatic motoneurons8 and on neurons in the cuneate nucleus 9, and implies a relatively lower affinity of bicuculline for G A B A receptors than strychnine for glycine receptors. The increased rate of discharge of SPNs with bicuculline suggests that it may have some direct excitatory effect on sympathetic preganglionic neurons (see also refs. 6 and 8 on cerebral cortical neurons, ref. 4 on spinal neurons and ref. 10 on cuneate neurons). In summary, these data indicate that G A B A and glycine mediate powerful, pharmacologically specific
1 Backman, S. B. and Henry, J. L., Responses of thoracic dorsal horn units to visceral inputs and to amino acids in the cat, Soc. Neurosci. Abstr., 6 (1980) 490. 2 Brock, L. G., Coombs, J. S. and Eccles, J. C., Intracellular recording from antidromically activated motoneurones, J. Physiol. (Lond.), 122 (1953) 429-461. 3 Chung, K., LaVelle, F. W. and Wurster, R. D., Ultrastructure of HRP-identified sympathetic preganglionic neurons in cats, J. comp. Neurol., 190 (1980) 147-155. 4 Curtis, D. R., Duggan, A. W., Felix, D. and Johnston, G. A. R., GABA, bicuculline and central inhibition, Nature (Lond.), 226 (1970) 1222-1224. 5 De Groat, W. C., The effects of glycine, GABA and strychnine on sacral parasympathetic preganglionic neurones, Brain Research, 18 (1970) 542-544. 6 Godfraind, J. M., Krnjevic, K. and Pumain, R., Doubtful value of bicuculline as a specific antagonist of GABA, Nature (Lond.), 228 (1970) 675-676. 7 Henry, J. L., Effects of substance P on functionally identified units in cat spinal cord, Brain Research, 114 (1976) 439-451.
and direct inhibition of sympathetic preganglionic neurons and support the possibility that G A B A and glycine may be chemical m e d i a t o r s of inhibitory synaptic inputs which have a n o r m a l physiological role in the control of excitability of sympathetic preganglionic neurons. S u p p o r t e d by the Q u 6 b e c H e a r t F o u n d a t i o n . S.B.B. is a Trainee of the C a n a d i a n H e a r t F o u n d a tion. Expert photographic assistance was provided by Mr. N. Schestakowich. The authors are i n d e b t e d to Ms. A. Cairns for typing the manuscript. The following gifts are gratefully acknowledged: halothane (Somnothane) from Hoechst Pharmaceuticals, Montr6al; p a n c u r o n i u m b r o m i d e (Pavulon) from Organon, West Hill, Ontario.
8 Hill, R. G., Simmonds, M. A. and Straughan, D. W., A comparative study of some convulsant substances as 7-aminobutyric acid antagonists in the feline cerebral cortex, Brit. J. Pharmacol., 49 (1973) 37-51. 9 Hill, R. G., Simmonds, M. A. and Straughan, D. W., Antagonism of 7-aminobutyric acid and glycine by convulsants in the cuneate nucleus of cat, Brit. J. Pharmacol., 56 (1976) 9-19. 10 Kelly, J. S. and Renaud, L. P., On the pharmacology of 7-aminobutyric acid receptors on the cuneo-thalamic relay cells of the cat, Brit. J. Pharmacol., 48 (1973) 369-386. 11 Krnjevic, K., Puil, E. and Werman, R., GABA and glycine actions on spinal motoneurons, Canad. J. Physiol. Pharmacol., 55 (1977) 658-669. 12 Lebedev, V. P., Petrov, V. I. and Skobelev, V. A., Do sympathetic preganglionic neurones have a recurrent inhibitory mechanism? Pflagers Arch., 383 (1980) 91-92. 13 Wyszogrodski, I., Central Inhibition in the Sympathetic Nervous System, Ph. D. Thesis, McGill University, Montr6al, 1972.