J. Physiol. (1979), 293, pp. 393-415 With 12 text-figurem Printed in Great Britain
393
GLYCINE, GABA AND SYNAPTIC INHIBITION OF RETICULOSPINAL NEURONES OF LAMPREY BY GARY MATTHEWS AND WARREN 0. WICKELGREN WITH AN APPENDIX BY A. R. MARTIN From the Department of Physiology, University of Colorado Medical School, Denver, Colorado 80262 US.A. (Received 11 January 1979) SUMMARY
1. Intracellular recordings were made from the cell bodies and axons of giant reticulospinal neurones (MUller cells) of the lamprey and the effects of a variety of putative neurotransmitters tested. Bath-applied acetylcholine, carbamyleholine, norepinephrine, dopamine, histamine and serotonin were without effect. Glycine and y-aminobutyric acid (GABA) hyperpolarized and reduced the input resistance of cell bodies but had no effect on the membrane conductance of axons. 2. The threshold dose of bath-applied GABA or glycine for a conductance change in somata was about 0.5 mm and the maximum effect was reached at about 10 mM. The maximum conductance change produced by glycine was always greater than that produced by GABA. 3. Replacement of the sodium in the bathing saline with lithium or choline prolonged the conductance change produced by ionophoretically applied glycine or GABA, suggesting the presence of sodium-dependent uptake systems for glycine and GABA. 4. The reversal potentials for responses to ionophoretically applied glycine and GABA average about -83 mV, the same as that for the inhibitory post-synaptic potential (i.p.s.p.) produced in Muller cells by stimulation of the ipsilateral vestibular nerve. 5. The i.p.s.p. and drug responses appeared to involve an increase in chloride conductance, since their reversal potentials were shifted appropriately by changes in either internal or external chloride. 6. Changes in extracellular potassium concentration also changed i.p.s.p. and drug reversal potentials. However, these effects could be attributed to secondary changes in internal chloride. 7. The receptors for GABA and glycine appeared to be different because of the absence of cross-desensitization and because, at doses below 20 /,M, picrotoxin and bicuculline selectively blocked GABA responses while strychnine selectively blocked glycine responses. 8. At concentrations of 20 /tM, strychnine eliminated the i.p.s.p. while picrotoxin and bicuculline had no effect. Further, the i.p.s.p. and glycine response of Muller cells located in the isthmic region of the midbrain had the same threshold sensitivity 0022-3751/79/4540-0954 $01.50 © 1979 The Physiological Society
394 3. MATTHEWS AND W. 0. WICKELGREN to strychnine. However, the glycine response of other Muller cells was more sensitive to strychnine than was the i.p.s.p. 9. We conclude that glycine is a better candidate for the inhibitory transmitter onto Muller cells than is GABA. INTRODUCTION
Among the most generally accepted candidates for inhibitory neurotransmitters in the vertebrate central nervous system are the amino acids glycine and y-aminobutyric acid (GABA) (see reviews by Curtis & Johnson, 1974 and Krnjevic, 1974). There is evidence implicating glycine in inhibition of spinal motoneurones (Werman, Davidoff & Aprison, 1968), of medullary reticular neurones (Tebecis & Dimaria, 1972), of retinal ganglion cells (Caldwell, Daw & Wyatt, 1978), and as the transmitter released by Renshaw cells (Curtis, Game, Lodge & McCulloch, 1976; Belcher, Davies & Ryall, 1976). GABA has been suggested to be the inhibitory transmitter released onto and by cerebellar Purkinje cells (Woodward, Hoffer, Siggins & Oliver, 1971; Obata, Takeda & Shinozaki, 1970), in cerebral cortex (Krnjevic & Phillis, 1963), in the cuneate nucleus (Kelly & Renard, 1973) and substantia nigra (Precht & Yoshida, 1971; Dray, Gonye & Oakley, 1976), the retina (Caldwell et al. 1978) and the visual cortex (Sillito, 1975). In the vertebrate C.N.S., however, it normally is not possible to compare in detail the effects of GABA and glycine with those of the natural transmitter and to assess quantitatively the effects of various pharmacological agents. This is so because of the difficulty of intracellular recording from most vertebrate C.N.S. neurones and the lack of control over their extracellular environment. The lamprey offers a number of advantages for the study of inhibitory synaptic transmission and the effects of putative inhibitory transmitters. Lampreys are vertebrates with a prototypic vertebrate C.N.S., and there are a number of large individually identifiable neurones in the brain, most notably the giant recticulospinal neurones (Muller cells) of the brainstem (Rovainen, 1967; Wickelgren, 1977). These cells and their spinal axons permit stable intracellular penetration for hours with one or more micropipettes. Further, the isolated lamprey brain and spinal cord survive well when placed in artificial saline, allowing experimental control of the ionic and pharmacological environment of the C.N.S. Finally, a pure, compound inhibitory post-synaptic potential (i.p.s.p.) can be evoked in Muller cells by electrical or mechanical stimulation of the ipsilateral vestibular capsule (Wickelgren, 1977). We have investigated the physiological and pharmacological properties of the actions of glycine and GABA on Muller cells and compared them with the i.p.s.p. evoked by vestibular stimulation. GABA, glycine and the natural inhibitory transmitter all had similar effects on the post-synaptic membrane in that the reversal potentials for all three were identical and responded similarly to manipulations of the ionic environment. However, experiments using the pharmacological agents strychnine, picrotoxin and bicuculline indicated that glycine is a better candidate for the inhibitory transmitter than is GABA. An abstract on this work has appeared (Matthews & Wickelgren, 1978a).
GLYCINE, GABA AND SYNAPTIC INHIBITION
395
METHODS
Experiments were conducted on isolated brains and spinal cord segments from young adult (Petromyzon marines). The isolated spinal cord was prepared from an anaesthetized lamprey (tricaine methanseulphonate, 0-5 g/l.) by cutting out a segment of trunk region between the gills and the first dorsal fin and removing everything except the notocord and overlying spinal cord. The preparation was then fixed to the bottom of a saline-filled Plexiglas chamber via minutien pins pushed through the notocord and into a layer of Sylgard resin (Dow Coming). For the brain experiments the lamprey was anaesthetized and decapitated in the mid-gill region. The head was then pinned dorsal side up in a dish of cold, oxygenated saline consisting of (mM): 104-5 NaCl, 2 KCl, 4 CaCl2, 8 MgCl2, 4 glucose, and 2 Tris; titrated to a pH 7-4 with HCl. (In some experiments the composition of the saline was altered as indicated in Results.) The brain, rostral spinal cord, and vestibular capsules were exposed and removed from the head together with the cartilaginous brain case and notocord. The preparation was then transferred to a saline-filled Plexiglas chamber which allowed change of solution via a push-pull syringe arrangement. Temperature was held at 7-5 ± 0 -5 C by electrothermal plates beneath the chamber. Transmission of light was improved by stripping away tissue adhering to the ventral surface of the brain case. The brain-spinal cord preparation was then pinned to the sea lampreys
bottom of the chamber dorsal side up. To expose the Muller cells, the choroid plexus covering the 3rd and 4th ventricles, the pineal and parapineal organs, and the habenula were removed and the commissure of the optic tectum and the cerebellum split at the mid line. The membranes of the vestibular capsules were exposed by removing the capsule cartilage dorsally and laterally. Electrical stimuli were applied to the membranes via bipolar platinum electrodes. Experiments on Muller cell bodies were done on the three pairs of cells in the midbrain (M1, M2 and M3), the single visible pair in the aqueductal region (I,) and the most anterior two pairs in the floor of the 4th ventricle (B1 and B2) (the nomenclature is from Rovainen, 1967). Muller axons were recognized by their unique electrophysiological characteristics (Rovainen, 1967; Wickelgren, 1977), but specific axons were not identified, except that the bursting Muller axon, 12, which appears to receive a chemical synaptic input (Rovainen, 1967), was not studied. Usually two micropipettes, one to pass current and the other to record membrane potential, were placed in a Muller cell body or axon. The electrodes were filled with 4 M-K acetate (pH 7-0) and had resistances in saline of 30-60 MO. For bath application, drugs were added to lamprey saline without compensation for osmolarity, but when necessary pH was readjusted to 7-4. Whenever a change in bathing saline was made during an experiment, at least 100 ml. of the new solution was washed through the chamber, which had a volume of 9 ml. Electrodes for ionophoresis of drugs were filled with 4 M-GABA (pH 2-7) or 1-5 M glycine (pH 2-0). Ionophoretic current was usually 0-5-0-7 ,#A and was provided by an isolated stimulator through a 100 MQl resistor. Leakage of drug from the pipette was retarded by applying a constant negative bias voltage. Conventional electrophysiological recording and stimulating techniques were employed. Fast eletrophysiological events were displayed on an oscilliscope and photographed. For slower events a strip chart recorder was used. RESULTS
Effects of bath-applied drugs on Muller cell bodies & axons The effects on Muiller cell bodies and axons of a number of putative neurotransmitters were screened using bath application. The procedure was to impale either a Muller cell body or axon with both a current-passing and a recording micropipette and measure the resting potential and current-voltage (I-V) characteristics of the membrane before, during, and after the addition of the drug. Acetylcholine, carbamylcholine, norepinephrine, histamine, dopamine, and serotonin were without effect on either cell bodies or axons when applied at a concentration of 1 mm. Higher concentrations of these drugs were not tested. Glutamate and aspartate
396 G. MATTHEWS AND W. 0. WICKELGREN depolarized Muller cell bodies and axons, and these effects are described in the next paper (Matthews & Wickelgren, 1979). Glycine, taurine, fl-alanine and y-aminobutyric acid (GABA) hyperpolarized Mfiller cell bodies and axons but appeared to affect the membrane conductance of only the cell bodies and not the axons. The results on Muller cell bodies will be described first. Soma
Axon
--60 E
-65~
~
(n A) -20 -15 -10 I
'
lC//
-75 '
-70
oa
~
l
Gly
~~~~~E -75 (nA) -20-15 -10 -80 ' ' ' '2 /5 +10+15 Contro -90 -95
-90
-~~~~~~100
-95 control -100
:-70-
~~~~~~~~E
W'
+5 +10 +15
393
GLYCINE, GABA AND SYNAPTIC INHIBITION OF RETICULOSPINAL NEURONES OF LAMPREY BY GARY MATTHEWS AND WARREN 0. WICKELGREN WITH AN APPENDIX BY A. R. MARTIN From the Department of Physiology, University of Colorado Medical School, Denver, Colorado 80262 US.A. (Received 11 January 1979) SUMMARY
1. Intracellular recordings were made from the cell bodies and axons of giant reticulospinal neurones (MUller cells) of the lamprey and the effects of a variety of putative neurotransmitters tested. Bath-applied acetylcholine, carbamyleholine, norepinephrine, dopamine, histamine and serotonin were without effect. Glycine and y-aminobutyric acid (GABA) hyperpolarized and reduced the input resistance of cell bodies but had no effect on the membrane conductance of axons. 2. The threshold dose of bath-applied GABA or glycine for a conductance change in somata was about 0.5 mm and the maximum effect was reached at about 10 mM. The maximum conductance change produced by glycine was always greater than that produced by GABA. 3. Replacement of the sodium in the bathing saline with lithium or choline prolonged the conductance change produced by ionophoretically applied glycine or GABA, suggesting the presence of sodium-dependent uptake systems for glycine and GABA. 4. The reversal potentials for responses to ionophoretically applied glycine and GABA average about -83 mV, the same as that for the inhibitory post-synaptic potential (i.p.s.p.) produced in Muller cells by stimulation of the ipsilateral vestibular nerve. 5. The i.p.s.p. and drug responses appeared to involve an increase in chloride conductance, since their reversal potentials were shifted appropriately by changes in either internal or external chloride. 6. Changes in extracellular potassium concentration also changed i.p.s.p. and drug reversal potentials. However, these effects could be attributed to secondary changes in internal chloride. 7. The receptors for GABA and glycine appeared to be different because of the absence of cross-desensitization and because, at doses below 20 /,M, picrotoxin and bicuculline selectively blocked GABA responses while strychnine selectively blocked glycine responses. 8. At concentrations of 20 /tM, strychnine eliminated the i.p.s.p. while picrotoxin and bicuculline had no effect. Further, the i.p.s.p. and glycine response of Muller cells located in the isthmic region of the midbrain had the same threshold sensitivity 0022-3751/79/4540-0954 $01.50 © 1979 The Physiological Society
394 3. MATTHEWS AND W. 0. WICKELGREN to strychnine. However, the glycine response of other Muller cells was more sensitive to strychnine than was the i.p.s.p. 9. We conclude that glycine is a better candidate for the inhibitory transmitter onto Muller cells than is GABA. INTRODUCTION
Among the most generally accepted candidates for inhibitory neurotransmitters in the vertebrate central nervous system are the amino acids glycine and y-aminobutyric acid (GABA) (see reviews by Curtis & Johnson, 1974 and Krnjevic, 1974). There is evidence implicating glycine in inhibition of spinal motoneurones (Werman, Davidoff & Aprison, 1968), of medullary reticular neurones (Tebecis & Dimaria, 1972), of retinal ganglion cells (Caldwell, Daw & Wyatt, 1978), and as the transmitter released by Renshaw cells (Curtis, Game, Lodge & McCulloch, 1976; Belcher, Davies & Ryall, 1976). GABA has been suggested to be the inhibitory transmitter released onto and by cerebellar Purkinje cells (Woodward, Hoffer, Siggins & Oliver, 1971; Obata, Takeda & Shinozaki, 1970), in cerebral cortex (Krnjevic & Phillis, 1963), in the cuneate nucleus (Kelly & Renard, 1973) and substantia nigra (Precht & Yoshida, 1971; Dray, Gonye & Oakley, 1976), the retina (Caldwell et al. 1978) and the visual cortex (Sillito, 1975). In the vertebrate C.N.S., however, it normally is not possible to compare in detail the effects of GABA and glycine with those of the natural transmitter and to assess quantitatively the effects of various pharmacological agents. This is so because of the difficulty of intracellular recording from most vertebrate C.N.S. neurones and the lack of control over their extracellular environment. The lamprey offers a number of advantages for the study of inhibitory synaptic transmission and the effects of putative inhibitory transmitters. Lampreys are vertebrates with a prototypic vertebrate C.N.S., and there are a number of large individually identifiable neurones in the brain, most notably the giant recticulospinal neurones (Muller cells) of the brainstem (Rovainen, 1967; Wickelgren, 1977). These cells and their spinal axons permit stable intracellular penetration for hours with one or more micropipettes. Further, the isolated lamprey brain and spinal cord survive well when placed in artificial saline, allowing experimental control of the ionic and pharmacological environment of the C.N.S. Finally, a pure, compound inhibitory post-synaptic potential (i.p.s.p.) can be evoked in Muller cells by electrical or mechanical stimulation of the ipsilateral vestibular capsule (Wickelgren, 1977). We have investigated the physiological and pharmacological properties of the actions of glycine and GABA on Muller cells and compared them with the i.p.s.p. evoked by vestibular stimulation. GABA, glycine and the natural inhibitory transmitter all had similar effects on the post-synaptic membrane in that the reversal potentials for all three were identical and responded similarly to manipulations of the ionic environment. However, experiments using the pharmacological agents strychnine, picrotoxin and bicuculline indicated that glycine is a better candidate for the inhibitory transmitter than is GABA. An abstract on this work has appeared (Matthews & Wickelgren, 1978a).
GLYCINE, GABA AND SYNAPTIC INHIBITION
395
METHODS
Experiments were conducted on isolated brains and spinal cord segments from young adult (Petromyzon marines). The isolated spinal cord was prepared from an anaesthetized lamprey (tricaine methanseulphonate, 0-5 g/l.) by cutting out a segment of trunk region between the gills and the first dorsal fin and removing everything except the notocord and overlying spinal cord. The preparation was then fixed to the bottom of a saline-filled Plexiglas chamber via minutien pins pushed through the notocord and into a layer of Sylgard resin (Dow Coming). For the brain experiments the lamprey was anaesthetized and decapitated in the mid-gill region. The head was then pinned dorsal side up in a dish of cold, oxygenated saline consisting of (mM): 104-5 NaCl, 2 KCl, 4 CaCl2, 8 MgCl2, 4 glucose, and 2 Tris; titrated to a pH 7-4 with HCl. (In some experiments the composition of the saline was altered as indicated in Results.) The brain, rostral spinal cord, and vestibular capsules were exposed and removed from the head together with the cartilaginous brain case and notocord. The preparation was then transferred to a saline-filled Plexiglas chamber which allowed change of solution via a push-pull syringe arrangement. Temperature was held at 7-5 ± 0 -5 C by electrothermal plates beneath the chamber. Transmission of light was improved by stripping away tissue adhering to the ventral surface of the brain case. The brain-spinal cord preparation was then pinned to the sea lampreys
bottom of the chamber dorsal side up. To expose the Muller cells, the choroid plexus covering the 3rd and 4th ventricles, the pineal and parapineal organs, and the habenula were removed and the commissure of the optic tectum and the cerebellum split at the mid line. The membranes of the vestibular capsules were exposed by removing the capsule cartilage dorsally and laterally. Electrical stimuli were applied to the membranes via bipolar platinum electrodes. Experiments on Muller cell bodies were done on the three pairs of cells in the midbrain (M1, M2 and M3), the single visible pair in the aqueductal region (I,) and the most anterior two pairs in the floor of the 4th ventricle (B1 and B2) (the nomenclature is from Rovainen, 1967). Muller axons were recognized by their unique electrophysiological characteristics (Rovainen, 1967; Wickelgren, 1977), but specific axons were not identified, except that the bursting Muller axon, 12, which appears to receive a chemical synaptic input (Rovainen, 1967), was not studied. Usually two micropipettes, one to pass current and the other to record membrane potential, were placed in a Muller cell body or axon. The electrodes were filled with 4 M-K acetate (pH 7-0) and had resistances in saline of 30-60 MO. For bath application, drugs were added to lamprey saline without compensation for osmolarity, but when necessary pH was readjusted to 7-4. Whenever a change in bathing saline was made during an experiment, at least 100 ml. of the new solution was washed through the chamber, which had a volume of 9 ml. Electrodes for ionophoresis of drugs were filled with 4 M-GABA (pH 2-7) or 1-5 M glycine (pH 2-0). Ionophoretic current was usually 0-5-0-7 ,#A and was provided by an isolated stimulator through a 100 MQl resistor. Leakage of drug from the pipette was retarded by applying a constant negative bias voltage. Conventional electrophysiological recording and stimulating techniques were employed. Fast eletrophysiological events were displayed on an oscilliscope and photographed. For slower events a strip chart recorder was used. RESULTS
Effects of bath-applied drugs on Muller cell bodies & axons The effects on Muiller cell bodies and axons of a number of putative neurotransmitters were screened using bath application. The procedure was to impale either a Muller cell body or axon with both a current-passing and a recording micropipette and measure the resting potential and current-voltage (I-V) characteristics of the membrane before, during, and after the addition of the drug. Acetylcholine, carbamylcholine, norepinephrine, histamine, dopamine, and serotonin were without effect on either cell bodies or axons when applied at a concentration of 1 mm. Higher concentrations of these drugs were not tested. Glutamate and aspartate
396 G. MATTHEWS AND W. 0. WICKELGREN depolarized Muller cell bodies and axons, and these effects are described in the next paper (Matthews & Wickelgren, 1979). Glycine, taurine, fl-alanine and y-aminobutyric acid (GABA) hyperpolarized Mfiller cell bodies and axons but appeared to affect the membrane conductance of only the cell bodies and not the axons. The results on Muller cell bodies will be described first. Soma
Axon
--60 E
-65~
~
(n A) -20 -15 -10 I
'
lC//
-75 '
-70
oa
~
l
Gly
~~~~~E -75 (nA) -20-15 -10 -80 ' ' ' '2 /5 +10+15 Contro -90 -95
-90
-~~~~~~100
-95 control -100
:-70-
~~~~~~~~E
W'
+5 +10 +15
