Gamma-aminobutyric acid (GABA) concentration in a single neuron--localization of GABA in Deiters' neuron

Unorthodox view of the functioning of a GABAA synapse

1. Studies about the permeation of labelled chloride and GABA across single plasma membranes microdissected from vestibular Deiters' neurons have yielded two unexpected results: (a) intracellular GABA stimulates chloride permeation in an asymmetric fashion (efflux being favoured); (b) under certain conditions GABA permeates by a diffusion mechanism in the out-->in direction across these plasma membranes. 2. These two main results have been obtained over many years together with a host of other indications about the fine mechanism of these events. Thus, a picture has emerged of their physiological meaning within the context of the functioning of the GABAA synapses between the Purkinje cells and the Deiters' neurons. 3. In short, it is proposed that at these synapses GABA accumulates into the postsynaptic neuron after its release and activation of the postsynaptic receptors. GABA accumulated in the Deiters' neurons is involved in the process of chloride extrusion to build an inward directed electrochemical gradient for chloride.

GABA and chloride permeate via the same channels across single plasma membranes microdissected from rabbit Deiters' vestibular neurones

The permeation of labelled gamma-aminobutyric acid (GABA) across single microdissected Deiters' membranes has been studied in a microchamber system. The GABA permeation is via pores which are blocked by 4,4'-diisothiocyanato stilbene-2-2'disulphonic acid (DIDS). As this substance blocks as well chloride permeation across these membranes we tested whether GABA and chloride permeate across the same pores. Membrane pre-treatment with different doses of corticotropin releasing factor (CRF), a membrane permeant cyclic AMP analogue and phalloidin parallelly block the permeation of the two substances. Thus, it is most probable that GABA and chloride pass across the same pores. These pores may be swelling activated ones, opened by the mechanical stress on the membranes in the microchamber system. The passage of GABA across these pores may be of physiological importance in the termination of GABA inhibitory action on the vestibular Deiters' neurones.

Intracellular GABA-activated in-->out permeation of chloride across the Deiters' neuron membrane: modulation by phosphorylating activities

The modulation of intracellular GABA activated 36Cl- in-->out permeation across single Deiters' neuron membranes has been studied in a microchamber system. Addition of Mg2+/ATP on the membrane cytoplasmic side reduces strongly the GABA effect as does ATP alone. However, the greatest inhibition of the GABA effect is given by the addition of Mg2+ to the intracellular side buffer: a complete block of the stimulation by GABA of 36Cl- in-->out permeation. This is interpreted as due to the presence in this case of a constant concentration of exogenous Mg2+ acting together with endogenous ATP in the small cytoplasmic layer on the membrane inner side. The addition of ADP to Mg2+/ATP increases the inhibitory effect of the latter. This is presumably due to an extra increase of ATP, locally under the membrane, due to phosphorylation of ADP by endogenous phosphocreatine. Overall, the data confirm that phosphorylating conditions impair the intracellular GABA action on 36Cl- in-->out permeation.

Chloride permeation across the Deiters' neuron plasma membrane: activation by GABA on the membrane cytoplasmic side

Single plasma membranes were microdissected from Deiters' neurons freshly obtained from the lateral vestibular nucleus of the rabbit and their chloride permeability was studied in a microchamber system. The basal in-->out 36Cl- permeation initially found was brought to zero by Zn2+, 4,4'-diisothiocyanatostilbene-2,2'-disulphonic acid and iodide. GABA on the membrane cytoplasmic side resulted in a measurable in-->out 36Cl- passage, which was blocked by the GABA(A) antagonists bicuculline and picrotoxin. This effect peaked at 1 microM GABA on the inner side of the membrane. At higher GABA concentrations, a strong desensitization of the effect was found. Stimulation of Cl- permeability by GABA on the extracellular side of the membrane peaked at much higher GABA concentrations, 10-100 microM. This excludes an effect due to passage of the neurotransmitter from the inner to the outer compartment in our microchamber device. Moreover, this possibility is also dismissed by the fact that 1 microM GABA on the membrane outside did not evoke any 36Cl- in-->out permeation. In addition, pentobarbitone by itself could also stimulate 36Cl- in-->out permeation when added on the cytoplasmic side of Deiters' membrane. On these bases and in agreement with our previous reports, we propose that structures behaving pharmacologically as GABA(A) receptors respond to low levels of GABA on the cytoplasmic side of these neurons' membranes. We suggest that these structures are devices that, at the expense of ATP consumed in their phosphorylation, extrude Cl- after postsynaptic GABA uptake into the Deiters' neuron.

Stimulation of chloride in-->out permeation across the Deiters' neuron membrane by pentobarbital on the cytoplasmic side: additional evidence of GABA(A) receptors acting as chloride extrusion pumps

Pentobarbital stimulates 36Cl- permeation across single Deiters' membranes in a microchamber system, acting on classical, extracellularly facing, GABA(A) receptors. However, when applied on the membrane cytoplasmic side it activates per se labeled chloride in-->out permeation. No effect was found on chloride out-->in permeation. Similarly, at lower concentrations it facilitates the increase of 36Cl- in-->out permeation by application of GABA on the membrane inside, again via asymmetric chloride channels allowing in-->out but not out-->in passage. These data confirm that on the Deiters' membrane cytoplasmic side there are structures behaving pharmacologically as GABA(A) receptors whose function is that of a Cl- extrusion pump. This mechanism involves a cycle of activation-phosphorylation/desensitization-reactivation of the receptor complexes.

The effects of glucose, mannose, fructose and lactate on the preservation of neural activity in the hippocampal slices from the guinea pig

Using hippocampal slices from guinea pigs, we investigated the effect of different concentrations of glucose and replacement of glucose with mannose, fructose and lactate on neural activity. As an index of neural activity, the population spikes (PS) were recorded in the granule cell layer of the dentate gyrus (DG) and the pyramidal cell layer of the CA3 area in the hippocampal slices. Lowering the concentration of glucose from 10 mM to 5, 3, 2, 1 and 0 mM caused a reduction in the PS amplitude. There were differences in the decay times of the PS evoked in these two regions. PS evoked in CA3 region decayed faster even at a concentration of 3 mM glucose at which PS in granule cell layer was well maintained. The decay time of the PS in the CA3 region in the presence of glucose up to a concentration of 3 mM was shorter than that evoked in the DG. After the replacement of glucose with mannose, fructose or lactate, the PS disappeared within 35 min and there were no significant differences between the decay times in the two regions of slices incubated in the same medium. ATP, creatine phosphate (CrP) and lactate levels in each slice were determined. To investigate whether mannose and fructose could be metabolized or not in the tissue slice, anaerobic production of lactate from glucose, mannose and fructose were measured during oxygen and glucose deprivation. Under anaerobic conditions for 60 min, the levels of high-energy phosphates decreased to 50% of the initial level and lactate was produced from glucose, mannose or fructose. However, there were significant differences in the rate of lactate production between the DG and CA3 areas during application of 3 mM glucose, 10 mM mannose and 10 mM fructose. These results indicate that mannose, fructose and lactate can be metabolized and are available for maintaining the levels of high-energy phosphates but not for neural activity in the tissue slices and that the presence of glucose is indispensable for the maintenance of neural activity.

Ipsilateral corticotectal pathway inhibits the formation of long-term potentiation (LTP) in the rat superior colliculus through GABAergic mechanism

The purpose of the present experiments was to clarify the possible mechanism for the depression of long term potentiation (LTP) induction in the superficial gray layer of the rat superior colliculus after optic nerve stimulation. A postsynaptic field potential was recorded in vitro in the superficial gray layer of superior colliculus slices after stimulation of the optic layer. Tetanic optic layer stimulation (50 Hz, 20 s) induced LTP of the postsynaptic field potential elicited in the superficial gray layer. The postsynaptic field potential, with unitary discharges, produced in the superficial gray layer by optic nerve stimulation in vivo was depressed by a conditioning stimulus to the visual cortex. Identical inhibition of the cortical response of the superficial gray layer was produced by optic nerve stimulation. The application of picrotoxin (2.5 mg/kg, i.p.), a GABAA antagonist or methoxypyridoxine (100 mg/kg, i.v.), an anti-glutamate decarboxylase agent which reduces GABA levels, blocked the inhibitory interaction between the optic nerve-superficial gray layer and visual cortex-superficial gray layer. Tetanic optic nerve stimulation (50 Hz, 20 s) failed to induce LTP in the superficial gray layer of the intact rat. LTP was only elicited by tetanic optic nerve stimulation when picrotoxin or methoxypyridoxine was administered prior to the tetanic stimulation and when the ipsilateral visual cortex was removed. These results indicate that GABAergic interneurons in the superficial gray layer activated by corticotectal input, may stop the formation of LTP in the superficial gray layer.

The increase in Cl- permeation across the Deiters' neuron membrane by GABA on its cytoplasmic side is abolished by protein kinase C (PKC) activators

1. Cl- ion outward permeation across microdissected Deiters' neuron plasma membranes is augmented by GABA on the membrane cytoplasmic side. When these neurons are preincubated with a PKC activator, phorbol-12,13-dibutyrate (PdBu), there is a complex pattern of effects on basal and GABA-activated 36Cl- in-->out permeation. A distinct fact is an increase in basal Cl- passage and a disappearance of the 10(-6) M GABA effect at [PdBu] = 0.1 microM. 2. Likewise, 0.1 microM oleylacetylglycerol (OAG) treatment erases the effect completely, further supporting a role for PKC in modulating GABA-stimulated Cl- in-->out permeation. 3. The inactive ester, phorbol-12,13-didecanoate (Pdd), at 0.1 microM, does not affect GABA stimulation of Cl- passage. 4. High concentration (15-20 microM) of OAG and PdBu block the "intracellular" GABA efefct. However, the 20 microM PdBu effect is reversed by 30 microM H7. 5. These results indicate a role of endogenous PKC in Cl- extrusion by GABAA receptors on the cytoplasmic side of the Deiters' neuron membrane.

Selective reduction of glutamate in the rat superior colliculus and dorsal lateral geniculate nucleus after contralateral enucleation

The effects of afferent lesions on the levels of glutamate, aspartate and gamma-aminobutyric acid (GABA) in the laminae of the superior colliculus (SC) and dorsal lateral geniculate nucleus (dLGN) of the rat were studied, using microassay methods for these amino acids. The analysis was performed 12-14 days after left eye enucleation, or ablation of right visual cortical area, or both left eye enucleation and ablation of right visual cortex. Superficial gray layer (SGL) and deep layers in the SC were dissected out from the thin-sectioned, freeze-dried sample. In the dLGN, the outer and inner laminae were separately dissected. The glutamate contents in the upper half of SGL and outer lamina of dLGN contralateral to eye enucleation decreased significantly (15%). Combination of eye enucleation and visual cortical ablation further decreased the glutamate content in the upper half of the right SGL (29.3%). On the other hand, aspartate and GABA concentrations in the SC and dLGN exhibited no significant reduction after deafferentations. These results indicate that the retino-tectal and retino-geniculate pathway of the rat may be glutamatergic in nature.

Can Cl- ions be extruded from a gamma-aminobutyric (GABA)-acceptive nerve cell via GABAA receptors on the plasma membrane cytoplasmic side?

1. In this commentary we discuss results obtained by a micromethod for the study of Cl- permeability across single nerve membranes from rabbit Deiters' neurons. 2. These results showed the presence of GABAA receptors on the nerve cell membrane cytoplasmic side. 3. We could show that these receptor complexes have a higher affinity for GABA than their extracellularly facing counterparts. Moreover, they present a phenomenon of desensitization. Another distinct property is that upon activation by GABA, they expose positive charges at their cytoplasmic mouths. 4. We propose that these receptor complexes could function in situ as a device for extruding Cl- anions from the nerve cell interior. This phenomenon would create an electrochemical gradient for Cl- penetration into the cell upon the action of extracellular GABA, after its presynaptic release.

GABA distribution in the central vestibular system after retroauricular galvanic stimulation. An immunohistochemical study

The changes of the neurotransmitter (GABA) distribution in the brain stem of rats by retroauricular galvanic stimulation were investigated using the immunohistochemical method. In the lateral vestibular nucleus GABA-like immunoreactivity was more intensive on the side ipsilateral to the anodal stimulation than on the other side. It is concluded that retroauricular galvanic stimulation causes some changes in the inhibitory activity of the lateral vestibulo-spinal tract and of the spinal motor neuron.

Differential effects of fasciculus retroflexus lesions on serotonin, glutamate and gamma-aminobutyrate content and choline acetyltransferase activity in the interpeduncular nucleus

After placing bilateral electrolytic lesions in the fasciculus retroflexus (FR) of the rat, the endogenous content of serotonin, glutamate and gamma-aminobutyrate (GABA) as well as choline acetyltransferase activity (ChAT) were measured in the interpeduncular nucleus (IPN) at the 7th, 28th and 120th survival days. Confirming earlier results, an almost total depletion of ChAT was obtained in the IPN following complete FR lesions at any survival day studied. In such cases, the following changes were observed; 1) the serotonin level increased consistently and roughly doubled at the 120th survival day, suggesting heterotypic sprouting of serotonergic fibers and/or enhanced serotonin synthesis in the serotonergic neurons in the IPN, 2) the glutamate level decreased by approximately one-half, while the activity of high affinity uptake of glutamate remained unaltered, at the 7th survival day, suggesting a lowered glutamate formation coupled with lowered glucose utilization in the IPN, and 3) the GABA level decreased at a slower rate and reached one-third of the control at the 120th survival day, for which either transsynaptic degeneration of GABA neurons in the IPN or a suppressed metabolic rate in the GABA shunt following the lowered glutamate formation is a possible explanation.

Masking effect of NMDA receptor antagonists on the formation of long-term potentiation (LTP) in superior colliculus slices from the guinea pig

After electrical stimulation of the optic layer (OL) of superior colliculus (SC) slices, the postsynaptic potential (PSP) was recorded in the superficial gray layer (SGL) of the SC. The degeneration studies of retinotectal or corticotectal inputs to the SGL of the SC indicated that this PSP evoked in the SGL of the SC slices was retinotectal in origin. Neurotransmission in this pathway may be mediated by glutamate, because the PSP amplitude was reduced and blocked by application of kynurenate or quinoxaline dione (DNQX) to the medium. Furthermore, the concentration of glutamate in the right SGL was significantly reduced by 32% after left optic denervation and by 30% after ablation of the right visual cortex, compared with that in the left SGL. Long-term potentiation (LTP) in the SGL was induced by tetanic stimulation (50 Hz, 20 s) to the OL. The LTP formation was facilitated by the removal of Mg2+ from the medium. The effects of glutamate antagonists D-amino-5-phosphonovalerate (D-APV), gamma-D-glutamylglycine (gamma-DGG), and (+)-5-methyl-10,11-dihydro-5H-dibenzo, a,d-cycloheptene-5,10-imine maleate (MK-801) on the induction of LTP were investigated. D-APV (100 microM) or gamma-DGG (1 mM) masked the expression of LTP by tetanic stimulation, however LTP was induced after removal of the agents. LTP formation was observed without further tetanic stimulation following the removal of D-APV from the medium even 80 min after the tetanic stimulation. LTP once formed was not influenced by application of D-APV.(ABSTRACT TRUNCATED AT 250 WORDS)

GABA A receptors on the cytoplasmic side of the Deiters' neurone plasma membrane: mechanism and functional implications

A micromethod has been used for studying the passage of 36Cl- ions between two microchambers across the GABA acceptive plasma membrane of the rabbit Deiters' neurone, under various different conditions. The presence of 3.3 mM GABA, steady state intracellular concentration in situ (Okada & Shimada, 1976), on the cytoplasmic side of those membranes prolongs remarkably the time for the achievement of 36Cl- equilibrium across the membranes (from 4 minutes to 30 minutes). The effect appears to be mediated by "intracellularly oriented" GABAA receptor complexes as studied with GABAA antagonists such as bicuculline and picrotoxin. Intracellular GABA is still active in this effect at 10(-6) M. This phenomenon is discussed as, when coupled with the intracellular ATP producing machinery, being of importance for synaptically released GABA hyperpolarizing action on these neurones. The action of intracellular GABA on "internally oriented" receptors appears to involve the exposure of positively charged loci at the cytoplasmic side of the membrane. The consequent accumulation of intracellular Cl- at the membrane inner side would account for the higher overall permeability of those ions in the in----out direction. This circumstance was assessed by blocking the intracellular GABA effect either working at a relatively basic pH (8.4) inside or increasing the ionic strength in the intracellular compartment.

Direct evidence for the presence of GABAA receptors on the cytoplasmic side of the Deiters' neurone membrane

A newly developed micromethod has been used for studying the rate of passage of 36Cl- ions across single nerve membranes from rabbit Deiters' neurones. The application of gamma-aminobutyric acid (GABA) on the cytoplasmic side of those membranes increases the rate of passage of 36Cl- ions from that side to the other one across the membrane. The maximal effect is exerted by 10(-6) M GABA and it fades at higher neurotransmitter concentrations (10(-5) M to 3.3 x 10(-3) M). The cause of this fading of the effect appears to be a receptor desensitization phenomenon. The 10(-6) M GABA effect is reversed by both 10(-4) M picrotoxin and 10(-5) M bicuculline. The overall pattern of the data indicates the presence of GABAA receptors on the cytoplasmic side of these nerve membranes.

Laminar distribution of GABA (gamma-aminobutyric acid) in the dorsal lateral geniculate nucleus, Area 17 and Area 18 of the visual cortex, and the superior colliculus of the cat

The laminar distribution of gamma-aminobutyric acid (GABA) was studied in certain structures of the visual system of the adult cat. A microassay method to measure GABA (10(-12) mol) was established using enzymatic cycling of NADP-NADPH. In the dorsal lateral geniculate nucleus, GABA concentration was highest in lamina A (average concentration 23 mmol/kg dry weight) and lowest in lamina C. In the visual cortex (Areas 17 and 18), the concentration of GABA was 10-12 mmol/kg dry weight in layers I-IV and 5-8 mmol/kg dry weight in layers V and VI. No significant difference was found in the GABA distribution in Areas 17 and 18. In the superior colliculus, the highest level of GABA was found in the upper part of the superficial gray layer (40 mmol/kg dry weight), whereas the deep layers contained GABA at a concentration of 23-28 mmol/kg dry weight. The results of the GABA distribution measurements revealed an orderly, layer-specific disposition of the neurotransmitter in the cat visual system. GABA may play an important role in the function of the visual system.

Fine distribution of gamma-aminobutyric acid, glutamic acid decarboxylase, and glutamic acid in the rabbit cerebellum

The fine distribution of GABA, glutamic acid decarboxylase, and glutamic acid within each layer of the rabbit cerebellar cortex was determined with microanalytical methods. The greatest glutamic acid decarboxylase activity and the highest GABA concentration were found in the Purkinje cell layer. In the distribution of GABA and glutamic acid decarboxylase the peak of glutamic acid decarboxylase activity was more pronounced than that of GABA; the concentration of glutamic acid did not show much variation between each layer.

"Intracellular" GABA affects the equilibrium distribution of Cl- across the plasma membrane of a GABA acceptive neuron

The permeability of Cl- ions through single microdissected plasma membrane from Deiters' neurons was studied by a microtechnique. In particular, the time course of the passage of 36Cl- ions from a microchamber, M1, to another one, M2, across the membrane was followed. This study was performed with or without gamma-amino-butyric acid (GABA) in the two microchambers. The results suggest that in basal conditions the high intracellular concentration normally present in these neurons, 3.3 mM (1), causes a higher permeability of Cl- in the direction inside----outside in the respect of the plasma membrane. "Extracellular" GABA, 0.1 mM, is able to abolish this imbalance in Cl- permeability in the two opposite directions. This event appears to be the basis for GABA induced hyperpolarization of these neurons.

GABAA receptor complexes are present on both sides of a GABA-acceptive neuronal membrane

A micromethod allowing the study of the characteristics of single GABA-acceptive membranes microdissected from Deiters' neurones was used in order to assess the effects of both "extra"- and "intra"-cellular GABA on Cl- permeability. The results indicate that GABA can activate Cl- permeability in the in----out direction when it is present on the cytoplasmic side of the membrane. Moreover, as already described, it can activate Cl- permeability in the opposite direction when present on the "extracellular" side of the membrane. Both these phenomena are blocked by GABAA receptor inhibitors, bicuculline and picrotoxin. The presence of GABAA receptors on both sides of the membrane is discussed as the possible basis for synaptically released GABA hyperpolarising action on these neurones.

Asymmetric diffusion into the postsynaptic neuron: an extremely efficient mechanism for removing excess GABA from synaptic clefts on the Deiters' neurone plasma membrane

Microdissected Deiters' neuron plasma membranes have been used for studying the passage of GABA through the membrane both in the inward and outward direction. Working with 0.2 mM GABA in the compartment simulating the outside of the neurone and with 2.0 mM GABA in the one simulating the inside we found a net transport of GABA towards the inside. This mechanism does not require a Na+ ion gradient across the membrane. The nature of the transport process involved was studied by determining the rate of [3H]-GABA inward passage as a function of GABA concentration (1 nM - 800 microM) on the outward side of the membrane. The results have shown that until 50 microM a diffusion process (v = D1 X C, where D1 = 3.1 X 10(-11) 1/micron 2 X sec) is the sole mechanism involved. Above 50 microM a second diffusion process is activated v = D2 X (C - 50 X 10(-6), where D2 = 2.8 X 10(-11) 1/micron 2 X sec. Taking in account both inward and outward directed diffusion, one can calculate 16 microM as the equilibrium concentration of GABA on the outward side of the membrane. From a kinetic point of view, these diffusion processes are able to reduce GABA concentration in a synaptic cleft from 3 mM to 20 microM within 3 mu sec. These diffusion systems are discussed as extremely efficient in removing the excess of released GABA in the synaptic cleft.

GABA imbalance in squirrel monkey after unilateral vestibular end-organ ablation

Using an antibody against GABA conjugated to bovine serum albumin, GABA-like immunoreactivity was measured in vestibular nuclei and adjacent structures in normal and unilaterally vestibular-deafferentiated squirrel monkeys. Three and 6 days after end-organ ablation, GABA levels increased in lateral vestibular nucleus (LVN) on the side ipsilateral to the lesion, while GABA decreased in LVN on the side contralateral to the lesion. GABA levels in ventral cochlear nucleus or inferior cerebellar peduncle did not differ from normal in either case.

gamma-Aminobutyric acid (GABA) removal from the synaptic cleft: a postsynaptic event?

In the present commentary we discuss the adequacy of Na+ transport-coupled presynaptic gamma-aminobutyric acid (GABA) uptake systems for the removal of GABA from the synaptic cleft. This discussion is based on the accepted stoichiometry for GABA presynaptic internalization, GABAout + 3Na+out + K+in in equilibrium GABAin + 3Na+in + K+out, on the parameters reported in the literature for typical synaptosomal preparations, and on the assumption that GABA removal must be a quick event (less than or equal to 2 msec), as derived from electrophysiological studies. On these bases, we have developed a calculation in order to evaluate the time course of synaptic cleft GABA removal by presynaptic systems and ended up with an overall value (t approximately 0.3 sec) which does not fit with the data derived from electrophysiological recordings. Moreover, we calculated that if such systems had the function of removing GABA within 2 msec, as it should be, a large depolarization would be brought about in GABAergic boutons, resulting ultimately in further GABA release. These considerations together with biochemical and pharmacological experimental results seem to exclude that presynaptic uptake systems have the function of removing GABA from the synaptic cleft. Our experimental data on the ability of a GABA-acceptive postsynaptic membrane (Deiters' neuron membrane) to transport GABA indicate that this system may have the correct characteristics for removing the neurotransmitter. This refers to both the kinetics and the electrophysiological consequences of the phenomenon.

Glutamate decarboxylase activities in single vertebrate neurons

An enzymatic microassay method for glutamate decarboxylase (GAD) and gamma-aminobutyric acid (GABA) was improved to a degree yielding high sensitivity and low blank. Single cell bodies of anterior horn cells and dorsal root ganglion cells were dissected out from the freeze-dried sections of rabbit and chicken spinal cords and Purkinje cell bodies from those of rabbit cerebellum. A minute amount of GABA, present in single neurons or synthesized by GAD in single neurons, was enzymatically converted to NADPH. The NADPH was amplified 10,000-350,000-fold and measured, using an enzymatic amplification reaction (NADP cycling). GAD was contained in all Purkinje cell bodies and its average activity was four- to fivefold higher than those of the molecular and granular layers of rabbit cerebellum. The GABA concentration was threefold higher in Purkinje cell bodies than in these layers. GAD activity, at a level similar to that in the cerebellar layers, was found in almost all the cell bodies of anterior horn cells from rabbit and chicken. GABA was detected in 40% of rabbit neurons and not in chicken neurons. Dorsal root ganglion cells from both species contained no measurable GAD or GABA.

Problems and ambiguities in the identification of autonomic neurotransmitters

Although the traditional classification of autonomic neurons as either cholinergic or adrenergic is still subscribed to in most textbooks of physiology, it is apparent that many of the postganglionic neurons do not fall readily into either of these categories. Upwards of a dozen other transmitter candidates have been proposed, but in no instance has the identity of a non-cholinergic, non-adrenergic autonomic transmitter been established. In this paper some of the problems and ambiguities relating to satisfaction of the classical criteria for neurotransmitter identification are briefly surveyed, with special reference to the neuropeptides.

Pre- and post-synaptic inhibition

We now know a great deal about central inhibitory mechanisms: how they are organized in various neuronal circuits ("feed-forward" and "feed-back" inhibitions, inhibition of inhibitory cells giving "disinhibition" which releases neuronal activity in a finely graded and particularly safe manner); how they exercise a preponderant control over much of central neurl activity (9, 13, 14, 57); and of course, how inhibition operates at the cell membrane, by increasing Cl- permeability; this has a stabilizing action post-synaptically, but presynaptically, in afferent fibers, it depresses transmitter release from afferent terminals. In addition, there is some evidence that GABA transport may be electrogenic and therefore may significantly modulate membrane excitability, and that GABA may selectively depress Ca2+ influx in afferent terminals and thus inhibit transmitter release particularly effectively. However, it is by no means certain that we are fully aware of all the possible ways in which GABA affects neuronal excitability and synaptic transmission, and there may well be further exciting surprises in store.