The role of GABA and serotonin in the mediation of raphe-evoked spinal cord dorsal root potentials

The possible involvement of bulbospinal serotonergic systems in the mediation of analgesia has created a need for a better understanding of the influence this system has on neuronal mechanisms in the spinal cord. Therefore, these studies were designed to examine the effects of caudal raphe stimulation on primary afferent depolarization and to determine the role of serotonin (5-HT) and GABA in the mediation of these stimulation-produced effects. Stimulation of the raphe evoked two electrotonically conducted dorsal root potentials (DRP-1 and DRP-2) and two compound action potentials (VRP-1 and VRP-2) which were recorded from the dorsal and ventral roots, respectively. Length constant measurements indicated that DRP-1 was generated in group II and DRP-2 in group I primary afferent fibers. Histological determination of stimulation sites revealed that short-latency potentials (DRP-1 and VRP-1) were evoked from many sites within the caudal brain stem, while the long-latency potentials (DRP-2 and VRP-2) were evoked primarily from sites within the caudal raphe nuclei. The role of serotonin in mediating these evoked potentials was assessed by administering various antagonists of serotonin (cinanserin, methysergide and D-lysergic acid diethylamide). These agents consistently attenuated the long-latency potentials (DRP-2 and VRP-2) but increased the magnitude of DRP-1. The possibility of a GABAergic neuron in the descending systems projecting to primary afferent terminals was studied. Depletion of GABA by semicarbazide blocked DRP-1, but had only a modest effect of DRP-2. However, the putative GABA antagonist, bicuculline, inhibited both DRP-1, and DRP-2. These results suggest that a GABA interneuron is not involved in the bulbospinal serotonergic depolarization of primary afferent terminals. This system appears to constitute a presynaptic filter of afferent input, with the capacity to inhibit different fiber groups.     

Presynaptic action of beta(4-chlorophenyl)-gaba.

Spinal presynaptic inhibition is thought to be mediated by γ-aminobutyric acid (GABA). Because of the relative impermeability of GABA through the blood-brain barrier, the GABA-derivative, β(4-chlorophenyl)-GABA (Lioresal), was used to investigate possible presynaptic function. In eight cats anesthetized with sodium pentobarbital the administration of Lioresal (1 mg/kg, iv) attenuated the dorsal root potential and dorsal root reflex. Both dorsal root potential and reflex are thought to reflect presynaptic inhibition. The decrease of the dorsal root potential was accompanied by a d-c shift of the potential between the dorsal root and the spinal cord. Using condition and test stimulation of the dorsal root, Lioresal attenuated the unconditioned ventral root reflex, but first increased the conditioned reflex before it finally diminished the conditioned ventral root reflex as well. These results suggest that Lioresal maximally depolarizes the afferent terminals, thereby blocking the dorsal root potential and dorsal root reflex as well as presynaptically inhibiting motoneuron reflexes. These data support the notion that the therapeutic use of Lioresal for the treatment of spasticity can be attributed to the presynaptic inhibitor mechanisms.     

Presynaptic modulation of synaptic effectiveness of afferent and ventrolateral tract fibers in the frog spinal cord

In the isolated frog spinal cord, antidromic stimulation of motor nerves produces intraspinal field potentials with a characteristic spatial distribution. When recording from the ventral horn, there is a short latency (1–2 msec) response corresponding to activity generated by antidromic activation of motoneuron cell bodies and proximal dendrites. In addition, in the dorsal horn, a delayed wave (12–13 msec latency) corresponding in time with the negative dorsal root potential is also recorded. This wave (VR-SFP) is positive at the dorsal surface of the cord and inverts to negativity at more ventral regions. The negative VR-SFP is maximum between 300–500 μm depth from the dorsal surface and decays with increasing depth towards the motor nucleus. Six days after chronic section of the dorsal roots L7 to L9 in one side of the spinal cord, stimulation of the motor nerves on the deafferented side produces only the early response attributable to antidromic activation of motoneurons. No distinctive VR-SFPs are recorded at any depth within the cord. These findings are consistent with the interpretation that afferent fiber terminals are the current generators of the VR-SFP. The presynaptic and postsynaptic focal potentials recorded in the motor nucleus after stimulation of the ventrolateral tract, as well as the corresponding synaptic potentials electrotonically recorded from the ventral roots, are not depressed during conditioning stimulations which produce primary afferent depolarization. This contrasts with the depression of the presynaptic and post-synaptic focal potentials and synaptic potentials produced by stimulation of sensory fibers. It is concluded that, unlike the afferent fiber terminals, the terminals of the ventrolateral tract are not subjected to a presynaptic modulation of the type involving primary afferent depolarization.     

The effect of intravenously administered gamma-aminobutyric acid on afferent fiber polarization.

(1) The effect of intravenously administered gamma-aminobutyric acid (GABA) on afferent fiber polarization in the feline spinal cord was ascertained from fluctuations induced in the DC level of dorsal root filaments. (2) A dose-related depolarization of the filament, with a concomitant reduction in the magnitude of the dorsal root potential, was observed after 50 and 100 mg/kg GABA. (3) GABA also depolarized filaments of preparations in which interneuronal activity was suppressed by pretreatment with tetrodotoxin. Since the magnitude of the depolarization induced in these preparations was equal to that observed in nonpretreated animals, it is likely that the depolarization in the latter preparations reflects a direct effect on afferent terminals or fibers rather than an action on interneurons. (4) GABA failed to depolarize filaments in animals pretreated with bicuculline. This suggests that intravenously administered GABA interacted with receptors that are identical with or similar to those involved in neurally evoked primary afferent depolarization (PAD). (5) The direct depolarization of afferent fibers by intravenous GABA and the blockade thereof by bicuculline are characteristics compatible with those of the endogenous axo-axonic transmitter operating in pathways mediating neurally evoked PAD. These data, therefore, support the involvement of GABA at this synapse in the mammalian spinal cord.     

The development of the dorsal root potential and the responsiveness of primary afferent fibers to gamma-aminobutyric acid in the spinal cord of rat fetuses

The development of the dorsal root potential (DRP) and the responsiveness of primary afferent fibers to gamma-aminobutyric acid (GABA) were investigated in the isolated spinal cord of rat fetuses. At embryonic day 15.5, stimulation of the lumbar dorsal root was first effective in eliciting the DRP, which was not inhibited by bicuculline. A bicuculline-sensitive component of the DRP appeared at embryonic day 17.5. GABA (10 microM to 1 mM) caused a dose-dependent depolarization of the primary afferent fibers from embryonic day 13.5. The amplitude of the depolarization gradually increased with age until embryonic day 17.5 and was maintained thereafter. If the bicuculline-sensitive DRP solely reflects GABAergic activity, it is suggested that GABAergic activity develops at embryonic day 17.5 and the development of the responsiveness of primary afferent fibers to GABA precedes the functional onset of GABAergic neurons.     

Some actions of catechol on synaptic transmission in the isolated spinal cord of the frog.

The isolated frog spinal cord was used to investigate the synaptic effects of the convulsant agent catechol. Addition of the compound to the superfusate consistently enhanced orthodromic reflex activity recorded from ventral roots and augmented primary afferent depolarization. Concomitantly catechol altered the polarization changes produced in ventral and dorsal roots by putative neurotransmitter amino acids when these compounds were applied in Mg2+-containing Ringer. Catechol reduced the hyperpolarizations induced in motoneurons by the neutral amino acids, GABA, beta-alanine, taurine and glycine, but did not affect the depolarizations produced by the dicarboxylic amino acids, L-glutamate and L-aspartate. In contrast, catechol increased the dorsal root depolarizations elicited by both neutral and dicarboxylic amino acids and also the depolarizations produced by elevated potassium concentrations. Catechol did not bring about significant changes in the passive electrical properties of motoneurons or dorsal root fibers. In addition, it did not alter either the high affinity uptake or the depolarization-evoked release of tritiated GABA, glycine, L-glutamate and L-aspartate. It appears that the postsynaptic actions of catechol explain its ability to enhance spinal reflexes.     

Effect of ammonium acetate on depolarization processes in the central terminals of primary afferents]

Experiments on cats determined that ammonium acetate injected intravenously (2-4 mM/kg) supressed the processes of primary afferent depolarization (PAD) which are thought to be responsible for the presynaptic inhibition of spinal reflexes. The supression was transient and proceeded in paralle to depression of postsynaptic inhibition of monosynaptic reflexes. Ammonium acetate slightly decreased the amplitude of the negative postsynaptic potentials recorded form the dorsal surface of lumbar cord in response to stimulation of hind limb afferent nerves and increased polysynaptic reflex discharges in appropriate ventral roots. These findings make it unlikely that the ammonium depression of PAD is a result of impairment of interneuronal activity. A suggestion is made that ammonium depression of PAD results from diminition of the EMF for synaptic currents producing PAD.     

gamma-Hydroxybutyric acid is not a GABA-mimetic agent in the spinal cord.

gamma-Hydroxybutyric acid (GHB), a pharmacologically active central nervous system constituent, has been postulated to function as a gamma-aminobutyric acid (GABA) agonist. This hypothesis was tested directly on GABAergic synapses in isolated, superfused frog spinal cord. Addition of GHB to the superfusate produced effects on primary afferent terminals that were distinctly different from the effects of GABA. Thus, although both compounds depressed dorsal root potentials, GHB hyperpolarized terminals while GABA depolarized the same structures. The GABA responses were antagonized by bicuculline and picrotoxin, but these alkaloids did not change GHB's actions. In addition, GHB altered neither high-affinity uptake by cord slices, nor potassium-evoked release of tritiated GABA from them. GHB did not directly release GABA from spinal slices preloaded with [3H]GABA. These observations suggest that the central nervous system actions of GHB are not dependent upon its ability to activate GABAergic synapses or to modify GABAergic mechanisms.     

Cholinergic modulation of synaptic transmission in the spinal cord of the frog

Superfusion of the isolated frog spinal cord by the Ringer solution containing arecoline (10 mumol/l) evoked depolarization and increase of the input resistance and PSP amplitude of motoneurons. Depolarizing electrotonic potentials and reflex discharges in the ventral roots also increased, but duration of dorsal root potentials decreased. The observed arecoline facilitation of synaptic transmission in the spinal cord has postsynaptic nature evoked by motoneuron M2-cholinoreceptor activation and bound to an increase of the cyclic nucleotide metabolism. Arecoline inhibited the synaptic transmission in the spinal cord under conditions when its postsynaptic action was eliminated. This effect was due to presynaptic M1-cholinoreceptor activation without changing the cyclic nucleotide metabolism.     

Excitation of mouse motoneurones by GABA-mediated primary afferent depolarization

Observations on the reflex properties of a mouse spinal cord preparation in vitro are reported. The findings show that the synaptically evoked, GABA-mediated, discharge of action potentials in primary afferent fibres, monitored as the dorsal root reflex, may lead to the excitation of motoneurones. Subthreshold, bicuculline-sensitive increases in motoneuronal excitability, followed by prolonged inhibition, may be seen in preparations in which the delayed reflex is not seen. Thus, primary afferent depolarization may both increase motoneuronal excitability and also cause presynaptic inhibition of afferent input.     

GABAergic terminals are presynaptic to primary afferent terminals in the substantia gelatinosa of the rat spinal cord

Multiple, dorsal rhizotomies were performed unilaterally at lumbar levels L1–L4 in adult rats. Following 24–48 h degeneration periods and fixation by intracardiac perfusions, spinal cord were removed and transversely cut into 150 μm thick sections. These sections were incubated in immunocytochemical reagents for the peroxidase-labeling of glutamic acid decarboxylase (GAD), the enzyme that synthesizes the neurotransmitter γ-aminobutyric acid (GABA). The sections were then prepared for electron microscopic examination, while other sections were processed for light microscopic, GAD immunocytochemistry and for Fink-Heimer staining of degenerating axons and axon terminals.Thirty-six hours following dorsal rhizotomies, the sections that were prepared for the light microscopic study of terminal degeneration showed large numbers of degenerating profiles in the ipsilateral substantia gelatinosa while degenerating profiles were virtually absent contralaterally. In electron microscopic preparations, degenerating primary afferent terminals were commonly observed at the centers of rosettes where they formed synaptic contacts with other axon terminals and with surrounding dendrites. Several types of synaptic relationships were observed in the rosettes which involved both GAD-positive and degenerating primary afferent terminals. Such synaptic relationships included those in which: (a) a single GAD-positive terminal was presynaptic to the central, primary afferent terminal, (b) two different GAD-positive terminals formed synapses with opposite sides of the same central, primary afferent terminal and were also closely apposed to the surrounding dendrites of the rosette, and (c) a GAD-positive terminal was presynaptic to a primary afferent terminal and both types of terminals were presynaptic to the same dendrite of the rosette.The synaptic relationships described in this study are discussed with respect to their possible functional roles in such GABA-mediated phenomena as: (a) primary afferent depolarization, (b) the dorsal root reflex and (c) primary afferent hyperpolarization. Our observations support the concept that GABAergic axon terminals are involved in the synaptic circuits which produce presynaptic inhibition and presynaptic facilitation of the primary afferent input to the dorsal spinal cord. Collectively the observed synaptic relationships could provide a morphological substrate that is compatible with an inhibitory surround system in the substantia gelatinosa.     

Depression of afferent-induced primary afferent depolarization at the lumbar spinal cord following concussive head injury

Afferent-induced primary afferent depolarization (PAD) was depressed for 2-5 min following concussive head injury in the cat, as assessed by dorsal root potentials and augmentation of antidromic dorsal root potentials, both evoked by stimulation of adjacent dorsal roots. These changes in PAD were abolished by spinal cord transection but not affected by midpontine transection. Spontaneous dorsal root potentials, resting amplitudes of antidromic dorsal root potentials and reductions of antidromic dorsal root potentials following tetanic root stimulation were not substantially altered by injury. These findings suggest that concussive head injury depresses spinal interneuronal transmission by neurally mediated processes involving the bulbar brainstem.     

Ultrastructural study of glutamate- and GABA-immunoreactive terminals contacting the primary afferent fibers in frog spinal cord. A double postembedding immunocytochemical study

The in vitro HRP application to the dorsal root of the frog spinal cord produced an intensive staining of primary afferent fibers. A double postembedding GABA and glutamate immunocytochemical study revealed GABA- or glutamate-immunopositive presynaptic boutons establishing axo-axonic synapses onto HRP-stained primary afferent fibers in the spinal cord intermedial zone.     

GABAergic control of rhythmic activity in the presence of strychnine in the lamprey spinal cord

Rhythmic ventral root activity has been induced in the spinal cord of the lamprey in vitro in the presence of strychnine. Bicuculline (a GABAA blocker) increases the initial frequency and decreases the episode duration of such activity, whereas diazepam (which potentiates GABA action) has the opposite effect. In addition, bicuculline slows the depolarisation of ventral horn neurones. However, voltage clamp at potentials positive to the interburst potential does not reveal net outward current, even in the presence of diazepam, indicating that the effects of GABA on the rhythmic activity must be presynaptic to the impaled cells.     

The possibility of different effects of primary afferent depolarization on spinal reflexes]

Intensive depolarization of central primary afferent terminals evoked by strong stimulation of afferent nerves or dorsal root produces recurrent discharges which may be recorded as antidromic dorsal root reflexes. It is shown that the discharges are simultaneously propagating in the dorso-ventral direction and thus produce facilitation of spinal reflexes. The obtained results allow suggesting the existence of two types of influences of the primary afferent depolarization on the reflex transmission to the spinal cord.     

Presynaptic GABAA and GABAB Receptor-mediated Phasic Modulation in Axons of Spinal Motor Interneurons

The lamprey spinal cord has been utilized to investigate the role of presynaptic inhibition in the control of the spinal motor system. Axons of the lamprey spinal cord are comparatively large because of their lack of myelination. Axons impaled with microelectrodes demonstrate depolarizing responses to the application of GABAA and GABAB receptor agonists, muscimol and baclofen. These depolarizing effects are counteracted by the specific GABAA and GABAB receptor antagonists, bicuculline and phaclofen. GABAA receptor activation leads to a gating of Cl- channels on the axons. However, the ionic mechanism leading to axonal depolarization following GABAB receptor activation is unknown. After initiation of fictive locomotion, these axons demonstrate oscillations in axonal membrane potential related to the locomotor cycle. During ficitive locomotion they depolarize in phase with the bursting of the ipsilateral ventral root of the same segment. These axonal membrane potential oscillations are due to a phasic GABAA and GABAB receptor-mediated gating of ion channels on the axonal membrane. Fictive locomotion in the lamprey spinal cord is largely unaffected by antagonism of one or other GABA receptor subtype alone, but is severely disrupted by simultaneous antagonism of both subtypes. In conclusions, we demonstrate, for the first time, an agonist-gated depolarization of a vertebrate presynaptic element measured by direct impalement of the axon under study. We also demonstrate that GABA-mediated presynaptic inhibition occurs in axons of spinal interneurons. It is not limited to the primary afferents as has previously been believed.     

Electrophysiological evidence for a presynaptic mechanism of morphine withdrawal in the neonatal rat spinal cord

Dorsal and ventral root depolarizing responses to capsaicin (1 microM) and substance P (SP; 1 microM) were measured from the isolated, hemisected spinal cord of the neonatal rat. Capsaicin depolarized the dorsal and ventral roots. The mechanism of ventral root depolarization was presynaptic; since dorsal root depolarization preceded the ventral, and the ventral depolarization was eliminated when synaptic transmission was blocked in the absence of calcium. SP depolarized the ventral root without affecting the dorsal root. The SP-induced depolarization of the ventral root was reduced but not abolished by blocking synaptic transmission with low calcium, suggesting that SP acted postsynaptically on motoneurons and excitatory interneurons to depolarize the ventral root. Morphine (10 microM) abolished the capsaicin-induced ventral root depolarization, but only slightly suppressed the SP response (30%). The capsaicin-induced depolarization of the ventral root was enhanced greatly (238%) when morphine, which had been in the superfusion for 1 h, was removed and naloxone (1 microM) was added to the superfusion solution, whereas the SP response was not augmented during withdrawal from acute morphine. Furthermore, a putative SP antagonist ([D-Arg1, D-Pro2, D-Tryp7,9, Leu11]-SP) prevented the augmented capsaicin ventral root response during precipitated withdrawal. These data provide electrophysiological evidence for a presynaptic mechanism of acute morphine withdrawal in the neonatal rat spinal cord.     

GABA and glycine-like immunoreactivity at axoaxonic synapses on 1a muscle afferent terminals in the spinal cord of the rat

The object of this study was to analyze the synaptic interactions of identified muscle spindle afferent axon terminals in the spinal cord of the rat. Group 1a muscle afferents supplying the gastrocnemius muscle were impaled with microelectrodes in the dorsal white matter of the spinal cord and stained by intracellular injection with Neurobiotin. Postembedding immunogold techniques were used to reveal GABA- and glycine-like immunoreactivity in boutons presynaptic to afferent terminals in the ventral horn and the deep layers of the dorsal horn. Serial-section reconstruction was used to reveal the distribution of synaptic contacts of different types on the afferent terminals. The majority of afferent boutons received axoaxonic and made axodendritic or axosomatic synaptic contacts. In the ventral horn, 91% of boutons presynaptic to the afferent terminals were immunoreactive for GABA alone and 9% were immunoreactive for both GABA and glycine. The mean number of axo-axonic contacts received per terminal was 2.7, and the mean number of synaptic contacts at which the terminal was the presynaptic element was 1.4. In the deep layers of the dorsal horn, 58% of boutons presynaptic to afferent terminals were immunoreactive for GABA alone, 31% were immunoreactive for GABA and glycine, and 11% for glycine alone. The mean number of axoaxonic contacts received per afferent terminal in this region was 1.6 and the mean number of synaptic contacts at which the terminal was the presynaptic element was 0.86. This clearly establishes the principle that activity in 1a afferents is modulated by several neurochemically distinct populations of presynaptic neuron.     

Presynaptic inhibition of the monosynaptic reflex produced by injections of nicotine or eserine in the spinal cat

In completely deafferented cats with spinal cord transected and paralyzed with gallamine, the close arterial injection of nicotine (5 to 33 μg) into the spinal cord circulation or the intravenous injection of eserine (2 mg/kg) produced a transient increase in excitability of the central terminals of primary afferent fibers. Continuous d-c records from dorsal roots during nicotine injections indicated that a depolarization of the terminals was probably responsible for the observed excitability increase. Mecamylamine prevented both the depolarization and the increase in excitability of primary afferent central terminals whereas atropine and gallamine did not. The monosynaptic reflex was always depressed following injections of nicotine or eserine, but the excitability of the motorpool was found to be increased by these drugs. It was concluded that the monosynaptic reflex depression following injections of nicotine or eserine had a significant presynaptic inhibitory component as a consequence of a drug-induced depolarization of the central terminals of Group Ia afferent fibers.     

The influence of bemegride on frog spinal cord root potentials]

The effect of the convulsant bemegride (beta-ethyl-beta-methyl-glutharimide) on the electronic potentials in the frog spinal roots was investigated in situ. Intravenous injection of the bemegride subconvulsant doses (6.8 +/- 2.7 mg/kg) depressed rapidly the electrotonic dorsal root potentials (DRP) evoked by a stimulation of the adjacent dorsal root (DR) or the ventral root (VR). They were reduced by 55--67% 3-6 min after the bemegride injection. The effect of bemegride was reversible and the DRP amplitude restored its initial value within an hour. Ventral root potentials after bemegride injection revealed only greater amplitude fluctuations. A conclusion is made that bemegride is a selective and potent depressor of the primary afferent depolarization in the frog spinal cord.