Elsevier

Brain Research

Volume 770, Issues 1–2, 3 October 1997, Pages 202-212
Brain Research

Research report
Pre- and postsynaptic modulation of spinal GABAergic neurotransmission by the neurosteroid, 5β-pregnan-3α-ol-20-one

https://doi.org/10.1016/S0006-8993(97)00809-3Get rights and content

Abstract

The neuroactive steroid 5β-pregnan-3α-ol-20-one (5β3α) modulates GABAA receptor function by potentiating postsynaptic GABA currents. While much is now known about the postsynaptic action of neurosteroids, far less is known about how they affect neurotransmission. We have investigated the synaptic actions of 5β3α in a simple vertebrate model, the embryo of the clawed toad, Xenopus laevis, in which a known GABAergic pathway, activated by the rostral cement gland, terminates swimming when the animal contacts an obstruction. Cement gland stimulation evokes bicuculline-sensitive inhibitory postsynaptic potentials (IPSPs) in motorneurones that terminate swimming and which are greatly enhanced by the presence of (1–5 μM) 5β3α. In the presence of TTX, depolarising inhibitory potentials are recorded with KCl-filled microelectrodes reflecting the spontaneous release of transmitter. The majority are glycinergic with durations of 20–80 ms and are blocked by strychnine while the remainder are GABAergic with durations of 90–200 ms and are abolished by bicuculline. We show here that, in the presence of 5β3α, the spontaneous GABA IPSPs lengthen dramatically in some cases to over 500 ms, but the glycine potentials are unaffected. The steroid has no other detectable postsynaptic effects in that the range of amplitudes of GABA potentials is unaffected and there is no change in the resting membrane potential. However, 5β3α also caused a marked increase in the rate of occurrence of spontaneous GABA potentials. This suggests a novel presynaptic site of action in which the steroid enhances the probability of vesicular GABA release from GABA terminals.

Introduction

In contrast to the classical, long-term action of steroid hormones, which act through intracellular receptor-mediated changes in protein synthesis, some neuroactive steroids have been shown to act within seconds to alter the excitability of neurones in the CNS. A substantial amount of evidence has now accumulated demonstrating that both synthetic 1, 7, 8, 14and naturally occurring neurosteroids 5, 22, possess anaesthetic properties via positive allosteric modulation of the GABAA receptor. Single-channel kinetic studies on GABAA receptors of mouse spinal neurones in culture suggest that neurosteroids increase the mean open time of the chloride channel by increasing the probability of naturally occurring longer open states, with some increase in the channel-opening frequency, especially at higher concentrations [33]. So far, there is no evidence for any additional postsynaptic mechanism of action for the neurosteroid agonists. For example, they have been shown not to affect the reversal potential or the GABA uptake system 5, 18. However, the augmentation of GABA-activated chloride currents by steroids occurs at low concentrations (nM), whereas at high concentrations (>1 μM), like the barbiturates, steroids can directly activate the chloride channel 5, 8, 18. This postsynaptic action of the steroid appears to be highly specific and despite the similarity of the GABAA receptor to the glycine receptor, neuroactive steroid agonists have no obvious effect on glycine-mediated postsynaptic currents 14, 18, 35. Furthermore, under voltage-clamp conditions, alphaxalone (5α-pregnan-3α-ol-11,20-dione) and 5β-preganan-3α-ol-20-one (5β3α) have no effect on NMDA- or kainate-induced currents [18].

Despite the resurgence of interest in the anaesthetic and therapeutic actions of neuroactive steroids 13, 19, their effects on GABAergic neurotransmission have yet to be examined in detail. In this study, we have explored the effects of 5β3α on GABAergic IPSPs in the spinal cord of a relatively simple and amenable preparation, the Xenopus embryo. This system has been intensively studied in the context of locomotor rhythm generation (see [25]for review). Following a brief sensory stimulus, such as a dimming of the illumination or after being touched on the trunk or tail skin, the embryo can sustain rhythmic swimming activity for up to several minutes. This activity is achieved by a co-ordinated pattern of contractions in the segmented myotomal muscles, which alternates between the two sides and passes down the body from head to tail, at around 15 cm·s−1. A `fictive' correlate of this activity can be recorded from the ventral roots of embryos paralysed in the neuromuscular blocking agent, α-bungarotoxin (α-BTX) and spinalisation studies have shown that the spinal cord itself possesses sufficient neural circuitry to produce alternating rhythmic activity [27]. The pattern of alternating activity is achieved by on-cycle excitation (with evidence for both glutamatergic [9], cholinergic and electrical components [21]) and crossed (glycinergic) inhibition [29]. At this early stage in development, GABAergic transmission appears not to be involved in the generation of the swimming rhythm [29]. However there is one pathway which is known to involve a population of GABAergic neurones, located in the brainstem, that influences the activity of spinal neurones. Swimming activity can either terminate spontaneously or after the animal contacts an obstruction. This latter `stopping' response can be mimicked in the immobilised animal by pressing either the head skin or the cement gland [3]. The pathway involves the activation of movement detector neurones, whose free nerve endings innervate the head skin and cement gland [15], and which, in turn, activate GABAergic midhindbrain reticulospinal neurones (mhr) whose axons descend into the spinal cord and inhibit neurones of the swimming pattern generator 3, 4.

Results from this present study show that the steroid, 5β3α, enhances the GABAergic stopping response and, therefore, that GABAA receptors have a similar sensitivity to neurosteroids to that already described in other vertebrate preparations. By investigating the effects of the steroid on spontaneous GABA potentials which remain in the presence of TTX, we were able to investigate, in some detail, the actions of 5β3α on synaptic transmission.

Section snippets

Materials and methods

Experiments were carried out on stage 37/8 embryos of the African clawed toad, Xenopus laevis[20], which were obtained by induced breeding following injection of human chorionic gonadotrophin (1000 U/ml, Sigma) into pairs of adults from a laboratory colony.

Firstly, the tail fins of the animals were gashed before placing them in a chamber containing 2 ml of α-bungarotoxin (1.25 μM, Sigma). After around 30 min, when the animals were fully paralysed, one was transferred to a preparation bath (ca.

The stopping response is potentiated by the neurosteroid 5β-pregnan-3α-ol-20-one

An episode of swimming activity can either terminate apparently spontaneously or be terminated prematurely by activating the cement gland pathway. Fig. 2A shows an intracellular recording made with a 2 M KCl-filled microelectrode, with an accompanying ventral root, of a typical episode of swimming activity in which there is a gradual decline in the cycle period between each burst of activity before the episode finally ceases. In Fig. 2B, another episode of swimming activity is prematurely

Discussion

In this study, we have shown that spinal motorneurones receive two types of spontaneous depolarising IPSP when recorded with KCl-filled microelectrodes, in contrast to previous studies that identified a single population of chloride-dependent IPSP 26, 32. The two types of IPSP are distinguishable on the basis of their durations and pharmacological sensitivity: the fast IPSPs are blocked by strychnine and are presumably glycinergic (cf. [28]); the slow IPSPs are blocked by bicuculline and are

Acknowledgements

We would like to thank Joe McDearmid for valuable comments on the manuscript. This work was supported by a Royal Society, 1983 University Research Fellowship (K.T.S.) and a grant from the Welcome Trust (to K.T.S.). C.A.R. was supported, in part, by a BBSRC studentship.

References (36)

  • H. Callachan et al.

    Modulation of the GABAA receptor by progesterone metabolites

    Proc. Roy. Soc. Lond. B

    (1987)
  • J.D.W. Clarke et al.

    Sensory physiology, anatomy and immunohistochemistry of Rohon–Beard neurons in embryos of Xenopus laevis

    J. Physiol.

    (1984)
  • G.A. Cottrell et al.

    Alphaxalone potentiates GABA and stimulates the GABAA receptor of bovine chromaffin cells in culture

    J. Physiol.

    (1986)
  • G.A. Cottrell et al.

    Modulation of GABAA receptor activity by alphaxalone

    Br. J. Pharmacol.

    (1987)
  • N. Dale et al.

    Dual-component amino-acid-mediated synaptic potentials: excitatory drive for swimming in Xenopus embryos

    J. Physiol.

    (1985)
  • N. Dale et al.

    Inhibitory neurons of a motor pattern generator in Xenopus revealed by antibodies to glycine

    Nature

    (1986)
  • K.W. Gee et al.

    EPALONS, the antiepileptic activity of neurosteroids active at the GABAA receptor

    Epilepsia

    (1995)
  • N.L. Harrison et al.

    Modulation of the GABAA receptor complex by a steroid anaesthetic

    Brain Res.

    (1984)
  • Cited by (27)

    • Developmental differences in neuromodulation and synaptic properties in the lamprey spinal cord

      2007, Neuroscience
      Citation Excerpt :

      The lack of insight into GABAergic synaptic inputs meant that no attempt was made to separate GABAergic and glycinergic inhibitory inputs in this study. This is a potential weakness, as in Xenopus GABAergic inputs have longer time-courses than glycinergic inputs (Reith and Sillar, 1997). It is possible that the differences in IPSP properties in different experiments could reflect variability introduced by recording from glycinergic or GABAergic IPSPs.

    • Neurosteroid modulation of GABA<inf>A</inf> receptors

      2003, Progress in Neurobiology
      Citation Excerpt :

      In agreement with earlier findings, these studies demonstrate that the GABAA receptor active pregnane steroids act primarily to prolong mIPSC, or sIPSC, decay recorded from neurons in the cerebellum (Purkinje, granule and stellate cells), hippocampus (CA1 and dentate granule cells), hypothalamus (oxytocin-containing neurons of the supraoptic nucleus) and cortex (Brussaard et al., 1997; Cooper et al., 1999; Harney et al., 2003; Fancsik et al., 2000; Jorge-Rivera et al., 2000; Hamann et al., 2002; Koksma et al., 2003; Puia et al., 2003; Vicini et al., 2002). Additionally, in rat hypothalamic/pituitary cell co-cultures, acutely isolated neurons of the medial preoptic nucleus, and spinal neurons of Xenopus tadpoles, the pregnane steroids have been reported to increase the frequency of IPSCs, indicative of a presynaptic facilitatory action (Reith and Sillar, 1997; Poisbeau et al., 1997; Haage and Johansson, 1999). Many central neurons have the capacity to express a complex array of GABAA receptor subunits.

    • Allopregnanolone modulates spontaneous GABA release via presynaptic Cl<sup>-</sup> permeability in rat preoptic nerve terminals

      2002, Brain Research
      Citation Excerpt :

      However, a few hypotheses have been discussed. According to one of these, steroids, in this case the progesterone metabolite pregnanolone, increase transmitter release by a direct potentiation of presynaptic Ca2+ channels [46]. This idea is, however, not supported by other reports where the progesterone precursor pregnenolone reduced the currents through high-voltage activated Ca2+ channels [15,50].

    View all citing articles on Scopus
    View full text