GABAergic systems in the vestibular nucleus and their contribution to vestibular compensation
Introduction
γ-Aminobutyric acid (GABA) mediates both fast and slow inhibitory synaptic transmission within the mammalian brain, and 20–50% of all synapses use GABA as a neurotransmitter (Sieghart, 1995). GABA is the endogenous ligand for GABAA, GABAB, and GABAC receptors (for a review see Chebib and Johnston, 1999, Bormann, 2000). These receptor subtypes are classified according to differences in their molecular structure and pharmacology (Table 1). Both GABAA and GABAC receptors are transmitter-gated ion channels and mediate fast synaptic inhibition. In contrast, GABAB receptors are G-protein linked receptors and mediate slow synaptic inhibition (Chebib and Johnston, 1999). Most of the evidence relating to GABA receptors in the process of recovery from vestibular damage (i.e., vestibular compensation) relates to GABAA receptors and therefore they will be discussed in detail. Fewer data are available on GABAB receptors; however, they will be discussed briefly due to the different roles that GABAA and GABAB receptors may play in vestibular compensation. There is no experimental evidence suggesting a role for the GABAC receptors in vestibular compensation. It is likely, however, that GABAC receptors are present within the vestibular nuclear complex (VNC) and they could conceivably contribute to recovery from vestibular damage.
The GABAA receptor belongs to the transmitter-gated ion channel superfamily that also includes the nicotinic acetylcholine and glycine receptors. In the mature brain, the GABAA receptor mediates fast inhibitory synaptic transmisson via the conduction of Cl− down their electrochemical gradient, resulting in hyperpolarization. In contrast, in the developing brain, the GABAA receptor mediates excitatory transmission and plays an important role in trophic effects, such as neuronal migration and synaptogenesis (for a review see McCarthy et al., 2002, Ben-Ari et al., 1997).
The GABAA receptor channel is a pentameric structure. The five subunits and their splice variants that form the GABAA receptor have been cloned and divided into different subfamilies based upon the deduced amino acid sequence homology (α1–6, β1–4 with two splice variants, γ1–3 with two splice variants, ɛ, δ, θ, and π) (for a review see Chebib and Johnston, 1999, Whiting et al., 1999). The sequence homology within a GABAA receptor subfamily is approximately 70–90%, and falls to 30–40% between the subfamilies (Moss and Smart, 1996). The structure of the individual subunits, as well as the suggested topological organization within the cell membrane, is illustrated in Fig. 1 (for a review see Smith and Olsen, 1995, Whiting et al., 1999, Cherubini and Conti, 2001). The occurrence within the brain and the functions of the better characterized GABAA receptor subunits are illustrated in Table 2 (for a review see Whiting et al., 1999, Cherubini and Conti, 2001). Table 2 illustrates that the most abundant GABAA receptor subunits are α1, β2, and γ2 and that GABAA receptor subunits are involved in a diverse range of neurological processes (for reviews see Mehta and Ticku, 1999, Rudolph et al., 2001).
There is a considerable debate over the native receptor stoichiometry of the GABAA receptor. Theoretically, a large number of diverse GABAA receptors could be derived from different combinations of the GABAA receptor subunits (i.e., 2000 different GABAA receptors; Chebib and Johnston, 1999). This is further complicated by the observation that GABAC receptor subunits can co-assemble with GABAA receptor subunits to form heteromeric GABAA receptors (Qian and Ripps, 1999). Nevertheless, based on immunoprecipitation studies, McKernan and Whiting (1996) have suggested that there are only eight major subunit combinations within the brain (McKernan and Whiting, 1996) (see Table 3). The preferred native receptor stoichiometry is dependent on the brain region being examined, but the most common receptor stoichiometry within the mammalian brain is two α, two β, and one of the γ, δ, or ɛ subunits (for a review see Whiting et al., 1999) (see Table 3). Sigel and Kannenberg (1996), however, have indicated that the limitation of immunoprecipitation studies is that only the solubilized GABAA receptors are detected, therefore, leaving the putative insoluble GABAA receptors undetected and hence potentially underestimating the diversity of GABAA receptors (Sigel and Kannenberg, 1996).
There are four commercially available GABAA receptor agonists: muscimol, isoguvacine, 4,5,6,7-tetrahydroisoxazolo[5,4-c]pyridin-3-ol (THIP) and piperidine-4-sulfonic acid (P4S). Muscimol, the most widely used GABAA receptor agonist, is a full GABAA receptor agonist, and at similar concentrations, a GABAC receptor agonist; therefore, muscimol cannot distinguish between responses mediated by GABAA and GABAC receptors (Woodward et al., 1993, Chang et al., 2000) (Table 4). The inability to distinguish between GABAA and GABAC receptors is not a problem associated with muscimol alone, as most of the available GABAA receptor agonists interact with GABAC receptors. Like muscimol, isoguvacine is a full GABAA receptor agonist and a partial GABAC receptor agonist (Woodward et al., 1993, Chang et al., 2000) (Table 4). Both THIP and P4S are partial GABAA receptor agonists and competitive GABAC receptor antagonists (Woodward et al., 1993, Ebert et al., 1997) (see Table 4). In summary, none of the available GABAA receptor agonists is selective for the GABAA receptor; all interact with the GABAC receptor.
There are only three GABAA receptor antagonists available: the competitive GABAA receptor antagonists, bicuculline and its related quaternary salts and gabazine (SR 95531 (2-(3′carboxy-2′-propyl)-3-amino-6-p-methoxyphenylpyridazinium bromide)); and the non-competitive GABAA receptor antagonist, picrotoxin. Both bicuculline and gabazine, however, have been suggested to act as allosteric inhibitors, reducing the influx of Cl− by decreasing the probability that the channel will open (Ueno et al., 1997). The first available competitive GABAA receptor antagonist was bicuculline; however, the limitation of bicuculline is that under physiological temperatures and pH, it is rapidly metabolised to the less active form, bicucine (Olsen et al., 1975). This instability led to the development of the soluble quaternary salts: bicuculline methiodide, bicuculline methochloride, and bicuculline methobromide. Despite their non-selectivity, the bicuculline salts are the most widely used GABAA receptor antagonists. As well as being GABAA receptor antagonists, bicuculline and its salts are also glycine, 5-hydroxytryptamine, nicotinic acetylcholine, and N-methyl-d-aspartate (NMDA) receptor antagonists (see Table 5). Furthermore, bicuculline salts can also block the afterhyperpolarization of thalamic reticular nucleus neurons, hippocampal interneurons, cortical pyramidal cells, and spinal neurons (Table 5). With only a 10-fold difference between the IC50 values for the blockade of the afterhyperpolarization and GABAergic transmission, it is impossible to differentiate between the two different mechanisms of action of the bicuculline salts (for review see Seutin and Johnson, 1999). In vitro assays suggest that bicuculline and its salts are also competitive inhibitors of brain acetylcholinesterase (Table 5), and can increase intracellular Ca2+ independently of their action on the GABAA receptor (Mestdagh and Wulfert, 1999). Seutin and Johnson (1999) suggest that bicuculline salts should not be used to characterize the role of GABAA receptors.
Gabazine (SR 95531) is the most potent competitive GABAA receptor antagonist of the pyridazinyl-GABA derivatives (for a review see Wermuth and Biziere, 1986). In clonal cell line studies (Ito et al., 1992, Woodward et al., 1993, Ebert et al., 1997) and in vitro assays (Heaulme et al., 1986, Maksay, 1998), gabazine is a more potent GABAA receptor antagonist than bicuculline. However, bicuculline is more potent than gabazine in inducing convulsions in mice when delivered systemically (Heaulme et al., 1986). The lower potency of gabazine in vivo has been attributed to it not crossing the blood–brain-barrier (Wermuth and Biziere, 1986). Unlike bicuculline, gabazine cannot block cholingeric receptors (Zhang and Feltz, 1991), afterhyperpolarizations (Seutin et al., 1997), or increase intracellular Ca2+ (Mestdagh and Wulfert, 1999). Furthermore, gabazine does not affect the strychnine, glutamatergic, dopaminergic D2, serotoninergic 5-HT1 and 5-HT2, α1-noradrenergic or muscarinic binding sites (Heaulme et al., 1986). Gabazine is, however, a relatively weak GABAC receptor antagonist (Woodward et al., 1993, Feigenspan and Bormann, 1994), and in the locus coeruleus, it has been suggested to inhibit monoamine oxidase (Luque et al., 1994).
Picrotoxin is a non-competitive GABAA receptor antagonist (i.e., Krishek et al., 1996), and, like bicuculline and gabazine, is also non-selective. Picrotoxin is a GABAC receptor antagonist with a 30-fold lower potency than for its action at the GABAA receptor (Woodward et al., 1993, Feigenspan and Bormann, 1994, Wang et al., 1995, Enz and Cutting, 1999). Clonal cell line (Lynch et al., 1995, Greka et al., 1998, Steinbach et al., 2000, Shan et al., 2001), neuronal culture (Yoon et al., 1998), and brain slice (Chattipakorn and McMahon, 2002) studies indicate that picrotoxin is also a glycine receptor antagonist (see Table 5). In summary, like the GABAA receptor agonists, the GABAA receptor antagonists are not selective for the GABAA receptor and do interact with other receptors.
There is an abundance of allosteric modulators for the GABAA receptor, including benzodiazepines, steroids (i.e., allopregnanolone), inhalation anaesthetics (i.e., isoflurane), ethanol, zinc, and furosemide (for a review see Costa, 1998, Chebib and Johnston, 1999). Since most studies of vestibular compensation employ anesthetics during inner ear surgery, any action of these anesthetics on the GABAA receptor could be an important confounding factor in the compensation process that follows. Inhalation anaesthetics enhance GABAergic transmission and this enhancement has been hypothesized to be one of the mechanisms underlying anaesthesia (see Little, 1996, Yamakura et al., 2001). It is likely that the use of inhalation anaesthetics, such as isoflurane, during an experiment would alter GABA-mediated transmission.
Stress-related steroids have also been implicated in the vestibular compensation process (for a review see Gliddon et al., 2003). Neuroactive steroids can positively or negatively modulate the GABAA receptor (for reviews see Zinder and Dar, 1999, Covey et al., 2001; Lambert et al., 2001a, Lambert et al., 2001b; Rupprecht et al., 2001). The modulation of GABAA receptors by neuroactive steroids is thought to play a role in the behaviour induced by the menstrual cycle (i.e., pre-menstrual syndrome) and stress (for a review see Zinder and Dar, 1999, Rupprecht et al., 2001).
The GABAB receptor belongs to the G-protein-linked receptor family and has the large extracellular N-terminal that contains the agonist/antagonist binding site, the seven transmembrane regions and the intracellular C-terminal structure of these receptors (for a review see Billinton et al., 2001). There are three GABAB receptor isoforms: GABA(B1a), GABA(B1b), and GABA(B2); it has been suggested that for a functional GABAB receptor to occur, either one of the GABAB(1a) or GABAB(1b) isoforms needs to form a heteromeric complex with the GABA(B2) (for a review see to Billinton et al., 2001, Bowery et al., 2002). Furthermore, it is believed that the GABAB(1a) is found pre-synaptically, whilst the GABAB(1b) is found post-synaptically (Billinton et al., 2001). Pre-synaptic GABAB receptors modulate the release of neurotransmitters.
GABAB receptors mediate slow synaptic inhibition via the modulation of adenylate cyclase and ion channels (i.e., K+ and Ca2+ channels (Bowery et al., 2002). Pre- and post-synaptic GABAB receptors decrease Ca2+ conductances and increase K+ conductances, respectively.
Baclofen, introduced in the 1970s, is the most commonly used GABAB receptor agonist and is the only GABAB receptor agonist that is used in the clinical setting. Other baclofen analogues have been synthesized, however, none have surpassed baclofen's affinity for the GABAB receptor. It was not until the carboxylic acid moiety of GABA was replaced by the phosphinic acid series that more potent GABAB receptor agonists were synthesized. Of this phosphinic acid series, 3-aminopropylphosphinic acid (CGP27492) and 3-aminopropylmethylphosphinic acid (SKF97541) are the most commonly used (for a review see Bowery et al., 2002).
The first available selective GABAB receptor antagonists, phaclofen, saclofen, and 2-hydroxysaclofen, were derivative of baclofen. The problem associated with these antagonists is that these drugs exhibit relatively low potency and cannot penetrate the blood–brain-barrier. The majority of the second generation GABAB receptor antagonists are derivatives of 3-aminopropylphosphinic acid and include CGP35348, CGP35348, CGP36742, and CGP55485 (for a review see Bowery et al., 2002). The advantage of these ligands is that they cross the blood–brain-barrier. Another GABAB receptor antagonist is the structurally distinct SCH 50911, which is an orally active, modestly potent GABAB receptor antagonist (for a review see Bowery et al., 2002).
GABAC receptors, like the GABAA receptor, belong to the transmitter-gated ion channel family and mediate fast synaptic transmission. However, the GABAC receptor is distinct from the GABAA receptor in terms of molecular biology, physiology, and pharmacology. There are three GABAC receptor isoforms, ρ1, ρ2, and ρ3, which share only 30–38% amino acid sequence identity with the GABAA receptor subunits (Bormann, 2000). The ρ1 isoform is located predominantly in the retina, whilst ρ2 is found throughout the CNS (Enz et al., 1995, Enz and Cutting, 1999). The GABAC receptor is insensitive to the GABAA receptor antagonist, bicuculline, and the GABAA receptor allosteric modulators (i.e., benzodiazepines, steroids, and barbiturates), but is selectively blocked by 1,2,5,6-tetrahydropyridin-4-methylphosphinic acid. GABAC receptors conduct less current and have a longer channel opening time than the GABAA receptor (Bormann, 2000).
Section snippets
GABA within the vestibular nucleus complex (VNC)
Neurons positive for GABA and its synthesizing enzyme, glutamic acid decarboxylase (GAD), are scattered throughout the rat (Nomura et al., 1984, de Waele et al., 1994, Zanni et al., 1995), guinea pig (Kumoi et al., 1987, Dupont et al., 1990), rabbit (Blessing et al., 1987), cat (Walberg et al., 1990, Tighilet and Lacour, 2001), and monkey (Holstein et al., 1996, Holstein et al., 1999a, Holstein et al., 1999b) VNC (for a review see Holstein, 2000). In the cat VNC, it has been estimated that less
Experimental evidence for the role of VNC GABAA receptors in vestibular compensation
In mammals, removal of the sensory inputs from one of the vestibular labyrinths by unilateral vestibular deafferentation (UVD) induces an ocular motor and postural syndrome. Over time, some of the symptoms of the UVD are reduced in severity or eliminated in a process known as ‘vestibular compensation’. The drive for vestibular compensation research is 2-fold. First, vestibular compensation has been proposed as a model of central nervous system lesion-induced plasticity. Second, damage to the
Does the modification of the GABAergic system correlate with the compensation of the static symptoms?
In the abovementioned studies, it is impossible to determine which neuronal population is exhibiting the changes in the GABAergic systems. The pre-synaptic changes (i.e., modification of GAD, GAT, and GABA concentrations) could occur in the terminals arising from the cerebellum, vestibular commissures, or local interneurons or from GABAergic neurons that project to the vestibular commissures, oculomotor and trochlear motoneurons, the inferior olive, or medial vestibulo-spinal tract (refer to
Are GABA receptors causally involved in vestibular compensation?
At the present stage, although there is some evidence to suggest that GABAergic systems may change during the development of vestibular compensation, there is no compelling and definitive evidence that such changes are causally implicated in the compensation process. The fact that various drugs that activate or block GABAA or GABAB receptors can modulate either the expression of UVD symptoms or alter the time course of their compensation, does not necessarily mean that the receptors on which
Acknowledgements
This research was supported by a Project Grant from the New Zealand Health Research Council. CG was supported by a FRST Bright Futures Ph.D. Scholarship.
References (258)
- et al.
The GABAA receptor γ1 subunit is expressed by distinct neuronal populations
Brain Res. Mol. Brain Res.
(1992) - et al.
Localization of GABAA-receptor γ2-subunit mRNA-containing neurons in the rat central nervous system
Neuroscience
(1992) - et al.
Region-specific expression of GABAA receptor α3 and α4 subunits mRNAs in the rat brain
Brain Res. Mol. Brain Res.
(1992) - et al.
Nystagmus induced by pharmacological inactivation of the brainstem ocular motor integrator in monkey
Vis. Res.
(1999) - et al.
Nonvesicular release of neurotransmitter
Neuron
(1993) - et al.
GABAA, NMDA and AMPA receptors: a developmentally regulated ‘menage a trois’
TiNS
(1997) - et al.
Advances in the molecular understanding of GABAB receptors
TiNS
(2001) - et al.
Vestibulospinal pathway in rabbit includes GABA-synthesizing neurons
Neurosci. Lett.
(1987) The ‘ABC’ of GABA receptors
TiPS
(2000)- et al.
Bicuculline-insensitive GABA receptors on peripheral autonomic nerve terminals
Eur. J. Pharmacol.
(1981)