ReviewGABA potentiation: a logical pharmacological approach for the treatment of acute ischaemic stroke
Introduction
Stroke is the third leading cause of death in major industrialised countries and is also the major cause of long lasting disability (Bonita, 1992). It therefore produces enormous social, financial and personal problems in the community, consuming for example 5% of the UK National Health Service budget (Lees, 1998). Since the adult mammalian CNS has not been considered to regenerate after damage, stroke therapy has focused on (1) reducing risk factors (e.g. antihypertensive agents and stopping patients smoking) and (2) minimising secondary brain damage by restoring and maintaining perfusion. However, a third approach is now being investigated, that of ‘neuroprotection’. This approach relies on the fact that while the cerebral tissue in the area immediately adjacent to the infarct (the core) is probably irretrievably damaged, the tissue surrounding this area (the penumbra), although compromised, is probably capable of full recovery if the right conditions are provided rapidly. The concept of the ‘neuroprotective agent’, a drug which if given soon after the stroke will protect the brain from the damage (Green and Cross, 1997b) has thus developed. This concept has been given considerable impetus over the last 10–15 years with the explosion of information on the biochemical mechanisms occurring in the brains of experimental animals following an acute hypoxic–ischaemic episode. It is felt that interfering with some of these pathological changes should result in damage being attenuated (Green and Cross, 1997b).
A major focus of stroke research has been the role of the excitatory neurotransmitter l-glutamate. The initial observations that exogenously applied glutamate (Olney et al., 1989) or N-methyl-d-aspartate (NMDA) (Coyle, 1983) were neurotoxic, were followed by the demonstration that extracellular levels of glutamate rise during ischaemia (Benveniste et al., 1984) and the report that 2-amino-7-phosphonoheptanoic acid, a competitive NMDA antagonist, was neuroprotective when given to rats prior to an ischaemic insult (Simon et al., 1984). Subsequently, Gill et al. (1988) showed that the high affinity non-competitive NMDA antagonist dizocilpine (MK 801) was an effective neuroprotective agent against cerebral damage when given after a global ischaemic insult. This was followed by reports of its efficacy in focal models of ischaemia (Buchan et al., 1992) [see Green and Cross (1997a) and Hunter et al. (1995) for descriptions of the different types of animal models of ischaemia and their relevance to stroke research].
Numerous compounds have been discovered over the last 15 years that inhibit glutamate function and thereby minimise the neuropathological consequences of an acute ischaemic episode. Many have been effective in at least some of the experimental models of stroke (Small and Buchan, 1997), but none has been shown unequivocally to be of clinical value. A probable reason for this is that some clinical trials were launched with compounds that produced dose limiting, clinically significant adverse events which meant that the drugs could not be given at doses high enough to provide neuroprotection. Problems with the design of some of the clinical trials may also have been a factor (Muir and Grossett, 1999). Consequently we have good reason to believe that inhibition of glutamate function is still a valid approach to preventing ischaemia-induced cell death in the central nervous system.
Because excessive glutamate function initiates a cascade of events (Fig. 1), a substantial number of experimental approaches have been proposed to modify these biochemical/pathological changes and a large number of compounds developed to act on them. Some of these are also illustrated in Fig. 1 and have been reviewed in detail elsewhere (Green and Cross, 1997a). However, none of these other approaches has been characterised or investigated to the depth of the glutamate excitotoxic hypothesis and inhibition of glutamate function remains the premier approach.
Meldrum (1990) suggested that the excitotoxic process probably depends on a balance between excitatory and inhibitory mechanisms. GABA (γ-aminobutyric acid) is the primary inhibitory neurotransmitter in mammalian brain and increasing GABA function via activation of the GABAA receptor complex results in increased chloride flux across the post-synaptic membrane (see 3 The GABA, 5 Compounds enhancing cerebral GABAergic function by acting at the GABA). There is good evidence that GABA exerts an inhibitory tone on glutamate mediated neuronal activity (Kanter et al., 1996) and several different approaches have revealed that increased GABA function can inhibit NMDA receptor-mediated responses. For example, application of GABA to cells can block NMDA stimulated calcium influx in the rat brain cortical slice (Riveros and Orrego, 1999) and cell death in the striato–nigral pathway induced by injection of the excitotoxin ibotenic acid (Saji and Reis, 1987). The GABAmimetic agents pentobarbitone and clomethiazole inhibit NMDLA-induced convulsions (Cross et al., 1993) and clomethiazole inhibits the NMDA-induced derangement of sensory evoked potentials (Thóren and Sjölander, 1993). More relevant to stroke research is the observation that the GABA agonist muscimol inhibited NMDA-induced neurotoxicity in primary cell cultures, an effect abolished by bicuculline (Ohkuma et al., 1994). Enhancing GABAergic inhibitory mechanisms in vivo therefore might also be expected to attenuate the excitotoxic process.
This review outlines the evidence that GABA function may be decreased following an ischaemic insult and then presents evidence that some, but by no means all, compounds that increase GABAergic function are neuroprotective. Because any clinical neuroprotective treatment regime will be initiated after the stroke, little weight has been placed on data on compounds that are only neuroprotective when administered to animals before the ischaemic insult. A compound must be protective when given some time after the start of the ischaemic episode for the finding to have clinical relevance (Green and Cross, 1997b, Hunter et al., 1995).
Section snippets
Changes in GABA biochemistry during an ischaemic episode
There is good evidence that an ischaemic episode results in a substantial acute increase in the extracellular concentration of GABA in the brain. Microdialysis probes implanted in selected brain regions have detected this increase in the permanent middle cerebral artery (MCA) occlusion model (Hagberg et al., 1985) the rat photochemical stroke model (Baldwin et al., 1993, Baldwin et al., 1994) and the gerbil global ischaemia model (Shuaib et al., 1994, Mainprize et al., 1995). Phillis and
The GABAA receptor complex
GABA actions are mediated by at least three different receptor classes. The GABAA receptor is linked to a chloride ion channel and is stimulated by GABA, muscimol and isoguvacine and inhibited by bicuculline (competitively) and picrotoxin (non-competitively). The GABAB receptor is linked via a second messenger system to Ca2+ channel inhibition or K+ channel activation and stimulated by baclofen and inhibited by phaclofen. The GABAC receptor is also linked to a chloride ion channel and may be
Drugs which increase synthesis or release
Various experimental attempts have been made to produce neuroprotection against ischaemic damage in the brain by increasing GABA function. However, any pharmacological approach to increase GABA synthesis and release requires the integrity of presynaptic GABA neurones. There are, in fact, not only in vitro observations that GABAergic neurones are relatively resistant to ischaemia-induced damage (Tecoma and Choi, 1989), but also crucially in vivo evidence using both focal (Johansen et al., 1991)
GABA agonists
Muscimol is a potent and selective GABAA agonist (Scotti DeCarolis et al., 1969) and there is substantial evidence that this compound has neuroprotective properties in a variety of experimental models of stroke. The earliest report was that of Sternau et al. (1989), who found muscimol to be efficacious in the gerbil global ischaemia model when given before the insult. A subsequent study also demonstrated neuroprotection when muscimol was infused intracerebroventricularly during three 2-min
Combination therapy
Lyden and Lonzo (1994) found that a combination of muscimol and dizocilpine provided enhanced neuroprotective efficacy in the microsphere embolism stroke model and it has been reported that administration of clomethiazole and dizocilpine provided at least additive protective effects against kainate-induced neurotoxic degeneration (MacGregor et al., 1997). The brain vacuolisation which can occur after administration of NMDA antagonists (Olney et al., 1989) is prevented by administering
Summary and conclusions
The evidence from experimental models of stroke that enhancing GABA function results in neuroprotection is considerable and there is a rational basis for this pharmacological approach to treatment. However, there are significant caveats. Most studies have been carried out using global models of ischaemia and many diverse compounds show efficacy in this model but are ineffective in focal models (particularly the MCA occlusion models) which are considered more relevant models of acute ischaemic
References (111)
Epidemiology of stroke
Lancet
(1992)- et al.
GABAC receptors
Trends in Neurosciences
(1995) - et al.
Stoichiometry of a ligand-gated ion channel determined by fluorescence energy transfer
Journal Biological Chemistry
(1999) - et al.
MK-801 is neuroprotective in gerbils when administered during the post-ischaemic period
Neuroscience
(1988) Clomethiazole (Zendra) in acute ischemic stroke: basic pharmacology and biochemistry and clinical efficacy
Pharmacology and Therapeutics
(1998)- et al.
Attenuation by chlormethiazole of oedema following focal ischaemia in the cerebral cortex of the rat
Neuroscience Letters
(1994) - et al.
The immediate consequences of middle cerebral artery occlusion on GABA synthesis in mouse cortex and cerebellum
Neuroscience Letters
(1992) - et al.
The neurotoxic effects of methamphetamine on 5-hydroxytryptamine and dopamine in brain: evidence for the protective effect of chlormethiazole
Neuropharmacology
(1992) - et al.
A behavioural and neurochemical study in rats of the pharmacology of loreclezole, a novel allosteric modulator of the GABAA receptor
Neuropharmacology
(1996) - et al.
Modulation of GABAA and glycine receptors by chlormethiazole
European Journal of Pharmacology
(1992)