Multiple facets of GABAergic neurons and synapses: multiple fates of GABA signalling in epilepsies

Because blocking GABAergic neurotransmission in control tissue generates seizures and because GABA boosters control epilepsy in many patients, studies on epilepsies have been dominated by the axiom that seizures are generated by a failure of GABA-mediated inhibition. However, GABAergic interneurons and synapses are heterogeneous and have many roles that go beyond the straightforward concept of ‘inhibition of the target’. Operation of such a diversified system cannot be ascribed to a single mechanism. In epileptic tissue, GABAergic networks undergo complex rewiring at the anatomical, physiological and functional levels; GABAergic synapses are still operative but show unique features, including excitatory effects. Therefore, inhibition is not a uniform notion and the concept of ‘failure’ of inhibition in epilepsies must be reassessed. Seizures are not generated in a normal circuit in which GABA-mediated inhibition is simply impaired, but in a profoundly rewired network in which several properties of GABA function are altered. This review is part of the TINS Interneuron Diversity series.

Introduction

The starting point for conventional understanding of epileptic seizures is a simple model whereby inhibition and excitation have roles somewhat similar to those of the brakes and accelerator in an engine, respectively. Consequently, agents and conditions that reduce the former or augment the latter will inevitably favour the generation of seizures. In keeping with this, blockers of GABA-mediated transmission generate seizures in a wide range of preparations and in humans, suggesting that GABAergic inputs prevent this generation. Conversely, strong activation of glutamatergic synapses generates seizures. The consequence of this simplistic scheme is that agents that augment inhibition will have antiepileptic properties.

It is now clear that this message is oversimplified. First, the concept of GABA-mediated ‘inhibition’ must be re-evaluated. Recent data suggest that GABA neurotransmission can be excitatory in basal conditions, not only in immature but also in adult tissue. GABAergic synapses are indeed endowed with a unique and fundamental feature – resulting from their Cl⁻ permeability – that enables them to shift from an inhibitory mode of operation to one that mainly excites. This feature is not shared by the cationic glutamate receptors, which do not shift to inhibition in pathological conditions. Second, the simple static view of ‘inhibition of the target’ is ‘the tree that hides the forest’. One major function of GABA-mediated neurotransmission is to synchronize neuronal networks. In such dynamic perspective, the role of GABA extends well beyond mere ‘excitation’ and ‘inhibition’. Third, studies on interneurons have unraveled an unsuspected heterogeneity of neuronal types that have very different roles. Such heterogeneity is an important parameter to consider when looking at the roles of GABAergic neurotransmission. Clearly, study of the fate of GABA-mediated functions in an epileptic neuronal network must take into account such complexity. The modifications that take place at various levels within GABAergic networks will be discussed in this review. But before addressing these issues, it is necessary to recall the modus operandi of GABA-mediated neurotransmission and interneuron heterogeneity in control tissue.