PubMed1993
In the olfactory bulb, the activity of the output neurons, the mitral cells, is under inhibitory control exerted by GABAergic interneurons, the granule cells. Although the mechanisms of this inhibition are well known from in vitro studies, its physiological role in controlling mitral cell activity in response to odours has never been investigated. This study planned to improve understanding of the involvement of granule cells. To do so, GABAA-synaptic mechanisms were altered using GABAA antagonists in order to observe the consequences on mitral cell electrophysiological responses to odours, delivered over a wide concentration range. Due to the laminar organization of bulbar cell populations, the antagonists picrotoxin or bicuculline were injected into the bulbar ventricle in order to block granule cell inhibitory action at first. Surprisingly, the early consequence of the antagonist injection was a decrease in cell responsivity: response spike frequencies were lowered while thresholds were occasionally shifted toward higher concentrations. This initial depressive effect was followed by a recovery of control excitability and, later, by an increase in excitability: spike bursts became more sustained in frequency and in duration. At the same time, in most of the cells studied, spontaneous activity became bursting. The early depressive effect of GABAA antagonists is discussed in terms of an enhancement of the inhibitory influence of granule cells on mitral cells. This might reflect a blocking action of the antagonists at the level of GABAergic synapses located on granule cells themselves. The late effect, an increase in excitability, is explained as the consequence of the alteration of the functioning of dendrodendritic synapses between granule and mitral cells leading to a disinhibition of the latter. The comparison of the present findings with others obtained when antagonists were applied on to glomerular layers led us to infer that granule cell inhibition would be devoted to limiting mitral cell responses in frequency and in duration rather than to adjusting their response threshold. The chronology of the effects observed strongly supports the fact that, following the intraventricular injection, the antagonists acted primarily in the deep layers of the bulb. Nevertheless, due to free diffusion starting from the injection site, the possibility that drugs act later in the glomerular layer can not be rejected. It can be concluded that, in addition to its extensive involvement through intrinsic interneurons, GABA might also control the strength of the inhibition exerted by granule cells on mitral cells via centrifugal fibres.