Role of antennal lobe inhibition in gain modulation and odor generalization across concentration

AYELEN NALLY; EMILIANO MARACHLIAN; FERNANDO LOCATELLI; RAMÓN HUERTA

Bordeaux

Conferencia; NeuroFrance 2017; 2017

Animals extract ecologically relevant information from the environment to form an internal image of the surroundings through a multi-sensory system. Different sensory modalities must be weighted and integrated to allow stimuli recognition and guide proper behavioral decisions. For this comparison the stimulus has to be encoded regardless of its intensity (in some range) and unaffected by complex or noisy backgrounds. In addition, the neuronal convergence that takes place at the first level of sensory systems requires signal adaptation. Therefore, the local inhibitory network at early sensory processing plays a fundamental role in sensory systems.In this work we use honey bees as model animal for studying odor generalization across intensities and to understand the neural computations that underlie this ability. Odors are detected and primarily encoded by the olfactory sensory neurons (approximately 60,000 that converge into 800 projection neurons organized in 160 glomeruli in the antennal lobe). Each specific odor activates a unique combination of receptors that provides the input for its internal representation. However odors presented at different concentrations produce different activity patterns, mainly in terms of the intensity of activation but also in the combination of receptors. The question is how animals recognize the same odor across different concentrations in spite of the different input patterns.Using different GABA blockers we describe the contribution of the local inhibitory network in stabilizing odor patterns across intensity. We found that GABA-A and GABA-B components contribute differentially in terms of the intensity range of odors and in the temporal profile of the activation. We perform calcium imaging to describe neural representation of high and low odor concentrations in the antennal lobe under control of the GABAergic neurotransmission and without it. We describe also behavioral generalization across odor concentrations. Finally, the results are formalized in a computational model of the antennal lobe that provides detail of the inhibitory network and the role of GABA. The simplicity and robustness of the circuit suggests that similar versions might be involved in other sensory systems and even at higher processing levels.