A model of competitive interactions among mixture components in early olfactory processing

MUEZZINOGLU M; HUERTA R.; LOCATELLI F; VILLARREAL, F.; GALIZIA G; SMITH B

San Diego, EEUU

Congreso; XL Annual meeting of the Society for Neuroscience; 2010

American Society for Neuroscience

A model of competitive interactions among mixture components in early olfactory processing M. K. MUEZZINOGLU1, R. HUERTA2, F. LOCATELLI3, F. VILLAREAL4, G. GALIZIA5, B. H. SMITH4; 1Inst. Nonlinear Sci., Univ. California San Diego, SAN DIEGO, CA; 2BioCircuits Inst., Univ. California San Diego, La Jolla, CA; 3Univ. de Buenos Aires, Buenos Aires, Argentina; 4Arizona State Univ., Tempe, AZ; 5Univ. Konstanz, Konstanz, Germany Associative plasticity modifies early olfactory processing both in the mammalian olfactory bulb (OB) and in the insect antennal lobe (AL). It is now well established that nonassociative plasticity also modifies activity in the AL and OB, but its mechanisms and effects are less well understood. Using calcium imaging in the honey bee AL, we showed that stimulation with a binary odor mixture (A+X) sets up a spatiotemporal ‘transient’ pattern of activity that is predictable from the transients for the two pure components. Unreinforced exposure to one component (A) leaves the component transients unchanged, but the transient to the mixture becomes less similar to the A and more like the transient for X, rendering the mixture representation in the AL perceptually more similar to X. These results suggest that nonassociative plasticity modifies the neural network in the AL in a way that affects local, competitive interactions among components. It also implies that the plasticity in the AL maximizes the information transmission by diminishing the AL activity of common odorants while enhancing other odors which are not so ordinary. This result matches what one would expect by maximizing Shannon information.In this study, we present a conductance-based single-compartment computational model of the honey bee AL that effectively explains the adaptation of the projection neuron (PN) trajectories as observed in the calcium readings. We hypothesize three Hebbian plasticity schemes and statistically evaluate them in explaining the phenomenon. The conclusion is that the most likely targets for modification are the inhibitory LN to PN synapses , which are potentiated on each reinforcement of the system by the training odor A.