Biophysical model of the olfactory system of honey bees predicts qualitatively different responses to mixtures
In their natural environment, insects typically encounter complex mixtures of odorants. It is an important open question whether and how responses to odorant mixtures differ from those to single components. To approach this question, we built a statistical model of the full receptor repertoire and antennal lobe of honey bees. Our model was developed to reproduce a variety of statistics of olfactory response patterns observed experimentally (Galizia, et al. (1999) Nat Neurosci), taking into account biophysical processes, such as receptor binding and activation, and spike generation and transmission in neurons. Our model can predict responses to both, single components and mixtures, and reproduces temporal and spatial features of neuronal odor representations that were not used when building it (Ditzen (2005) FU Berlin; Szyszka, et al. (2014) PNAS; Deisig, et al. (2010) J Neurophysiol). In particular, it reproduces the strong correlations among receptor neuron responses that weaken among projection neurons (Galizia, et al. (1999) Nat Neurosci), and suggests that this decorrelation is predominantly due to inhibition.
Using simulations and mathematical analysis, we found that the receptor dynamics alone already lead to significant differences between the responses to mixtures and to single components, long before any neural processing takes place. The model predicts that a) the response latency of olfactory receptor neurons decreases and b) response patterns become less variable across concentrations with increasing number of chemical components in the mixture. These effects are preserved in the projection neurons. We confirm our prediction for response latencies by single sensillum recordings in Drosophila and Ca imaging in projection neurons of the antennal lobe of bees. Our results suggest that the insect olfactory system encodes mixtures more efficiently than single odorants, which resonates well with the observation that chemical signaling in nature predominantly utilizes mixtures.

Acknowledgements
This work was supported by HFSP, RGP0053/2015 and EPSRC, EP/J019690/1.
Date: 11 October 2017, 13:30
Venue: Le Gros Clark Building, off South Parks Road OX1 3QX
Venue Details: Lecture theatre
Speaker: Prof Thomas Nowotny (University of Sussex)
Organisers: Dr Friedemann Zenke (University of Oxford), Dr Rafal Bogacz (University of Oxford), Dr Tim Vogels
Organiser contact email address: friedemann.zenke@cncb.ox.ac.uk
Part of: Oxford Neurotheory Forum
Topics:
Booking required?: Not required
Audience: Members of the University only
Editor: Friedemann Zenke