Proceedings of The Physiological Society

University of Cambridge (2008) Proc Physiol Soc 11, C78

Oral Communications

Odor learning under anaesthesia: behavioural, neurochemical and electrophysiological effects

A. Nicol1, G. Sanchez-Andrade1, H. Fischer1, P. Collado2, A. Segonds-Pichon1, K. Kendrick1

1. Babraham Institute, Cambridge, United Kingdom. 2. Cd Universitaria, Madrid, United Kingdom.


Carbon disulphide (CS2) in the exhaled breath of rodents increases the attractiveness of food odours. Thus information about foods which are safe to eat is transmitted between individuals. When one individual smells a food odour on the breath of another, it eats more of that food than it would otherwise. Here we investigate this system of odour learning in anaesthetised mice. Food odours were prepared by mixing the ingredient with normal feed, and introducing the altered food in non-airtight capsules into gas-sampling bags containing N2. Bags were also prepared with CS2 (10µM), or the unaltered feed, both in N2. Mice were anaesthetised with isofluorane in 21% N2 and 79% O2. Stimuli were delivered by switching from this supply to a matched supply carrying a food odour and/or CS2. In experiment 1, Mice were trained by exposure to a novel food odour combined with CS2 for 60s, then allowed to recover from anaesthesia. Controls were exposed to a novel food odour alone, or to CS2 alone. When tested 24h after odour exposure, food preference was biased towards an odour that had been presented with CS2 during anaesthesia. In experiment 2, mice were anaesthetised and a microelectrode array (6x4 tungsten electrodes, tip separation 350µ) was positioned in the OB. Action potentials (spikes) were sampled from neurons in the mitral cell layer of the OB (≤8 neurons per electrode). Ten second presentations were made of the plain food odor, and the odors of food with ginger or coriander added: ~10 trials per odor with 5min inter trial intervals. Ten presentations were then made of the ginger food odor combined with CS2. In subsequent tests with the altered and plain food odors, neuronal responsiveness was biased towards the training odor relative to either of the other two odors. In experiment 3, mice were anaesthetised and trained by nose to nose contact with an anaesthetised mouse which had consumed a novel flavoured food (trained group), or plain food (control group). When tested 24h after odour exposure, food preference again reflected odour exposure during anaesthesia. Immediately after testing, the mice were re-anaesthetised, and neurochemical samples were collected from the OB by microdialysis during three consecutive 1min deliveries of the flavoured or plain food. In animals trained with flavoured food odour during anaesthesia, presentation of the flavoured food odour, but not the plain food odour, produced an increase in glutamate (142.5%±20.4SEM), GABA (155.3%±19.2) and noradrenaline (137.3%±28.4) in the OB. This was not so in animals exposed only to plain food odour under anaesthesia. These experiments demonstrate that this system of odour learning functions effectively in anaesthetised mice exposed to an artificial mix of food odour with CS2. Neurochemical changes and altered OB neuronal responsiveness also reflect odour exposure during anaesthesia.

Where applicable, experiments conform with Society ethical requirements