Environmental adaptation of olfactory receptor distributions

Olfactory receptor usage is highly non-uniform, with some receptor types being orders of magnitude more abundant than others. Intriguingly, in mammals the receptors are replaced regularly and the receptor distribution changes in response to olfactory experience. We propose a theoretical model that explains these experimental observations. Our model suggests that the receptor distribution is tuned to maximize the amount of information about the olfactory environment, in accordance with the efficient coding hypothesis. The optimal receptor distribution depends on natural odor statistics, implying that a change in environment should lead to changes in receptor abundances. The effect is more pronounced when olfactory receptors respond strongly to a smaller number of odors. Our model also predicts that, between closely-related species, those with more neurons in the olfactory epithelium should express more receptor types. We show that simple dynamical rules for neural birth and death processes converge to the predicted optimum.