Harvard University COVID-19 updates

Department News

Smelling Complexity [ Murthy Lab]

Smelling Complexity [ Murthy Lab]

You walk into a restaurant and immediately become aware of all the smells coming from the kitchen. These scents arise from small, volatile molecules called odorants. While there are a nearly infinite number of these odorants, the number of receptor types in your nose responsible for detecting these odorants is significantly smaller. In fact, in humans, there are only ~400 different types of receptors available to detect such complicated odor environments and transmit information to the brain. A fundamental question in olfaction, then, is how the olfactory receptor repertoire in the nose can detect such a vast array of molecules, while at the same time preserving the availability of some receptors to detect new smells that may be added to the environment. In other words, on top of detecting the complex smells coming from the kitchen, your nose should ideally leave some receptors available to smell your own meal when it arrives.

In a collaboration between the Vergassola lab at the University of California, San Diego and the Murthy lab at Harvard, published in eLife, Reddy et al. constructed a theoretical model describing how competitive interactions between odorant molecules can be used as a mechanism to help the brain parse complex odor environments. The model they constructed builds on previous theoretical work by incorporating a two-step process through which odors bind to, and then activate, receptors in the nose. In the first step, odor molecules compete for the same binding site of an olfactory receptor. Then, in the second step, the bound molecule goes on to generate activity in olfactory receptor neurons at differing levels of efficacy. A surprising effect occurs when the efficacy of an odor in activating different receptors is independent of how well it binds to these receptors: the number of receptors activated by a complex mixture is completely independent of how many components are present in the mixture. As the number of components in an odor mixture is increased, this independency of the binding-activation process is crucial for preventing saturation of the limited repertoire of olfactory receptors. This process, known as normalization, allows for new odors to be detected when added to complex mixtures and leads to superior performance in odor discrimination and identification tasks, as shown by the group through simulations.

Such a `normalization’ of activity is prevalent in other sensory modalities and brain regions, particularly throughout the visual cortex, and is thought to be a canonical neural computation. An important aspect of the model generated by Reddy and colleagues is that it provides a powerful mechanism for sensory normalization without the need for any recurrent neural processing. In effect, the competition between odorant molecules for the same receptors provides an indirect gain control mechanism that maximally preserves the information carrying capacity of the receptor neuron ensemble.

Reddy and colleagues also used their model to explain psychophysical observations from humans. In one such example, humans have reported a perceived decrease in odor intensity when the number of distractor odors present in a mixture increases. The model reliably captured this same phenomenon on the cellular level and provides a potential explanation for how we perceive mixtures of odors from a neurobiological perspective.

The results of Reddy et al. provide new theoretical evidence that competition between odorants in complex olfactory environments allows the brain to encode an immense number of odors despite a relatively limited complement of receptors. The model produced is not only relevant to olfaction, but also has broad applicability to the field of biological sensing.

by Gautam Reddy and Joseph Zak


Venki Murthy faculty profile

Murthy lab website

(l to r) Gautam Reddy (inset), Massimo Vergassola, Joe Zak, and Venki Murthy

(l to r) Gautam Reddy (inset), Massimo Vergassola, Joe Zak, and Venki Murthy