Antagonism in olfactory receptor neurons and its implications for the perception of odor mixtures

Gautam Reddy, Joseph D. Zak, Massimo Vergassola, Venkatesh N. Murthy

TL;DR

This paper presents a biophysical model of olfactory receptor neurons (ORNs) that explains how antagonistic interactions between odorants lead to a "normalization" of the overall neural signal. This normalization mechanism, arising from the distinct statistical properties of odorant binding and activation, improves odor discrimination and explains human perception of complex smells.

What Problem Does This Paper Address?

The natural world bombards our olfactory system with complex mixtures of odors. While we can often distinguish individual scents within these mixtures, the initial processing by olfactory receptor neurons (ORNs) is known to be highly non-linear and often antagonistic. This means that the response to a mixture isn't simply the sum of individual responses; instead, some odorants can suppress the activity of others. The challenge is to understand how these complex, non-additive interactions at the ORN level contribute to our ability to perceive and discriminate odors, and whether these interactions have a functional role beyond simply being a biological quirk. Specifically, the authors aim to bridge the gap between the known non-additive responses of individual ORNs and the overall glomerular response in the olfactory bulb, ultimately linking this to perceptual phenomena.

Methods

The authors developed a biophysical model of mammalian ORNs based on a two-step activation process: odorant binding to a receptor and subsequent activation. This model incorporates parameters for receptor binding affinity and activation efficacy. They then extended this to model mixture responses, focusing on competitive antagonism where odorants compete for receptor binding. Crucially, they also incorporated a "masking" mechanism, a form of non-competitive antagonism, to explain a broader range of observed phenomena.

To validate their model, they performed in vivo two-photon imaging of ORN axonal terminals in the olfactory bulb of live, breathing mice. They measured calcium signals in response to single odorants and binary mixtures at various concentrations. The model's parameters were then fitted to this experimental data.

For analyzing encoding and performance, they developed a computational framework that defines odorants by vectors of sensitivities and activation efficacies across a large number of ORN types. They introduced an "antagonistic factor" (ρ) to quantify the degree of antagonism and explored its impact on odor discrimination tasks using information theory and linear classifiers.

Datasets

The study utilized experimental data collected from in vivo imaging of ORN axonal terminals in OMP-GCaMP3 mice. Specific odorant pairs were used: ethyl valerate and allyl butyrate (pair one, with overlapping glomerular responses) and methyl tiglate and isobutyl propionate (pair two, with more distinct glomerular responses). The authors also referenced existing experimental data on masking agents to validate their masking model. No new benchmarks were introduced; rather, the study aimed to explain existing psychophysical observations.

Key Insights and Ideas

  • Antagonism as a Feature, Not a Bug: The core idea is that the non-additive, antagonistic interactions in ORNs are not just noise but a fundamental mechanism for efficient olfactory processing.
  • Normalization via Antagonism: The model demonstrates that competitive antagonism naturally leads to a "normalization" of the overall glomerular response. This means the total neural signal doesn't explode with increasing complexity but stays within a manageable range, preventing saturation.
  • Distinct Binding and Activation Properties: The model hinges on the separation of odorant binding affinity (how well it sticks) and activation efficacy (how much signal it generates once bound). Competitive antagonism arises when an odorant binds strongly but activates weakly, or vice-versa, relative to other odorants.
  • Masking Explains Synergy and Inhibition: The inclusion of a non-competitive masking mechanism is crucial for explaining a wider range of perceptual phenomena, including synergy (where a mixture elicits a stronger response than expected) and inhibition.
  • Linking Receptor-Level Dynamics to Perception: The paper successfully connects biophysical ORN responses to higher-level perceptual phenomena like masking, synergy, suppression, and overshadowing, providing a neurobiological basis for these experiences.

Key Results

  • The biophysical model accurately captures experimentally observed ORN response curves to single odorants and mixtures, including non-additive and inhibitory interactions.
  • The model predicts that antagonism leads to a "normalization" of the ensemble glomerular response, which is crucial for maintaining signal integrity and preventing saturation.
  • Antagonistic encoding (with ρ=0) significantly outperforms non-antagonistic encoding (ρ=1) in odor discrimination tasks, especially in the presence of increasing background odor complexity.
  • The model successfully explains various psychophysical observations, including inhibition, synergy, masking, symmetric/asymmetric suppression, and overshadowing. For instance, Figure 7 shows how different levels of antagonism and masking can reproduce these perceptual effects.

Limitations and Future Work

  • Inferred Parameters: The parameters for odorant-receptor interactions (κ and η) were inferred indirectly from mixture response curves, leading to potentially noisy estimates. The authors suggest that direct measurement in single ORNs would be beneficial.
  • Masking Mechanism Simplification: The masking model is phenomenological and assumes masking agents bind to lipid bilayers. While it explains observed effects, a more detailed biophysical mechanism for masking could be explored.
  • Generality of Antagonism: While the study highlights the benefits of antagonism, the authors note that structural constraints at the receptor level might prevent perfect decorrelation (ρ=0), and the extent of these benefits depends on various factors like odorant properties and mixture complexity.
  • Future Directions: The authors suggest further exploration of non-competitive antagonistic mechanisms in mice, direct measurement of single ORN responses, and leveraging experimental techniques to further investigate the role of ORN antagonism in psychophysical phenomena.

Summary updated

Abstract

Abstract Natural environments feature mixtures of odorants of diverse quantities, qualities and complexities. Olfactory receptor neurons (ORNs) are the first layer in the sensory pathway and transmit the olfactory signal to higher regions of the brain. Yet, the response of ORNs to mixtures is strongly non-additive, and exhibits antagonistic interactions among odorants. Here, we model the processing of mixtures by mammalian ORNs, focusing on the role of inhibitory mechanisms. Theoretically predicted response curves capture experimentally determined glomerular responses imaged by a calcium indicator expressed in ORNs of live, breathing mice. Antagonism leads to an effective “normalization” of the ensemble glomerular response, which arises from a novel mechanism involving the distinct statistical properties of receptor binding and activation, without any recurrent neuronal circuitry. Normalization allows our encoding model to outperform noninteracting models in odor discrimination tasks, and to explain several psychophysical experiments in humans.
BibTeX
@inproceedings{reddy2018antagonism,
  author = {Gautam Reddy and Joseph D. Zak and Massimo Vergassola and Venkatesh N. Murthy},
  title = {{Antagonism in olfactory receptor neurons and its implications for the perception of odor mixtures}},
  booktitle = {eLife},
  year = {2018},
  doi = {10.7554/elife.34958},
}
Figure