Top-Down Feedback Enables Flexible Coding Strategies in Olfactory Cortex

In chemical sensation, multiple models have been proposed to explain how odors are represented in olfactory cortex. One hypothesis is that the combinatorial identity of active neurons within sniff-related time windows is critical, whereas another model proposes that it is the temporal structure of neural activity that is essential for encoding odor information. Interestingly, we find that top-down feedback to main olfactory bulb dictates the information transmitted to piriform cortex and switches between these coding strategies. Using a detailed network model, we demonstrate that feedback control of inhibition influences the excitation-inhibition balance in mitral cells, thereby restructuring the dynamics of piriform cortical cells. This results in performance improvement in odor discrimination tasks. These findings present a framework for early olfactory computation, where top-down feedback to the bulb flexibly shapes the temporal structure of neural activity in piriform cortex, allowing early olfactory system to dynamically switch between two distinct coding models.