Reverse-Correlation Analysis of Mechanosensation Circuit in C. elegans Reveals Temporal and Spatial Encoding

As animals navigate through complex environments, they must integrate the activity of multiple mechanoreceptors, sensing forces throughout their bodies and allowing them to move in appropriate directions. In Caenorhabditis elegans, the only organism with a fully mapped connectome, the neural circuit involved in mechanosensation is well characterized. Although the general roles of the neurons in this circuit have been defined, most studies involve experiments with a small number of unnatural stimuli, leading to quantitative descriptions that may be biased towards the tested stimuli. In this work, we elucidate unbiased descriptions of the mechanosensory system in C. elegans by using reverse correlation analysis. We use a custom tracking and optogenetics platform to characterize and compare two mechanosensory systems in C. elegans: the gentle touch sensing TRNs and harsh touch sensing PVDs. This method yields linear filters that capture dynamics that are consistent with previous findings, as well as providing new insights on the spatial encoding of the TRN and PVD neurons. Our results suggest that the tiled network of the TRNs allow for spatial encoding with better resolution than PVD. Additionally, linear-nonlinear models accurately predict behavioral responses based only on sensory neuron activity. Our results capture the overall dynamics of behavior induced by the activation of sensory neurons, providing simple transformations that fully characterize these systems.