Modeling Hydra from neuron to muscle to behavior

Abstract

The fundamental question in neuroscience is to understand how single neurons combine to function as a network, and how their activity drives muscles to produce variety of behaviors to interact with the environment and maintain the dynamic equilibrium of the whole system. Hydra, with a simple nervous system and a transparent body that is ideal for recently developed calcium imaging techniques, provides an ideal system in which to build a complete model to understand neural activity and the transformation from neuron to muscle to behaviors. While recent calcium imaging techniques have been applied in recording both neuronal and muscle activities, the mechanisms and principles that support these dynamics are still rarely known. In this work, I build a complete model describing the neuromechanics, single neuron dynamics and the neural network of Hydra, interpreting the experimental findings from a biophysical perspective. I first review the basic facts about Hydra and some modeling approaches that provide a toolbox for our modeling work. Next, I summarize the questions raised by recent calcium imaging of Hydra. I develop a complete mechanistic model that simulates the transformation from neural drive to calcium dynamics in the muscular system, which are converted into active forces that drive a biomechanical model to generate different movements. The model addresses detailed questions such as the mechanisms underlying different timescales of muscle dynamics and the interaction of different muscle layers, and provides a testbed to understand how behaviors arise from specific neural drive. Thirdly, I explore the mechanism behind the bursting dynamics of the contraction burst network in Hydra. I propose that the periodic activity of single neurons is driven by mechanosensory signals due to water transportation in the myoepithelial tissue, and build an LIF-based neuronal model that successfully simulates the bursting pattern and reproduces the dynamical features observed in experiments. Finally, I construct a mutual inhibition network composed of two observed subnetworks in Hydra, simulating the network-level neural activity that has been recorded in both intact and bisected Hydra. Our models reveal a commonality between Hydra's nervous system and the autonomic nervous system in advanced organisms.