Animal brains integrate information to make crucial developmental and behavioral decisions. The brain adapts to life experiences by anatomical, functional, and molecular changes, but understanding the strategic value of these changes requires a comprehensive model that interconnects neural circuits and behavioral dynamics. To develop such models, it is useful to start with animals where entire brain circuits can be interrogated with full molecular, synaptic, and cellular resolution across development. A key decision made by Caenorhabditis elegans is whether to enter immediate reproductive development in good environments or diapause under poor environments. Environmental quality is assessed by a number of defined ethologically relevant signals: quantity and quality of food, population density via defined chemical social cues, and environmental stressors. This decision involves ~5% of the compact C. elegans nervous system, with at least five types of sensory neurons and five types of interneurons, yet is tractable enough to allow a comprehensive analysis. We will tightly couple modeling and experiment, designing perturbations and analyses to elucidate the pertinent dynamics of the decision-making. We will develop, test, and refine a model of the decision circuit starting with an algorithmic model based on decision theory and proceeding towards a circuit model that connects individual neurons and molecules to the decision circuit. We will apply recently molecular genetic tools for circuit manipulation and interrogation, define the transcriptomes of key cell types in the circuit to identify potential neuromodulatory inputs into the computation at longer time scales, and test how changes in the connectome caused by the decision modulate circuit dynamics and behavior. By functional imaging of the decision-making circuit, we will quantify total information flow throughout the worm nervous system that leads to the developmental decision. We will use a tracking microscope to optically record the dynamics of the decision-making circuit from sensory inputs to the brain starting from birth. We will use microfluidic and optogenetic rigs to control multidimensional sensory inputs on the 0.01-100 minute time scale to monitor and manipulate the decision-making circuit in real time. We will correlate neurophysiological analysis with endpoint analysis to map the dynamics of the developmental decision. We will enhance tools for efficient circuit analysis and define relevant neuromodulatory connections. To determine what neural circuit changes occur during and after the decision, we will use high-throughput methods in ssEM to reconstruct the connectome of animals, before, during, and after the decision is made. We will correlate functional imaging and anatomy to determine the functional basis of behavioral variation at all developmental stages. We will use single neuron microdissection to obtain gene expression profiles for about 10 neuron types that are key to the circuit of interest. We will store data in a project database, and visualize the system, results of simulations, and data including neural connectivity data in a 3D model through a network graph viewer and ontology-based browser.
Behavioral pathologies result from an animal's genotype and its life history, but it has not yet been possible to dissect the molecular, synaptic, and circuit differences that contribute to differences in animal behavior spanning an entire nervous system from birth through maturity. We propose to undertake analysis of an experience-dependent developmental decision in C. elegans ? the entry into diapause ? to determine how the animal uses a complex array of ethologically relevant cues to calculate a major developmental event that restructures brain, behavior and physiology. This project will provide computational models of an entire decision-making circuit that fully integrates anatomical, dynamical imaging, and behavioral data, as well as reagents and methods that will allow future extension to the entire nervous system.
