Neural network models of chemotaxis in C. elegans
T. C. Ferrée, B. A. Marcotte, J. T. Pierce, S. R. Lockery.
Institute of Neuroscience, University of Oregon, Eugene, OR 97403.
Nematodes move up chemical gradients by detecting changes in concentration at the tip of the nose and biasing their normal sinusoidal locomotion toward higher concentrations (klinotaxis). Laser ablations of identified neurons in C. elegans have shown the neural circuit for chemotaxis is a highly interconnected network of chemosensory neurons, interneurons, and motor neurons with feedforward and feedback connections (C. Bargmann, unpublished). Moreover, physiological recordings from neurons in C. elegans suggest that they are functionally isopotential and that signal propagation is largely graded. To understand how such a nervous system might control chemotaxis we constructed a reduced model and used a neural network optimization algorithm (simulated annealing) to find sets of connections that produce chemotaxis. Many such sets were found. These results suggest that the biological network may be one of many different possible solutions for chemotaxis in graded neural networks.
Supported by NIMH MH11373, NIMH MH51383, NSF IBN 9458102, ONR N00014-94-1-0642, the Sloan Foundation, and the Searle Scholars Program.