Dendrite Patterning as a Model for Self-Organizing Systems

Abstract

Biological systems self-organize into complex well-ordered structures and can evolve new patterns when perturbed. To identify principles underlying self-organization, I turned to the model organism C. elegans for its simple, highly reproducible anatomy and powerful genetics. In particular, I studied the sensory neuron PVD, which extends a highly elaborate dendrite arbor that covers the body wall, and the amphid sense organ, which consists of a bundle of 12 dendrites arranged in a highly reproducible order. First, I asked how the patterning of the PVD dendrite was affected when I introduced additional copies of this neuron. I found that the ectopic PVDs extended smaller arbors that filled the body wall without overlapping, a pattern called dendrite tiling that is not normally seen in C. elegans. I found that dendrite tiling arose as a byproduct of molecular pathways normally used for dendrite self-avoidance. This example shows how even a simple nervous system can re-purpose existing pathways to generate well-ordered structures in a way that could explain the evolution of new patterns. Next, I asked how the sensory dendrites of the amphid are organized into a bundle. Each amphid neuron extends an unbranched dendrite to the nose, together with the processes of two glial cells. Previous work suggested that the dendrites are bundled together in a reproducible order. I developed a method to visualize and quantify dendrite order in wild-type animals. Then, I used a candidate genetic screen to identify three CAMs that alter dendrite order. Loss of CDH-4/Fat-like cadherin causes a complete randomization of dendrite order. In contrast, loss of PTP-3/LAR or SAX-7/L1CAM surprisingly causes dendrites to take on a new, reproducible order. Further, misexpressing SAX-7 also leads to a new dendrite order. Altogether, my results suggest that differential expression of CAMs can organize dendrites in a way that is stereotyped across a wild-type population, yet that can easily give rise to novel patterns. These examples demonstrate how a simple nervous system makes use of self-organizing principles to generate ordered structures without sacrificing its evolvability.