We are using C. elegans male sensory rays as a model to study how neurons form precise and selective connections with their targets during neuronal development. Each ray consists of a structural cell and two neurons (type A and B) with distinct position, and possibly neurotransmitter. For example, we have found that rays 1, 3 and 9B neurons express serotonin, while rays 5, 7 and 9A neurons synthesize dopamine (see Lints and Emmons). These results raise the possibility that A and B neurons in different rays may synapse on different targets. Though previous studies by Sulston et al (1980) indicated that the ray axons terminate in the preanal ganglion, the precise synaptic connectivity of A and B ray neurons is still unknown. We are using both genetic and molecular approaches to address these issues. By looking at the expression patterns of a
pkd-2::gfp reporter, which is strongly expressed by one neuron in each ray (except ray 6) (Barr et al, 1999), we find that axons of the ray neurons exhibit a stereotyped pattern. The axons of anterior rays first extend anteriorly (ray 1) or ventrally (rays 2, 3, 4 and 5). Ray 1 then progresses to the ventral side before turning posteriorly, while rays 2-5 turn anteriorly and all run into the preanal ganglion. Axons of rays 7, 8 and 9 migrate anteriorly and ventrally to the same region. However, it is still unclear whether the target is a single cell or several cells specific to different neurons. We intend to ablate the putative target(s) in the preanal ganglion (the EF interneurons) to address whether the development of stereotyped connection is target-dependent. Many transcription factors have been shown to regulate ray development. We found that axonal trajectories are dramatically changed in
egl-5 (a HOX gene) and
unc-37 (a groucho-like corepressor) mutants, indicating these genes specify neuronal connectivity. We are also investigating the LIM homeodomain proteins. To look for molecules involved in target recognition, we have begun a screen using
pkd-2::gfp transgenic lines, and obtained 5 mutants. Most of these mutants have abnormal axon patterns with anterior axons (always ray 1, also rays 2, 3 and/or 4 in some cases) extending anteriorly and apparently failing to reach their targets.