To learn how neural segments are structured in a simple vertebrate, we have characterized the embryonic zebrafish hindbrain with a library of monoclonal antibodies. Two regions repeat in an alternating pattern along a series of seven segments. One, the neuromere centers, contains the first basal plate neurons to develop and the first neuropil. The other region, surrounding the segment boundaries, contains the first neurons to develop in the alar plate. The projection patterns of these neurons differ: those in the segment centers have descending axons, while those in the border regions form ventral commissures. A row of glial fiber bundles forms a curtain-like structure between each center and border region. Specific features of the individual hindbrain segments in the series arise within this general framework. We suggest that a cryptic simplicity underlies the eventual complex structure that develops from this region of the CNS. axons at about 18 h (Mendelson, 1986b). As embryogenesis continues, neurons are added to the segments in a stereotyped manner(Hanneman et al., 1988). Shortly after the first neurons appear, additional ones develop nearby, forming single clusters of young neurons bilaterally in each segment. Another population of neurons appears later near the segment boundaries. The arrangement of these two neuronal sets, in the center and border regions, produces a striking alternating pattern at 24 h (Hanneman et al., 1988), reminiscent of the anterior and posterior halves of vertebrate somites (Keynes and Stern, 1984) and the anterior and posterior compartments of Drosophila segments(Garcia-Bellido et al., 1973; Morata and Lawrence, 1975).

We have further characterized the structure of zebrafish hindbrain segments and have found new evidence for their division into two separate regions. We generated a library of monoclonal antibodies and used them to determine the arrangements of specific groups of neurons, commissural tracts, neuropil areas, and radial glial fibers. Our findings strongly support the division of the segments into two parts. Furthermore, they suggest possible mechanisms by which the two regions might be kept separate, bow axons may be guided through them, and how differences between the individual segments may arise within this general plan.