The mode of embryonic and larval development and the ethology of metamorphosis in the spider crab and the American lobster are very different, and we took advantage of this to compare neuronal development in the two species. The goals of this study were to discover whether the differences in the maturation of the neuromuscular system in the pereopods and the metamorphic changes of motor behavior between the two species are reflected at the level of the developing nervous system ('neurometamorphosis'). Furthermore, we wanted to broaden our understanding of the mechanisms that govern neuronal development in arthropods. Proliferation of neuronal stem cells in thoracic neuromeres 4-8 of the lobster Homarus americanus and the crab Hyas araneus was monitored over the course of embryonic and larval development using the in vivo incorporation of bromodeoxyuridine (BrdU). Neuropil structure was visualized using an antibody against Drosophila synapsin. While proliferation of neuronal precursors has ceased when embryogenesis is 80 % complete (E80%) in the lobster thoracic neuromeres, proliferation of neuroblasts in the crab persists throughout embryonic development and into larval life. The divergent temporal patterns of neurogenesis in the two crustacean species can be correlated with differences in larval life style and in the degree of maturation of the thoracic legs during metamorphic development. Several unusual aspects of neurogenesis reported here distinguish these crustaceans from other arthropods. Lobsters apparently lack a postembryonic period of proliferation in the thoracic neuromeres despite the metamorphic remodeling that takes place in the larval stages. In contrast, an increase in mitotic activity towards the end of embryonic development is found in crabs, and neuroblast proliferation persists throughout the process of hatching into the larval stages. In both E20% lobster embryos and mid-embryonic crabs, expression of engrailed was found in a corresponding set of neurons and putative glial cells at the posterior neuromere border, suggesting that these cells have acquired similar specific identities and might, therefore, be homologous. None of the BrdU-labeled neuroblasts (typically 6-8 per hemineuromere over a long period of embryogenesis) was positive for engrailed at this and subsequent stages. Our findings are discussed in relation to the spatial and temporal patterns of neurogenesis in insects.
Embryonic and larval development of Thai Pangas (Pangasius sutchi Fowler, 1937)
