Cell-cell communication in nematode embryos: differences between Cephalobus spec. and Caenorhabditis elegans

During early nematode embryogenesis a series of asymmetric cleavages in the germ line generates several somatic founder cells and a primordial germ cell. We have found previously that the two soil nematodes Cephalobus spec. and Caenorhabditis elegans express considerable differences in the order of events and spatial arrangement of cells during early embryogenesis. With the help of microinjected fluorescent marker dyes, we show here that these dissimilarities partner major differences in the pattern of intercellular communication. Whilst in C. elegans all early blastomeres become dye-coupled simultaneously, in Cephalobus communication is established progressively in the sequence in which cells are born. In addition, in Cephalobus but not C. elegans, sequential lucifer yellow accumulation indicates stepwise changes in the state of early blastomeres: if injected into the uncleaved zygote, for example, the dye becomes equally distributed to all cells at first but rapidly accumulates in a single blastomere in the 4-cell stage. We speculate that such a redistribution mechanism may be involved in the differential segregation of cytoplasmic components to individual blastomeres. The most dramatic difference between the two species was found with respect to the transfer of high molecular weight molecules. In contrast to C. elegans, in Cephalobus not only small lucifer dyes but also high molecular weight dextrans can diffuse along specific pathways between early somatic cells indicating the presence of large communication channels. However, a transfer of dextran into or out of germ line cells never takes place. The origin of these channels as midbodies of previous mitoses and their potential role for normal development is discussed. Tissue-specific dye-coupling compartments in the slow developing Cephalobus are established in the same order but at a considerably earlier developmental stage than in C. elegans suggesting that this process may depend more on parameters like available time for transcription rather than the number of cell cycles passed through.