Expression Patterns of Drosophila Segmentation Gene Homologs
“Modern” studies of segmentation in Helobdella were initiated in the 1970s, prior to the reorganization of animal phylogenies based on molecular sequence comparisons. At that time, it was commonly assumed that annelids and arthropods arose from a segmented common ancestor. In light of this assumption, which is now very much in doubt, the dramatic discoveries elucidating the molecular genetic basis of segmentation in Drosophila led to speculation as to how the various types of “segmentation genes” might function in what was obviously the very different cellular process of segmentation that was being worked out for leeches.
Within the prevailing belief that annelids and arthropods share a segmented common ancestor, for example, the discovery of the grandparental N and Q lineages in Helobdella resonated with the anterior and posterior compartments of Drosophila segments. This led to the hypothesis that engrailed or other segment polarity genes might be specifying the differences between the f and s blast cells in the N and Q lineages. Initial immunohistochemical studies characterizing the expression of an Engrailed homolog in segmentally iterated transverse stripes in the Helobdella germinal plate seemed to support this hypothesis (Wedeen and Weisblat 1991; bans et al. 1993; Ramirez et al. 1995), but the late onset of Engrailed expression was somewhat disconcerting, as was the eventual realization that Engrailed-expressing cells appear to play no role in regulating segment polarity in leeches (Seaver and Shankland 2001) and that the stripes of Engrailed-expressing cells only appear after morphological events of segmentation were well underway (Shain et al. 1998). In other work, the leech hedgehog gene was characterized, but turned out not to be expressed in a segmentally iterated pattern within the germinal plate during the early process of segmentation; more as in vertebrates, the leech hedgehog gene plays an important role in gut formation (Kang et al. 2003). Thus, despite considerable effort, there is no evidence as yet that the segment polarity genes identified in Drosophila play homologous roles in Helobdella segmentation.
An alternative possibility for establishing homology at the cellular and molecular levels between Drosophila segmentation genes and the f and s cells of the grandpa- rental teloblast lineages in Helobdella was via homologs of the pair-rule genes. To investigate this possibility, homologs of even-skipped (eve) and hairy/enhancer-of- split (hes) were characterized (Song et al. 2002,2004). Curiously, both genes proved to be expressed at the earliest zygotic stages. This expression was maintained in apparently all the teloblasts and early blast cells of the PGZ, but no differences in expression between the f and s cells of the N or Q lineages were observed.
Even more curiously, transcripts for both hes and eve genes were associated with the chromatin of dividing cells (Song et al. 2002, 2004); in the case of the leech hes homolog, it was shown that transcripts accumulated during mitosis and nuclear accumulation of translated protein began as cells exited mitosis. The significance of this antiphasic, cell cycle-driven expression remains to be determined. The injection of anti- sense morpholinos targeting eve expression led to a partial disruption of germinal band morphology and of neurogenesis in the injected lineage, but both these genes are ripe for reinvestigation using genome editing approaches that were unavailable at the time.
Yet another attempt to draw parallels between the segmentation processes in leeches and flies involved the characterization of a leech nanos gene (Pilon and Weisblat 1997). In this undertaking, the rationale was that since nanos expression marks the posterior pole of the early Drosophila embryo, it might either be expressed in the teloblasts of the PGZ throughout its function, since they constitute the posterior pole of the growing embryo, or perhaps just during their last divisions, which generate founder cells for the posteriormost segments. Rather than either of these possibilities, however, the leech nanos is an abundant and broadly expressed maternal transcript that declines markedly during cleavage, is essentially absent from the PGZ, and then is reexpressed broadly within the germinal bands and germinal plate, before finally becoming restricted to presumptive germline precursors (Kang et al. 2002). During cleavage, expression of Nanos protein appears to turn on during the two-cell stage and peak at stage 4b. At this stage, both transcript and protein levels are higher in DNOPQ than in sister cell DM (Pilon and Weisblat 1997).
In summary, while the leech homologs of Drosophila segmentation genes all show interesting and/or suggestive patterns of expression, there is no obvious point of homology with the critical roles of related genes in Drosophila segmentation. In retrospect, these results are not surprising given the vast phylogenetic separation between annelids and arthropods and the real possibility that segmentation arose independently in Ecdysozoa and Lophotrochozoa. Given the complex signaling requirements associated with evolution of segmentation in any taxon, we do not find it surprising that the transcription factor pair-rule genes eve and lies are expressed in the PGZ, along with other widely used developmental regulators in the Notch, WNT, and BMP signaling pathways (Rivera et al. 2005, Rivera and Weisblat 2009; Cho et al. 2010; Yoo et al. in preparation; Kuo and Weisblat 2011), some of which regulate segmentation of vertebrates and/or arthropods.