Segment Identities and the Paradoxical Specification of Regional Differences along the Anterior–Posterior Axis
The extent to which segmentation in leeches is viewed as being homonymous versus heteronomous or tagmatized is partly a matter of definition and partly a matter of how closely one looks. For the purposes of this discussion, the essential point is that there are reproducible regional and segment-specific differences along the anterior-posterior (А-P) axis. Most obvious is the organization of the four rostral segments and seven caudal segments (segments R1-R4 and C1-C7, respectively) into specialized anterior and posterior suckers, innervated by fused segmental ganglia (Figure 7.1). The four fused anterior ganglia connect via circumesophageal connectives to a non-segmental, dorsal anterior ganglion, comprising progeny descended from the first micromere quartet, which arises at third cleavage (Weisblat et al. 1984). In addition, and notwithstanding their rather uniform external morphologies, the midbody segments (segments M1-M21) also exhibit a variety of segment-specific features such as: specialization of segments 5 and 6 for male and female reproductive functions (Sawyer 1986), respectively; segment-specific distribution and orientation of metanephridia (Weisblat and Shankland 1985; Martindale and Shankland 1988); segment-specific distribution of primordial germ cells and definitive testisacs (Sawyer 1986; Kang et al. 2002); and segment-specific differences in the occurrence and distribution of various neuronal phenotypes (Stuart et al. 1987; Shankland and Martindale 1989; Martindale and Shankland 1990; Shankland et al. 1991).
There is also ample evidence for the specification of segmental differences within the endodermal derivatives of the midgut (Figure 7.1). The crop (anterior midgut) comprises six segmentally iterated, bilaterally paired outpocketings called crop ceca that serve in food storage; the posterior pair of crop ceca are particularly large and extend rearward through most of the midbody. Crop ceca are particularly important for the blood-feeding leeches, which may feed only rarely and yet can take in several times their own weight at one feeding. The intestine (posterior midgut) features smaller and more uniformly sized ceca.
These morphological markers indicate that segments acquire specific identities along the А-P axis in leeches. Molecular markers corresponding to segment-specific identities come largely from analyses of the expression of the leech Hox genes, but this work has led to fascinating new puzzles as well.
Initial studies of leech Hox genes were carried out without benefit of having a sequenced genome or transcriptome. A number of putative Hox orthologs were identified, however, and their expression patterns were interpreted as showing the expected succession of А-P boundaries consistent with the canonical Hox clusters of Drosophila and vertebrates, all of which seems in line with the conclusions that (1) Hox genes function to specify segment or regional identities along the А-P axis, and; (2) their expression is driven by spatiotemporal colinearity responding to their sequence w'ithin the evolutionarily conserved Hox clusters (Kourakis et al. 1997). However, a significant body of work has revealed that the situation is more complicated than it first appeared.
First, a variety of ablation experiments indicate that ectodermal blast cells have assumed specific identities at least by the time they enter the germinal bands. For example, as described earlier, blast cell ablation experiments were used to confirm the commitments of blast cells to their f and s identities in the grandparental N lineage (Bissen and Weisblat 1987; Shain et al. 2000). More definitively, precise combinations of lineage tracing and immunostaining were used to identify segment-specific patterns of presumptive peptidergic neurons within the N lineage (Shankland and Martindale 1989). Then, blast cell ablations were carried out to induce the rearward slippage of о blast cells as previously demonstrated (Shankland 1984; Martindale and Shankland 1990). The experimentally induced slippage was timed to reposition о blast cells into segments posterior to those that would normally contain the identifiable peptidergic neurons—the repositioned о blast cells still gave rise to the kinship group appropriate to the segment for which it was originally destined rather than adapting to its new position. This result indicates that the blast cell had already acquired a segmental identity by the time it had entered the germinal band, well prior to its first mitosis. Similar approaches yielded similar results for blast cells in the mesodermal (M) lineages of a different glossiphoniid leech species (Gleizer and Stent 1993). An important advance was that in this work, it was possible to ablate individual m blast cells soon after they had been produced by the parent M teloblasts when they were only a few hours in the lineage.
Thus, the results indicated that these m blast cells had acquired their segment- specific fates at or within a couple of hours after their birth from the parent stem cell. And yet, most of the leech Hox genes that have been examined are not expressed until much later, when the blast cells have undergone many subsequent divisions and the differentiation of segments is w'ell underway within the germinal plate (Kourakis et al. 1997; Nardelli-Haefliger et al. 1994; Cho et al. in preparation).
Moreover, notwithstanding the general result that Hox gene expression boundaries fall within particular segments across multiple teloblast lineages, blast cell ablation experiments followed by immunostaining revealed that these boundaries of Hox gene expression are established by the cells of each lineage independently of interactions with other lineages, as are the other segment-specific markers (Nardelli- Haefliger et al. 1994).
Yet another aspect of this puzzle emerged when whole genome sequencing was carried out on three lophotrochozoan species (Simakov et al. 2013). The genomes of a mollusk (Lottia) and a polychaete annelid (Capitella) each contain 11 mutually orthologous Hox genes. The molluscan Hox genes appear in a single genomic scaffold and the polychaete genes appear in two or possibly three scaffolds, but lie in the same syntenic order as the molluscan genes, in accord with expectations based on vertebrates and Drosophila. In stark contrast, the Helobdella genome contains 18 or 19 Hox genes with multiple duplications (two paralogs each of labial, deformed, and lox4 three of post2 and five of sex combs reduced) and loss of proboscipedia and postl. The organization of the cluster is also severely disrupted. Many genes appear as singletons, flanked by non-Hox genes, and the longest single cluster contains only four genes, none of which are neighbors in the inferred ancestral lophotrochozoan Hox cluster (Simakov et al. 2013).
Thus, since the last shared ancestor of Capitella and Helobdella, there has been an extensive disruption of the annelid Hox cluster in the evolutionary lineage leading to leeches. This disruption is emblematic of a comparably extensive rearrangement of the entire leech genome relative to other sequenced taxa; available evidence suggests that such rearrangements are characteristic of clitellate annelids as a group.
In summary then, it appears that the segmental founder cells in leeches acquire specific segmental identities before they undergo their first mitoses and possibly as they are born from the parent stem cells/teloblasts. In addition, these segmental identities are assigned independently within each lineage and before the onset of most Hox gene expression, by mechanisms that remain to be determined. And yet, despite the extensive disruption of the ancestral Hox cluster organization, these genes are expressed in patterns that are fairly consistent with those inferred for the ancestral annelid Hox genes.