Grandparental Stem Cell Lineages
Analyzing the segmental contributions of individual blast cells in the N and Q lineages, as described earlier for the M, O, and P lineages, also yielded surprising results (Figure 7.4). Namely, in these lineages, the respective blast cells adopt two different fates in exact alternation. In addition to generating distinct complements of final, differentiated progeny to the respective N and Q kinship groups, the different blast cell types exhibit characteristic differences in the timing and asymmetry of their subsequent mitoses (Zackson 1984; Bissen and Weisblat 1987, 1989; Zhang and Weisblat 2005). These distinct blast cell identities are designated f and s, based on differences in the timing of their initial mitosis. Cells nf and qf have shorter cell cycle times than cells ns and qs, respectively.
One possible explanation for the fact that n (and q) blast cells arise as two distinct cell types in exact alternation is that the N (and Q) teloblasts from which these cells arise switch back and forth with each division, in terms of the two different cell types produced in these lineages. Thus, if we define canonical stem cell divisions as being “parental” in nature because one daughter cell at each division recapitulates the fate of the parent cells, then the N and Q teloblasts are undergoing “grandparental divisions.” The other possible explanation is that the differences between f and s blast cell types are imposed only after the cells arise from the N and Q teloblasts, through precise patterning processes of one sort or another. Unfortunately, currently available evidence (summarized later) is not conclusive in favor of either the intrinsic or extrinsic models for specifying the f and s fates in the N and Q lineages. Therefore, we designate these intriguing phenomena as grandparental stem cell lineages, rather than as grandparental stem cells per se.
Given the differences between f and s blast cells in cell cycle duration, it was originally thought that differences in the duration of their G1 phase might yield information about when the f and s cells become different, and this possibility was addressed using BrdUTP incorporation to label S phase nuclei in carefully timed experiments (Bissen and Weisblat 1989). This approach revealed that blast cells enter
S phase immediately upon birth from the teloblast, and that the differences in cell cycle duration reflect differences in duration of the G2 phase. However, dye-coupling analysis revealed differences in the persistence of cytoplasmic bridges between the parent teloblast and nascent nf and ns blast cells, which suggested the possibility of more subtle differences in the cell cycles by which N teloblasts give rise to nf and ns blast cells (Bissen and Weisblat 1987). In another approach, photoablating individual blast cells within the n bandlet revealed that the flanking cells within the bandlet have assumed distinct nf or ns identities prior to their first mitoses (Bissen and Weisblat 1987). Unfortunately, experimental constraints imposed by the geometry of the embryos made it impossible to carry out these experiments until the blast cell had entered the germinal bands, which is many hours after their birth from the parent teloblast. Intriguingly, such ablation experiments have also revealed differences in cellular dynamics between nf and ns blast cells. As will be discussed further elsewhere (Section 7.13), it had been shown that killing one blast cell in a germinal bandlet has the potential to delay the forward progress of cells in that bandlet that are behind the lesioned cell (Shankland 1984). When this experiment was carried out in the N teloblast-derived bandlet, it was discovered that the effect of the lesion depends in part on the identity (nf or ns) of the cell that was killed (Shain et al. 2000). Specifically, when an nf cell is killed, the two ns cells flanking the lesion often maintain their positions and even close the gap in the bandlet within the germinal band, so that the only deficit in the resultant nerve cord is the clone that would otherwise have arisen from the ablated cell. In contrast, when an ns blast cell is killed, the cells behind the lesion invariably slip backward by one or more segments relative to the ipsilateral bandlets. Thus, these experiments suggest a positive affinity between ns cells that is lacking between nf cells.
Two different molecular approaches offer hope for eventually elucidating the mechanisms by which f and s blast cell fates are assigned in the grandparental stem cell lineages. One set of experiments suggests that differentially upregulated expression and activation of a CDC42 member of the small Rho GTPase family is associated with the ns cell fate (Zhang et al. 2009). Given the connections between CDC42 activity and filopodia formation (Nobes and Hall 1995), we speculate that the observed differences between nf and ns cells in CDC42 activity may also underlie the differences observed by Shain et al. (2000) in the behavior of nf and ns cells following the ablation of neighboring blast cells.
More recently, it has been discovered that some components of the WNT signaling pathway are differentially expressed between pre-mitotic f and s blast cells and within their early clones (Cho, Yoo et al. in preparation); the timing of these expression differences suggests that it is more likely to be a consequence of the initial specification of f and s fates, but perhaps characterizing the cell-type specific enhancers of genes will allow us to work our way upstream toward the initial specification event.
Studies in oligochaetes suggest that the mix of parental and grandparental teloblast lineages may be a general feature of clitellate annelid development (Goto et al. 1999a, 1999b; Arai et al. 2001; Storey 1989a). Apart from certain lineage mutations in Caenorhabditis elegans that exhibit a similar cell fate alternation (Chalfie et al. 1981), we are not aware of any other examples outside this group.