Our second case study focuses on the patterns of cell division underlying elongation in the hemipteran insect Oncopeltus fasciatus. Oncopeltus is classified as an intermediate germband embryo (Liu and Kaufman, 2004) and during embryogen- esis develops the last nine abdominal segments sequentially from a small posterior region (Ben-David and Chipman, 2010; Birkan et al., 2011; Liu and Kaufman, 2004, 2005; Stahi and Chipman. 2016).
Posterior Cell Division, Coupled with Cell Division in the Already Segmented Region, Is Required for Normal Abdominal Segmentation Auman et al. (2017) quantified both changes in dimensions and the distribution of cell division throughout the posterior region during the development of the abdominal segments. They compared the cell division in already specified abdominal segments (expressing invected stripes), and the growth zone, defined as the region posterior to the last inverted stripe. The growth zone shrinks over the course of abdominal segmentation, consistent with the cells in this region contributing to segment addition. However, as in Thamnocephalus, its depletion is not sufficient to account for all the tissue of the newly added segments. Thus, the growth zone must continue to proliferate, or cells must arrive from elsewhere. There is also a significant degree of growth in regions of the embryo other than the growth zone. Consistent with this, Auman et al. (2017) found that the length of a new segment increases as segments continue to be added behind them.
The Anterior Growth Zone Is Characterized by a Lack of Cell Division To more precisely define when and where cell division is taking place, Auman et al. (2017) marked cells with pH3. This led to the discovery that differences in cell division effectively subdivide the growth zone itself into anterior and posterior domains: the anterior growth zone has very little cell division whereas the posterior has, on average, over twice as much (Figure 3.5). Auman et al. (2017) also found that no more than 5% of the cells in any region were undergoing division during the time points analyzed. The region with already specified segments showed the highest percentage of proliferating cells. EdU incorporation to determine any patterns of cells in S-phase was not done.
The Growth Zone Is Regionalized with Respect to Cell Cycling, and Those Regions Correspond to Changes in Segmentation Gene Expression Patterns Auman et al. (2017) also found a correlation with the region of reduced/no cell division and a cell’s transition from undifferentiated to differentiated. In Oncopeltus, the region of low cell division in the anterior of the growth zone is coincident with striped even-skipped (eve) expression versus the region of higher cell division in the posterior that is coincident with caudal expression and a broad eve expression domain (Auman et al., 2017). Again, the correlation of cell division and gene
FIGURE 3.5 Rates of mitosis are different in the anterior and posterior GZ in Oncopeltus and correspond to domains of eve expression. A. Multiple рНЗ-stained germbands, 46-54 hours after egg laying, merged into a heat map. The zone of low cell proliferation is indicated by dotted lines. B. In the anterior GZ (black line), eve is expressed in a dynamic striped pattern. In the posterior GZ, eve is expressed in a continuous domain (red line). (From Auman et al., 2017.) expression domains is consistent with the hypothesis of a posterior signaling region where cells are maintained in a multipotent state until they transit anteriorly, toward segment specification. However, as in Thamnocephalus, the broad posterior expression domains of these particular segmentation genes do not provide clues to understand w'hich 5% of the growth zone cells are given the signal to divide.
Evidence for Blastoderm Specification
This model is in contrast to an older study that showed that the cells in the posterior region might not be multipotent at early stages. Lawrence (1973) used x-ray irradiation to fate map the Oncopeltus embryo. When cells in the early embryo are irradiated, the irradiated cells and their progeny express a different pigment in the larval epidermis than nonirradiated cells and their progeny. Using this method, Lawrence conducted clonal analyses of abdominal segments. When individual cells were irradiated at the cleavage stage, clones contributed to multiple abdominal segments of the fifth-stage larva. However, when individual cells were irradiated at the blastoderm stage, each clone was restricted to a dorsal/ventral or left/right quadrant within a single abdominal segment of the fifth-stage larva. It should be noted though that the positions of the initial clones were not reported, making it difficult to draw precise inferences about the specificity of single-cell fates and minor contributions of a clone (Lawrence, 1973). The results are nonetheless intriguing and suggest segmental prepatterning precedes sequential segmentation of the posterior. Alternatively, it might simply indicate that the cells irradiated early undergo more cell division. If we abandon the idea that posterior cells are highly proliferative and recognize that a posterior cell in the late blastoderm only undergoes one to three division cycles and does not move around much, clones from single blastoderm cells may, in the majority of cases, ultimately lie within a single segment. However, the fact that all the clones respected segmental boundaries favors the interpretation that cells acquire some degree of segmental fate in the late blastoderm. Additional fate mapping studies would increase our understanding of when cells commit to their segmental fates.
Thus, as in Thamnocephalus, a model that emerges is that, at any point in time, the growth zone is shrinking but only a small fraction of cells in the posterior are dividing and additional tissue is required for elongation. The highest rates of division occur within the already segmented regions. The growth zone is regionalized with respect to cell cycling, and regionalized cell cycling corresponds with segmentation gene expression domains.