Our final case study involves the embryos of the flour beetle Tribolium castaneum. Tribolium embryos add labial, thoracic, and abdominal segments sequentially in early embryogenesis and thus fit within the category of short germband embryos. As with the other case studies, as segments are added the length and area of the posterior growth zone decreases. Thus, segment addition depletes the field of cells constituting the growth zone. The original growth zone needs to be -18% larger to equal the total area of all segments added during extension (Nakamoto et ah, 2015; note that these measurements did not account for the ingression of the mesoderm in the posterior region, which could potentially increase the percentage growth required by as much as 50%).
Does the Additional Growth Originate from Localized Cell Division in the Posterior?
Surprisingly, the answer to this question is not completely resolved. No evidence for higher rates of division in the posterior were found by counting DAPI-stained nuclei in fixed preparations (Sarrazin et al., 2012; Constantinou and Williams, unpublished observations). Similarly, when cultured early or mid-germband embryos are exposed to EdU for 60 minutes, cells in S-phase are found dispersed throughout the germband (Cepeda et ah, 2017). When EdU staining is quantified in these embryos, only early germband embryos show significantly more cells in S-phase in the growth zone relative to the already segmented trunk. The growth zone or SAZ is defined by these researchers as coincident with the posterior caudal expression domain. That the cells in the growth zone are not dividing more frequently than trunk cells is supported by the fact that small clones marked along the anteroposterior axis of the blastoderm do not show a significant difference in the number of cell divisions undergone by the end of embryo elongation (Nakamoto et ah, 2015). Oberhofer et ah (2014) come to a different conclusion and report that 38% of embryos injected with EdU (10-14 hours after egg laying; 32°C) and cultured for 3 hours show enhanced EdU uptake in the growth zone. The differences between these results are difficult to explain. Unfortunately, neither DAPI staining or EdU uptake, nor measuring division cycles in clones are ideal for quantifying proliferation. DAPI or Hoechst staining capture cells in M-phase, which spans a short segment of the cell cycle. The short time window of mitosis can be overcome with EdU labeling, where labeling records any cell that underwent some or all of S-phase during the duration of the exposure. However, the length of the cell cycle is unknown in Tribolium (and in most species), so interpreting how cells in S-phase translate into proliferation rates is difficult, and EdU incorporation cannot be taken as a straight metric of cell division per se. Another caveat is that in all these Tribolium studies, many, sometimes most, of the EdU or DAPI positive cells are either in the presumptive mesoderm or hindgut. Increased division in these two populations may correlate with their midline invagination or posterior internalization and may not directly contribute to segmental elongation. Without an analysis that distinguishes cells of different presumptive fates, it is difficult to confirm a posterior prevalence of cell division in the segmental anlage in these experiments. Despite these caveats, it is clear that some cell division occurs both in the growth zone and throughout the developing germband at all stages analyzed.
Tribolium Embryos Undergo Temporal Bursts of Cell Division A small pulse of cell division during early segmentation was found by counting DAPI-stained mitotic nuclei in cohorts of staged embryos (Sarrazin et al., 2012; Constantinou and Williams, unpublished observations). Interestingly, when the mitotic measures of staged animals are mapped to developmental stages (Figure 3.6), the relatively higher levels of growth zone mitoses are found when the thoracic segments are being added. Whether this is to supplement thoracic segmental anlage with cells in preparation for limb outgrowth or to pad the growth zone itself for the
FIGURE 3.6 Higher numbers of mitosis during embryogenesis correspond to the addition of thoracic segments in Tribolium. A. Fully extended Tribolium germband stained with the 4D9 antibody that recognizes both Engrailed and Invected proteins, with position and numbers of segments added in the previous 2 hours (after egg laying). B. The highest numbers of cells in mitosis occur just prior to thoracic segments are being added (En stripes 4-6). (Panel A from Nakamoto et al., 2015.) upcoming rapid production of abdominal segments is unknown in the absence of lineage tracing. These counts also do not distinguish between mesodermal versus ectodermal fates (as the aforementioned data).
Functional Approaches Confirm a Requirement for Cell Division in Elongation
Other approaches to determining the requirement for cell division in Tribolium elongation are consistent with a requirement for a mid-segmentation burst in cell division. To functionally test a requirement for cell division, Cepeda et al. (2017) injected embryos with a cocktail of hydroxyurea and aphidicolin, two inhibitors of DNA replication (and thereby cell division) that work by distinct mechanisms. Hydroxyurea inhibits ribonucleotide reductase, inhibiting DNA replication once nucleotide stores are depleted (Timson, 1975), while aphidicolin inhibits DNA polymerase alpha (Krokan et al., 1981). Embryos treated at an early stage when the first Engrailed/Invected (En/In) stripe is visible, only add an additional four to five En/In stripes, albeit at a slower pace than control embryos. Extension to a roughly similar stage was also seen in experiments that exposed embryos to microtubule inhibitors, expected to arrest the progress of cell division (Macaya et al., 2016). Arrest at this stage correlates well with the hypothesis that the pulse of cell division during thorax formation (En/In stripes 4-6) is required for the subsequent rapid addition of abdominal segments (Nakamoto et al., 2015). It would be interesting to extend the postblock examination of these embryos to see whether En/In stripes past the thorax (T6) were eventually added.
An alternative functional approach to testing a requirement for cell division was reported in the PhD thesis of Fu (2014). Fu injected adult females with dsRNA to knockdown either cyclin D or string function. Cyclin D (Gl-cyclin) initiates the transition from G1 to S-phase (reviewed in Massague, 2004), while string (cdc25) regulates the transition from G2 to M-phase (reviewed in Mailer, 1991) and regulates the three post-blastoderm cell cycles underlying the Drosophila mitotic domains (Edgar and O’Farrell, 1990). While the resulting embryonic phenotypes from these dsRNA injections were quite variable, in one distinct class of cuticular phenotypes the abdomen was truncated after the third thoracic segment. In addition to truncated abdomens, additional cuticular phenotypes included tiny heads, small malformed appendages, defects in dorsal closure, and asymmetric segment formation. While still preliminary, these experiments to functionally disrupt the cell cycle confirm a requirement for cell division in elongation, particularly of the abdominal segments.
Is the Tribolium Growth Zone Regionalized?
A subdivision of the Tribolium growth zone into distinct regions is apparent from the cycling domains of the pair-rule genes (Patel et al., 1994; Choe et al., 2006; Sarrazin et al., 2012; El-Sherif et al., 2012) as well as the recent mapping of Dichaete and odd paired expression showing separate stable domains within the posterior Caudal domain (Clark and Peel, 2018). But whether the anterior and posterior regions of the growth zone vary in cell cycle regulation has not been reported. Maybe cell cycle domains don’t exist or perhaps they haven’t been discovered. Regional synchronic- ity in cell cycles between individuals is easier to detect if collection times are short (e.g., 10 minute collections were used in Thamnocephalus hatchlings), at least short enough to not span more than the duration of a single-cell cycle. At the moment, cell cycle duration at any point in development is not known for any species other than Drosophila, and collecting large clutches of eggs in short amounts of time isn’t feasible in many species.
Computational Modeling Points to Cell Rearrangements As Key Motor behind Elongation
In another approach to determined a requirement for cell division in Tribolium elongation, Nakamoto et al. (2015), modeled the rapid phase of abdominal elongation of Tribolium embryos (18-20 h 30°C; Engrailed/Invected stripes 7-12). The initial growth zone in the model mirrored length and cell counts in an embryo just prior to abdominal segment addition. Cells in the model did not divide but underwent directional movements paralleling those well documented in an earlier stage embryo (Sarrazin et al., 2012). The simulated elongation mimicked elongation in the embryo. Although this simulation did not account for loss of the midline mesodermal cells, it is consistent with a model in which cell rearrangements rather than a highly proliferative posterior growth drive axial elongation in the abdomen after a pulse of cell division during thoracic segmentation.
Thus, like Thamnocephalus and Oncopeltus, cell division occurs throughout the germband during Tribolium embryo elongation. The size of the growth zone is not sufficient to account for the additional tissue added during segmentation, and cell division is required in both the growth zone and the trunk to elongate the embryo. However, there doesn’t appear to be a difference between the number of divisions a cell in either the trunk or growth zone undergoes during elongation. It is not known whether proliferation or DNA replication are restricted in the anterior growth zone, nor whether the cell cycle is regulated through control of S-phase. Novel to this species is the hypothesis that a temporal burst of cell division may fuel abdominal elongation.