Studies on new arthropod models such as the beetle Tribolium castaneum are shifting our knowledge of embryonic patterning and morphogenesis beyond the Drosophila paradigm. In contrast to Drosophila, Tribolium embryos exhibit the short-germ type of development and become enveloped by extensive extra-embryonic membranes, the amnion and serosa. The genetic basis of these processes has been the focus of active research. Here, we complement genetic approaches with live fluorescence imaging of Tribolium embryos to make the link between gene function and morphogenetic cell behaviors during blastoderm formation and differentiation, germband condensation and elongation, and extra-embryonic development. We first show that transient labeling methods result in strong, homogeneous and persistent expression of fluorescent markers in Tribolium embryos, labeling the chromatin, membrane, cytoskeleton or combinations thereof. We then use co-injection of fluorescent markers with dsRNA for live imaging of embryos with disrupted caudal gene function caused by RNA interference. Using these approaches, we describe and compare cell and tissue dynamics in Tribolium embryos with wild-type and altered fate maps. We find that Tribolium germband condensation is effected by cell contraction and intercalation, with the latter being dependent on the anterior-posterior patterning system. We propose that germband condensation drives initiation of amnion folding, whereas expansion of the amniotic fold and closure of the amniotic cavity are likely driven by contraction of an actomyosin cable at the boundary between the amnion and serosa. Our methodology provides a comprehensive framework for testing quantitative models of patterning, growth and morphogenetic mechanisms in Tribolium and other arthropod species.

One of the key morphogenetic processes used during development is the controlled intercalation of cells between their neighbors. This process has been co-opted into a range of developmental events, and it also underlies an event that occurs in each major group of bilaterians: elongation of the embryo along the anterior-posterior axis [1]. In Drosophila, a novel component of this process was recently discovered by Paré et al., who showed that three Toll genes function together to drive cell intercalation during germband extension [2]. This finding raises the question of whether this role of Toll genes is an evolutionary novelty of flies or a general mechanism of embryonic morphogenesis. Here we show that the Toll gene function in axis elongation is, in fact, widely conserved among arthropods. First, we functionally demonstrate that two Toll genes are required for cell intercalation in the beetle Tribolium castaneum. We then show that these genes belong to a previously undescribed Toll subfamily and that members of this subfamily exhibit striped expression (as seen in Tribolium and previously reported in Drosophila [3-5]) in embryos of six other arthropod species spanning the entire phylum. Last, we show that two of these Toll genes are required for normal morphogenesis during anterior-posterior embryo elongation in the spider Parasteatoda tepidariorum, a member of the most basally branching arthropod lineage. From our findings, we hypothesize that Toll genes had a morphogenetic function in embryo elongation in the last common ancestor of all arthropods, which existed over 550 million years ago.

Patterning of the terminal regions of the Drosophila embryo is achieved by an exquisitely regulated signal that passes between the follicle cells of the ovary, and the developing embryo. This pathway, however, is missing or modified in other insects. Here we trace the evolution of this pathway by examining the origins and expression of its components. The three core components of this pathway: trunk, torso and torso-like have different evolutionary histories and have been assembled step-wise to form the canonical terminal patterning pathway of Drosophila and Tribolium. Trunk, torso and a gene unrelated to terminal patterning, prothoraciotrophic hormone (PTTH), show an intimately linked evolutionary history, with every holometabolous insect, except the honeybee, possessing both PTTH and torso genes. Trunk is more restricted in its phylogenetic distribution, present only in the Diptera and Tribolium and, surprisingly, in the chelicerate Ixodes scapularis, raising the possibility that trunk and torso evolved earlier than previously thought. In Drosophila torso-like restricts the activation of the terminal patterning pathway to the poles of the embryo. Torso-like evolved in the pan-crustacean lineage, but based on expression of components of the canonical terminal patterning system in the hemimetabolous insect Acyrthosiphon pisum and the holometabolous insect Apis mellifera, we find that the canonical terminal-patterning system is not active in these insects. We therefore propose that the ancestral function of torso-like is unrelated to terminal patterning and that torso-like has become co-opted into terminal patterning in the lineage leading to Coleoptera and Diptera. We also show that this co-option has not resulted in changes to the molecular function of this protein. Torso-like from the pea aphid, honeybee and Drosophila, despite being expressed in different patterns, are functionally equivalent. We propose that co-option of torso-like into restricting the activity of trunk and torso facilitated the final step in the evolution of this pathway; the capture of transcriptional control of target genes such as tailless and huckebein by this complex and novel patterning pathway.

Organizers play important roles during the embryonic development of many animals. The most famous example is the Spemann organizer that sets up embryonic axes in amphibian embryos. In spiders, a group of BMP secreting mesenchymal cells (the cumulus) functions as an organizer of the dorsoventral axis. Similar to experiments performed with the Spemann organizer, transplantation of the cumulus is able to induce a secondary axis in spiders. Despite the importance of this structure, it is unknown which factors are needed to activate cumulus specific gene expression. To address this question, we performed a transcriptomic analysis of early embryonic development in the spider Parasteatoda tepidariorum. Through this work, we found that the transcription factor Pt-Ets4 is needed for cumulus integrity, dorsoventral patterning and for the activation of Pt-hunchback and Pt-twist expression. Furthermore, ectopic expression of Pt-Ets4 is sufficient to induce cell delamination and migration by inducing a mesoderm-like cell fate.DOI: http://dx.doi.org/10.7554/eLife.27590.001

In Drosophila melanogaster, the germband forms directly on the egg surface and solely consists of embryonic tissue. In contrast, most insect embryos undergo a complicated set of tissue rearrangements to generate a condensed, multilayered germband. The ventral side of the germband is embryonic, while the dorsal side is thought to be an extraembryonic tissue called the amnion. While this tissue organisation has been accepted for decades and has been widely reported in insects, its accuracy has not been directly tested in any species. Using live cell tracking and differential cell labelling in the short germ beetle Tribolium castaneum, I show that most of the cells previously thought to be amnion actually give rise to large parts of the embryo. This process occurs via the dorsal-to-ventral flow of cells and contributes to germband extension (GBE). In addition, I show that true ‘amnion’ cells in Tribolium originate from a small region of the blastoderm. Together, my findings show that development in the short germ embryos of Tribolium and the long germ embryos of Drosophila is more similar than previously proposed. Dorsal-to-ventral cell flow also occurs in Drosophila during GBE, and I argue that the flow is driven by a conserved set of underlying morphogenetic events in both species. Furthermore, the revised Tribolium fate map that I present is far more similar to that of Drosophila than the classic Tribolium fate map. Lastly, my findings show that there is no qualitative difference between the tissue structure of the cellularised blastoderm and the short/intermediate germ germband. As such, the same tissue patterning mechanisms could function continuously throughout the cellularised blastoderm and germband stages, and easily shift between them over evolutionary time.