In animals, gastrulation is a period of time in early development during which a sphere of cells is reorganized into an embryo with cells arranged into three distinct layers (called germ layers). The process has changed substantially during the course of evolution and thus provides a great experimental system to investigate the genetic basis for differences in animal form and shape.

As an example, true flies use at least two different mechanisms to make the middle germ layer (the mesoderm). In both cases, the mesoderm is made up of cells that move inwards from the boundary of the outer germ layer. In midges and some other flies, these cells migrate individually, while in others including fruit flies, the cells move together as a sheet. Fruit flies and midges shared their last common ancestor 250 million years ago and although the genes that make the mesoderm in fruit flies are well understood, little is known about how the mesoderm forms in midges.

Urbansky, González Avalos et al. investigated which genes were responsible for the evolutionary transition between the different types of cell migration seen in flies. The experiments identified two genes – called folded gastrulation and t48 – that seem to operate as a simple switch between the two ways that mesoderm cells migrate. Both of these genes are active in fruit fly embryos and are required for the group migration of mesoderm cells. However, the genes do not appear to play a major role in the movement of individual mesoderm cells in midges.

Further experiments demonstrate that switching on these genes in midge embryos is sufficient to invoke group mesoderm cell migrations similar to those seen in fruit flies. These findings show that it is possible to identify genetic changes that underlie substantial differences in animal form and shape over hundred million of years. The ease by which Urbansky, González Avalos et al. were able to switch between the two types of mesoderm migration might explain why similar transitions in gastrulation have evolved repeatedly in animals. The next step is to test this hypothesis in other animals.

The evolution of shape and form is generally associated with changes of cell and tissue behavior during the course of development (Carroll et al., 2013). Such changes in morphogenesis can originate from modifications of molecular patterning, by adjusting the interpretation and integration of molecular patterning within and between cells, or by altering the translation of cellular decisions into changes of cytoskeleton and cell behavior, either alone or in combination (Davies, 2013). Understanding principal transitions in the evolution of morphogenesis thus requires the precise description of differences in cell and tissue behavior, the identification of coincident changes in gene activity, and a functional validation to support the link between molecular and morphogenetic divergence.

The evolutionary gain or loss of gene activity has been shown to be a major source of morphological innovation, in particular between closely related species and in systems where patterns of gene expression correspond to a direct phenotypic output like insect wing and body pigmentation (Arnoult et al., 2013; Gompel et al., 2005; Prud'homme et al., 2006; Wittkopp et al., 2002a, 2002b). Well-studied examples of morphogenetic evolution have correlated differences in tissue and cell behavior to modifications in gene expression (Cleves et al., 2014; Indjeian et al., 2016; Shapiro et al., 2004), occasionally supported by experimental evolution (Abzhanov et al., 2006, 2004), but rarely have analyzed morphogenesis in reference to a genetically well-understood developmental context (Rafiqi et al., 2012). It is currently not known whether major, macro-evolutionary changes of morphogenesis between distantly related species are driven through accumulative genetic tinkering, the recruitment or modification of pre-existing developmental modules in a switch-like fashion, or by the acquisition of entirely new gene function; neither is it understood how many or what kinds of modifications are required to instruct transitions between different modes of morphogenesis.

To address these questions, we have studied gastrulation in the fruit fly Drosophila melanogaster and the midge Chironomus riparius, two fly species that shared their last common ancestor about 250 million years ago (Wiegmann et al., 2011). Their evolutionary history is characterized by an evolutionary transition between two distinct modes of cell internalization during gastrulation, i.e. mesoderm invagination in D. melanogaster and mesoderm ingression in C. riparius (Figure 1). Similar transitions between distinct modes of cell internalization have been observed repeatedly and in either direction during the evolution of metazoan gastrulation (Leptin, 2005; Nielsen, 2012; Solnica-Krezel and Sepich, 2012). In insects, the extremely fast and coordinated invagination of mesoderm cells constitutes a derived morphogenetic process that evolved from stochastic migration of individual cells (Johannsen and Butt, 1941; Roth, 2004). Mesoderm invagination in flies has been studied extensively in D.melanogaster and thus provides an excellent reference system (Leptin, 2005); mesoderm ingression is less well understood but has been previously reported for C.riparius as well as other species representing the most basal branches of flies (Goltsev et al., 2007; Ritter, 1890).

Figure 1

Download asset Open asset

Mesoderm invagination in D.melanogaster builds on molecular patterning through the transcription factors Twist (Twi) and Snail (Sna) along the ventral midline of the blastoderm embryo. These transcription factors repress mitosis (Grosshans and Wieschaus, 2000; Seher and Leptin, 2000) and instruct specific shape changes in the cells of the mesoderm anlage (Costa et al., 1994; Kölsch et al., 2007; Leptin, 1991; Manning et al., 2013; Martin et al., 2009; Rauzi et al., 2015). Their positional information is relayed to the cytoskeleton by G-protein coupled receptor (GPCR) signaling through Folded gastrulation (Fog) as ligand, Mist as receptor, and Concertina (Cta) as associated Gα-unit. Upon binding of Fog to Mist, Cta is released and activates RhoGEF2, which is enriched at the cell apex of mesoderm cells through the transmembrane anchor T48. In turn, RhoGEF2 dependent activation of Rho1 invokes changes in cell shape through constriction of the apical actomyosin network (reviewed in Leptin, 2005; Manning and Rogers, 2014). We used this defined morphogenetic module as reference to identify the genetic innovations that distinguish mesoderm invagination in D.melanogaster from ingression in C.riparius. We then asked whether the evolutionary transition between ingression and invagination could be experimentally reproduced, and finally tested whether the origin of mesoderm invagination in flies has been associated with putative developmental advantages.

An initial comparison of mesoderm internalization in C.riparius and D.melanogaster indicated differences in the mode of cell internalization, which were consistent with an evolutionary transition from ingression to invagination (Figure 1A,B). To establish whether differences between ingression and invagination could be studied by a careful analysis of cell positions in fixed tissue, we used staining of F-actin and DNA as a proxy for the cytoskeletal cell outline and nuclei, respectively (Figure 1C,D). In both species, we found that the position of mesoderm cells during progression of mesoderm formation could be approximated by their nuclei (Figure 1E,E’). To establish equivalent time points of mesoderm internalization during the courses of gastrulation in C.riparius and D.melanogaster, we compared the timing of mesoderm internalization in both species relative to a conserved and independent process of fly gastrulation, i.e., germband extension (Anderson, 1966). We found that relative to the onset of germband extension, mesoderm internalization started and ended earlier in D.melanogaster than in C.riparius (Figure 1F). In transversal sections of C.riparius embryos that were about halfway through internalization, mesodermal cells clustered on top of a shallow groove along the ventral midline, while at a comparable stage in D.melanogaster they formed part of a deep ventral furrow (Figure 1A,B,F). Taken together, the net behavior of mesoderm cells in the two species appeared to be either ingressive (C.riparius) or invaginative (D.melanogaster).

Notably, when we used cell position in transversal sections to assess the coherence of mesoderm cell behavior in individual embryos (see Materials and methods), we found mesoderm cell behavior to vary substantially along the anterior-to-posterior axis (Figure 2A–B’’). These differences were more pronounced in C.riparius embryos than in D.melanogaster and indicated that the classification of cell internalization in fixed embryos as either ingressive or invaginative could not be based on qualitative measures of transversal sections alone. Thus, to account for variation in the mode of cell internalization within and between individuals, we established measures to quantitatively distinguish mesoderm ingression and invagination in whole embryos. Specifically, we quantified cell behavior during mesoderm internalization by using parameters that have been used to characterize this process in D.melanogaster. These included the total number of cells within the future mesoderm that underwent mitosis, the width and depth of the ventral furrow, the number of cells along the ventral midline that were internalized, a measurement of epithelial integrity, and the maximal depth at which an internalized cell could be observed (Costa et al., 1993; Kam et al., 1991; McMahon et al., 2008). Mitosis was detected by antibody staining against phosphorylated histone H3. All remaining features were quantified by approximating cell position based on the position of the nuclei relative to the computationally reconstructed egg and epithelial surfaces (Figure 2C,D, Figure 2—figure supplement 1). In the assessment of maximal cell internalization and furrow depth, the analysis based on nuclear cell position proved particularly beneficial as it was independent of cell orientation and avoided discrimination against distinct modes of mesoderm internalization. All features were evaluated for nine transversal sections of a ventral window for each embryo (see Materials and methods, Figure 2—figure supplement 1). While mitosis was rarely detected in either species during mesoderm internalization, wild-type embryos of C.riparius and D.melanogaster exhibited significant divergence in all other parameters (Figure 2E).

Figure 2 with 1 supplement see all

Download asset Open asset

To address the extent to which each individual parameter contributed to the global differences in C.riparius and D.melanogaster mesoderm formation and detect biases in the analysis, we performed a principal component analysis of all the sections and parameters. We found three principal parameters that, regardless of intra-species variation in space and time, permitted the best discrimination between ingression and invagination: cell internalization, the depth of the ventral furrow and epithelial integrity (Figure 2F–F’’’). Accordingly, invagination in D.melanogaster was characterized by high tissue integrity, a deep ventral furrow, and a high number of internalized cells, while ingression in C.riparius was characterized by low tissue integrity, a shallow ventral furrow, and a lower number of internalized cells. The number of internalized cells as well as the depth of the ventral furrow may be in part depended on the size of the mesoderm anlage and thus contain a species-specific component, whereas integrity was independent of the size of the mesoderm anlage. Collectively, the identified features made it possible to distinguish between ingressive and invaginative mesoderm formation in C.riparius and D.melanogaster precisely and quantitatively, despite variations within and between individuals.

To identify the genetic basis for these differences, we used the established genetic network of D.melanogaster mesoderm internalization as a reference (Figure 3A) and asked how the expression and function of orthologous genes differed in C.riparius. In D.melanogaster, mesoderm internalization is orchestrated by two zygotic transcription factors, Twist (Twi) and Snail (Sna) (Leptin and Grunewald, 1990). Both genes are expressed along the ventral midline of the embryo (Figure 3B–C’), where they instruct zygotic expression of fog (Figure 3D,D’), t48 (Figure 3E,E’), and mist (Figure 3F,F’), while cta and RhoGEF2 are expressed ubiquitously (Figure 3G–H’) (Costa et al., 1994; Kölsch et al., 2007; Manning et al., 2013). Orthologues of twi and sna, t48, and RhoGEF2, along with genes encoding the GPCR ligand Fog, the receptor Mist, and the Gα subunit Cta, could be identified in C.riparius and in major families of winged insects where genome sequence data is available (Pterygota, Figure 3—figure supplement 1). The universal conservation of this gene set demonstrates its evolutionary importance and suggests that it originated over 400 million years ago (Misof et al., 2014).

Figure 3 with 2 supplements see all

Download asset Open asset

To assess how members of this conserved set of genes contributed to gastrulation in C.riparius, we analyzed their expression (Figure 3I–P’) and function (Figure 4). We found that Cri-twi and Cri-sna were both expressed along the ventral midline of the blastoderm embryo (Figure 3I–J’). The knockdown of Cri-twi by RNAi resulted in embryos that lacked a visible ventral groove. Despite the knockdown of Cri-twi, nuclei of internalized cells could still be observed within the ventral-most 10–15% of the embryonic circumference (Figure 4A,B), which corresponded to a slightly narrower domain than that of wildtype Cri-twist expression. The knockdown of Cri-sna produced an abnormally thin ventral epithelium with spherical nuclei and a failure of proper mesoderm internalization (Figure 4C). Similar expression and phenotypes have been reported in experimental manipulations of twi and sna in D.melanogaster (Leptin and Grunewald, 1990), indicating that mesoderm cells receive and require essentially the same zygotic patterning information in C.riparius and D.melanogaster. Expression of Cri-mist along the ventral midline and in the domain of Cri-twi expression (Figure 3N,N’), as well as ubiquitous expression of Cri-cta (Figure 3O,O’) and Cri-RhoGEF2 (Figure 3P,P’) was consistent with a RhoGEF2-mediated machinery of cytoskeletal regulation like in D.melanogaster.

Figure 4

Download asset Open asset

Analysis of fog and t48 gene expression patterns yielded significant differences between the two species (compare Figure 3D–E’ with Figure 3K–M’). In D.melanogaster, both molecules are initially subject to maternal loading (Figure 3—figure supplement 2) (Costa et al., 1994; Strutt and White, 1994; Zusman and Wieschaus, 1985) and are later expressed zygotically along the ventral midline of blastoderm embryos in a Twist-dependent manner (Costa et al., 1994; Kölsch et al., 2007). In C.riparius, Cri-fog1 and Cri-t48 transcripts were not detected until mesoderm internalization had been completed (Figure 3K,K’,M,M’, Figure 3—figure supplement 2). A second fog orthologue, Cri-fog2, was expressed prior to mesoderm internalization in the central third of the embryo, in a domain broader than that seen for the expression of Cri-twi (Figure 3L,L’). This suggested that the GPCR ligand it encodes and, by extension, its receptor Cri-Mist, may not play the same prominent role in C.riparius mesoderm formation as they do in D.melanogaster. This interpretation was consistent with a visual inspection of embryos after RNAi knockdown of either gene. Following the quantitative analysis, we found that the median cell internalization and furrow depth differed in Cri-mist RNAi embryos from water injected embryos (Figure 4D). In Cri-fog2 RNAi embryos, the difference was smaller and significant only for an increase in the furrow and maximal cell depth (Figure 4E). All effects after Cri-mist and Cri-fog2 knockdown were substantially smaller than in Cri-twi and Cri-sna RNAi embryos and close to the sensitivity of our method (<10% of nuclear size), and we cannot exclude that either Cri-fog or Cri-mist do not contribute to mesoderm formation.

Based on the observed differences in fog and t48 expression in D.melanogaster and C.riparius, we wondered whether the lack of substantial pre-gastrulation activity of both genes in C.riparius could be functionally linked to differences in the modes of mesoderm internalization between the two species. Specifically, we asked whether mesoderm invagination might be provoked in C.riparius by emulating Drosophila-like gene expression of fog and t48. We tested this hypothesis by injecting in vitro transcribed mRNA of the D.melanogaster genes fog and t48 into C.riparius embryos at the early blastoderm stage. In response to this unspecific and early ubiquitous expression of fog and t48, C.riparius embryos exhibited an invagination of the mesoderm. This invagination was seen in several ways. First, we observed the formation of a distinct continuous ventral furrow, qualitatively comparable to that of D.melanogaster (Figure 5A,B). Then in particular, the induced invagination in C.riparius embryos was characterized by increased epithelial integrity and an increased furrow depth (Figure 5B).

Figure 5 with 1 supplement see all

Download asset Open asset

To test whether the induction of mesoderm invagination in C.riparius required the activity of the D.melanogaster-specific protein sequence of fog and t48, we repeated the experiments with Cri-fog and Cri-t48. Following early ubiquitous expression of Cri-fog and Cri-t48, C.riparius embryos responded with an invagination of the mesoderm which resembled that seen with the D.melanogaster forms of fog and t48 (Figure 5C), suggesting that a gain of early ubiquitous expression of one or both of these molecules was sufficient for the induction of mesoderm invagination. To test whether both fog and t48 had been necessary for the evolution of mesoderm invagination, we tested whether a single factor alone was sufficient to induce invagination in C.riparius. Upon early ubiquitous expression of either Cri-fog or Cri-t48 alone, C.riparius embryos displayed induced furrows overall, with increased epithelial integrity and deeper internalization very similar to the features observed in Cri-fog and Cri-t48 double-injected embryos (Figure 5D). Differences were noted in the behavior of cells within individual embryos: these appeared less variable in embryos expressing Cri-fog than in embryos expressing Cri-t48 (Figure 5—figure supplement 1). Our results indicate that fog and t48 operate differently in transmitting instructions to the cytoskeleton, while at the same time demonstrating that gaining early expression of either gene alone was sufficient to invoke a change in the mode of mesoderm internalization. This interpretation was further supported by converse experiments in D.melanogaster. While the loss of either fog or t48 activity in the mesoderm alone affects ventral furrow formation, it does not block it (Costa et al., 1994; Kölsch et al., 2007), but the combined knockdown of fog and t48 functions in the ventral blastoderm abolished ventral furrow formation and resulted in ingression-like mesoderm cell behavior reminiscent of C.riparius wildtype development (Figure 5E).

In D.melanogaster, the expression of fog and t48 both contribute to coordinated mesoderm invagination (Figure 3A) (Manning and Rogers, 2014). Our observations in C.riparius suggested that expression of either Cri-fog or Cri-t48 was already sufficient for a morphogenetic response that appeared indistinguishable from the expression of both genes together. This seemingly redundant control of mesoderm invagination in D.melanogaster prompted us to speculate that features beyond cell internalization were associated with a tissue-wide invagination. In comparison with individual cells, a coherent epithelium provides increased mechanical stability (Lecuit and Lenne, 2007). In the context of mesoderm internalization, this stability of a coordinated invagination could provide robustness against physical perturbations that are widely present during fly gastrulation as the germband expands and the midgut invaginates (Costa et al., 1993). To test whether variation within C.riparius gastrulation could be reduced by early ectopic expression of Cri-fog and Cri-t48, we collected and compared all time-independent measures of mesoderm internalization between control-injected and fog/t48 injected embryos (Figure 6A). Following the induced expression of Cri-fog and Cri-t48, we observed an overall decrease in the variability of phenotypes (Figure 6B). The decrease in the variability of the measured features was observed within embryos (Figure 6C) and may reflect the increase in cell coordination due to the induced change in the mode of gastrulation. Consistent with conceptual consideration of robustness (Félix and Barkoulas, 2015), we detected a decrease in variability also between embryos (Figure 6C), which indicated that C.riparius embryonic development became more robust against cellular mechanical perturbations.

Figure 6

Download asset Open asset

To test whether the expression of Cri-fog and Cri-t48 affected developmental timing, we counted how many cells had been internalized immediately after the onset of germband extension. Following the expression of Cri-fog and Cri-t48, we found a significant increase in the number of cells that were internalized (Figure 6D). An increased robustness against mechanical perturbation and developmental efficiency did not appear to trigger conflicts with subsequent embryonic muscle formation and enervation: after 48 hr, twitching and moving was observed in embryos injected with Cri-fog and Cri-t48 at a frequency (40%; n=500) similar to control-injected embryos (44%; n=531). Our results thus indicate that, beyond a simple coordination of cell behavior, mesoderm invagination induced by fog and t48 is associated with a measurable increase in the robustness and efficiency of gastrulation. Given the simplicity of the system and the extent of its fixation over long periods of evolutionary time, such traits could have provided an adaptive advantage in specific environments and contexts encountered by the predecessors of D.melanogaster after divergence from the last common ancestor shared with C.riparius.

Our analysis of C.riparius mesoderm morphogenesis has allowed us to associate two distinct modes of mesoderm internalization with specific differences in developmental gene expression. We could demonstrate that mesoderm ingression in C.riparius correlates at the genetic level with the absence of significant fog and t48 activity in the early embryo. In comparison with mesoderm invagination in D.melanogaster, these results suggested that an evolutionary gain of early fog and t48 activity may constitute a genetic switch to change the mode of fly gastrulation. We functionally tested this hypothesis by early ubiquitous expression of fog and t48 in C.riparius embryos, and we found that the activities of either gene alone or both genes together were sufficient to invoke a Drosophila-like invagination of the mesoderm in C.riparius embryos. Our data suggest that mesoderm invagination in C.riparius increases developmental robustness and efficiency. Additionally, invagination preserves the epithelial character of the mesoderm during internalization and thus decouples epithelial-to-mesenchymal transition in the mesoderm from the positional cell exchange during germband extension. As a result, the anterior-to-posterior order of mesoderm cells may be more effectively preserved throughout gastrulation. During the course of evolution, this could have facilitated the use of positional information of pre-gastrulation blastoderm patterning for post-gastrulation differentiation of the internalized mesoderm.

In the following, we are building on our results in C.riparius to assess the evolution of highly coordinated mesoderm invagination as described for D.melanogaster. Unspecific injection of fog and t48 mRNA, which at best corresponds to a strong maternal loading, had a specific effect on C.riparius mesoderm internalization instead of leading to pleiotropic morphogenetic defects (Figure 5). This observation was unexpected and surprising. In order for C.riparius embryos to respond to the early ubiquitous expression of fog or t48 in a spatially restricted manner along the ventral midline, the activity of either gene was apparently realized differently in different regions of the C.riparius blastoderm embryo. A strong and exclusively ventral response to fog and t48 activity in C.riparius could explain such a local change in the mode of morphogenesis in spite of ubiquitous gene expression. Recent work has shown that ancestral modes of gastrulation can be built upon a gene repertoire that is in part distinct from the network known in D.melanogaster (Pers et al., 2016; Stappert et al., 2016), and the local response to ubiquitous fog and t48 activity in C.riparius embryos may thus be elicited by an unknown fog/t48 sensitive pathway under control of twist and/or snail. Alternatively, ectopic fog and t48 in C.riparius act through a conserved genetic cascade present also in D.melanogaster. In this case, the activity of at least one component of this pathway downstream of fog and t48 would have to be locally restricted to the ventral blastoderm. A possible candidate for such a conserved component with spatially restricted activity downstream of fog and t48 is the G-protein coupled receptor Mist. In embryos of both species, C.riparius and D.melanogaster, mist is expressed in a narrow domain along the ventral midline (Figure 3F,F’,N,N’).

Mist transduces signaling of the extracellular ligand Fog to the intracellular receptor-coupled G-protein Gα_12/13 Cta, which during D.melanogaster mesoderm invagination results in apical RhoGEF2 activation. RhoGEF2 is intracellularly enriched at the apical cell membrane by its transmembrane anchor T48, and apical RhoGEF2 activation subsequently results in constriction of the apical actomyosin network and thus coordinated cell shape changes that lead to ventral invagination (Kerridge et al., 2016; Manning et al., 2013; Parks and Wieschaus, 1991; Dawes-Hoang et al., 2005; Morize et al., 1998; Barrett et al., 1997; Häcker and Perrimon, 1998; Fox and Peifer, 2007; Kölsch et al., 2007). In blastoderm embryos of D.melanogaster, fog and mist are expressed in partially overlapping domains, and tissue invagination is observed only where the expression of both genes coincides (Manning et al., 2013). Similarly, the activity of ectopic ubiquitous fog and t48 expression in C.riparius appears to be limited to a ventral domain that overlaps with endogenous Cri-mist expression.

Based on the expression of Cri-cta and Cri-mist and the induction of mesoderm invagination through ectopic fog and t48, we speculate that mesoderm ingression as an ancestral mode of cell internalization is based on comparatively low-level RhoGEF2 activity (Figure 7A). This low level of RhoGEF2 activity may be invoked by scattered expression of the ligand Fog, or, as has been suggested before, by low-frequency self-activation of Mist (Manning and Rogers, 2014). Our results in wildtype C.riparius embryos suggest that under such conditions, cells undergo an early epithelial-to-mesenchymal transition and leave the blastoderm stochastically due to individual rather than collective cell shape changes. This cell behavior changes under conditions that increase RhoGEF2 activity at the apex of mesoderm cells, which may be achieved in two ways: either non-cell-autonomously, by increasing GPCR signaling strength through elevated levels of the secreted ligand Fog, or cell-autonomously, by improving GPCR signaling efficiency through the apically enriched membrane anchor T48 (Kölsch et al., 2007). It is tempting to speculate that nature used similar changes during fly evolution (Figure 7B), which ultimately led to the highly efficient GPCR signaling in the ventral blastoderm of D.melanogaster (Figure 7C).

Figure 7

Download asset Open asset

Although C.riparius and D.melanogaster diverged from a last common ancestor about 250 million years ago (Wiegmann et al., 2011), our functional analysis of mesoderm internalization in both species suggests that surprisingly simple genetic changes are sufficient to reversibly switch between cell ingression and tissue invagination. Without an overt genetic directionality, mesoderm formation in flies has thus evolved either from a Drosophila-like ancestor that essentially lost maternal as well as zygotic fog and t48 expression in C.riparius, or from a Chironomus-like ancestor that gained early embryonic fog and t48 expression in successive steps leading to the developmental network described for D.melanogaster. We cannot exclude that C.riparius evolved from a Drosophila-like ancestor by a secondary loss of early fog and t48 expression, and recent lineage-specific gene duplications, resulting, e.g., in two fog copies or pangolin (Klomp et al., 2015), suggest that C.riparius may have diverged significantly from the last common ancestor of flies. Overall however, we favor the hypothesis that C.riparius rather than D.melanogaster is more reminiscent of ancestral mesoderm internalization in flies. Supporting this view, mesoderm ingression has been reported for an independent family at the base of the insect order of flies (Diptera), A.gambiae (Goltsev et al., 2007), in the beetle Tribolium castaneum (Handel et al., 2005), and reflects the general view that mesoderm formation in insects evolved from an ancestor with an ingression-like mode of cell internalization (Roth, 2004).

Experimental induction of mesoderm invagination by early fog or t48 expression in C.riparius embryos suggests that the evolutionary origin of the D.melanogaster gene network is linked to the gain of a ubiquitous maternal and a ventral zygotic enhancer for fog and t48. Our results in C.riparius suggest that high levels of maternal-like ubiquitous expression of fog and t48 are sufficient to invoke mesoderm invagination. Consistent with a critical role of maternal fog and t48 in the evolutionary origin of mesoderm invagination, both genes are also expressed ubiquitously in pre-blastoderm embryos of D.melanogaster (Costa et al., 1994; Strutt and White, 1994) (Figure 3—figure supplement 2). However, so far only a weak maternal contribution has been functionally demonstrated for fog (Zusman and Wieschaus, 1985), and the main function of both genes in D.melanogaster mesoderm invagination is derived from their zygotic, twist-dependent ventral expression (Costa et al., 1994; Kölsch et al., 2007). Thus, mesoderm invagination in D.melanogaster originated by the origin of highly specific Twist-dependent enhancers in fog or t48 and later the gain of maternal expression in both genes, or vice versa. We favor the idea that the origin of mesoderm invagination was built on the gain of a maternal enhancer, which may evolve more likely de novo from random sequence turnover.

fog and t48 may have acquired their role in coordinating cell behavior during epithelial morphogenesis coincedently with the evolution of fly gastrulation. However, in D.melanogaster, the functions of fog, mist, and RhoGEF2 do not appear to be limited to gastrulation; the genes have been additionally implicated in gastrulation-like folding of epithelia during wing, leg, and salivary gland morphogenesis (Manning et al., 2013; Nikolaidou and Barrett, 2004; Ratnaparkhi and Zinn, 2007), and the presence of fog and t48 orthologues in insect genomes overlaps with the phylogenetic occurrence of winged appendages and salivary glands (Krenn and Aspöck, 2012) (Figure 3—figure supplement 1). We therefore favor a model in which fog and t48 have played a role in epithelial morphogenesis, at least during late stages of development, long before the origin of mesoderm invagination. Both genes were then subsequently recruited for a function in modern, Drosophila-like gastrulation with a delayed epithelial-to-mesenchymal transition. Such recruitments may occur independently and repeatedly, and it will be informative to study a possible role of fog and t48 in hymenopterans. Species in this insect order also internalize their mesoderm in a coordinated way, but as a stiff, coherent plate along the ventral midline (Fleig and Sander, 1988; Roth, 2004; Sauer, 1954).

Recruiting genes with roles in later developmental stages into the maternal germline may provide the missing link between the divergence seen in late phases of a species' development with diversity that appears earlier – often referred to as a developmental 'hour glass' (Duboule, 1994; Raff, 1996; Kalinka and Tomancak, 2012). This provides a simple, and possibly non-gradual, evolutionary path toward altering early embryonic stages prior to the onset of zygotic transcription, and it may reflect a more general mechanism that facilitated the diversification of early embryonic development.

1. 1. Abzhanov A
   2. Kuo WP
   3. Hartmann C
   4. Grant BR
   5. Grant PR
   6. Tabin CJ

   (2006) The calmodulin pathway and evolution of elongated beak morphology in Darwin's finches

   Nature 442:563–567.

   https://doi.org/10.1038/nature04843

   • Google Scholar
2. 1. Abzhanov A
   2. Protas M
   3. Grant BR
   4. Grant PR
   5. Tabin CJ

   (2004) Bmp4 and morphological variation of beaks in Darwin's finches

   Science 305:1462–1465.

   https://doi.org/10.1126/science.1098095

   • Google Scholar
3. 1. Altschul SF
   2. Madden TL
   3. Schäffer AA
   4. Zhang J
   5. Zhang Z
   6. Miller W
   7. Lipman DJ

   (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs

   Nucleic Acids Research 25:3389–3402.

   https://doi.org/10.1093/nar/25.17.3389

   • Google Scholar
4. 1. Amaya E
   2. Musci TJ
   3. Kirschner MW

   (1991) Expression of a dominant negative mutant of the FGF receptor disrupts mesoderm formation in Xenopus embryos

   Cell 66:257–270.

   https://doi.org/10.1016/0092-8674(91)90616-7

   • Google Scholar
5. 1. Anderson DT

   (1966) The comparative embryology of the diptera

   Annual Review of Entomology 11:23–46.

   https://doi.org/10.1146/annurev.en.11.010166.000323

   • Google Scholar
6. 1. Arnoult L
   2. Su KF
   3. Manoel D
   4. Minervino C
   5. Magriña J
   6. Gompel N
   7. Prud'homme B

   (2013) Emergence and diversification of fly pigmentation through evolution of a gene regulatory module

   Science 339:1423–1426.

   https://doi.org/10.1126/science.1233749

   • Google Scholar
7. 1. Ashraf SI
   2. Hu X
   3. Roote J
   4. Ip YT

   (1999) The mesoderm determinant snail collaborates with related zinc-finger proteins to control Drosophila neurogenesis

   The EMBO Journal 18:6426–6438.

   https://doi.org/10.1093/emboj/18.22.6426

   • Google Scholar
8. 1. Attrill H
   2. Falls K
   3. Goodman JL
   4. Millburn GH
   5. Antonazzo G
   6. Rey AJ
   7. Marygold SJ
   8. Consortium F
   9. FlyBase Consortium

   (2016) FlyBase: establishing a Gene Group resource for Drosophila melanogaster

   Nucleic Acids Research 44:D786–792.

   https://doi.org/10.1093/nar/gkv1046

   • Google Scholar
9. 1. Barrett K
   2. Leptin M
   3. Settleman J

   (1997) The Rho GTPase and a putative RhoGEF mediate a signaling pathway for the cell shape changes in Drosophila gastrulation

   Cell 91:905–915.

   https://doi.org/10.1016/S0092-8674(00)80482-1

   • Google Scholar
10. 1. Camacho C
    2. Coulouris G
    3. Avagyan V
    4. Ma N
    5. Papadopoulos J
    6. Bealer K
    7. Madden TL

    (2009) BLAST+: architecture and applications

    BMC Bioinformatics 10:421.

    https://doi.org/10.1186/1471-2105-10-421

    • Google Scholar
11. 1. Caroti F
    2. Urbansky S
    3. Wosch M
    4. Lemke S

    (2015) Germ line transformation and in vivo labeling of nuclei in Diptera: report on Megaselia abdita (Phoridae) and Chironomus riparius (Chironomidae)

    Development Genes and Evolution 225:179–186.

    https://doi.org/10.1007/s00427-015-0504-5

    • Google Scholar
12. Book

    1. Carroll SB
    2. Grenier JK
    3. Weatherbee SD

    (2013)

    From DNA to Diversity ((2nd ed))

    Oxford: Wiley-Blackwell.

    • Google Scholar
13. 1. Cleves PA
    2. Ellis NA
    3. Jimenez MT
    4. Nunez SM
    5. Schluter D
    6. Kingsley DM
    7. Miller CT

    (2014) Evolved tooth gain in sticklebacks is associated with a cis-regulatory allele of Bmp6

    PNAS 111:13912–13917.

    https://doi.org/10.1073/pnas.1407567111

    • Google Scholar
14. Book

    1. Costa M
    2. Sweeton D
    3. Wieschaus E

    (1993)

    Gastrulation in Drosophila: cellular mechanisms of morphogenetic movements

    In: Bate M, Martinez Arias A, editors. The Development of Drosophila Melanogaster. New York: Cold Spring Harbor Laboratory Press. pp. 425–465.

    • Google Scholar
15. 1. Costa M
    2. Wilson ET
    3. Wieschaus E

    (1994) A putative cell signal encoded by the folded gastrulation gene coordinates cell shape changes during Drosophila gastrulation

    Cell 76:1075–1089.

    https://doi.org/10.1016/0092-8674(94)90384-0

    • Google Scholar
16. Book

    1. Davies JA

    (2013)

    Mechanisms of Morphogenesis ((2nd ed))

    London/New York: Academic Press.

    • Google Scholar
17. 1. Dawes-Hoang RE
    2. Parmar KM
    3. Christiansen AE
    4. Phelps CB
    5. Brand AH
    6. Wieschaus EF

    (2005) folded gastrulation, cell shape change and the control of myosin localization

    Development 132:4165–4178.

    https://doi.org/10.1242/dev.01938

    • Google Scholar
18. 1. Duboule D

    (1994)

    Temporal colinearity and the phylotypic progression: a basis for the stability of a vertebrate Bauplan and the evolution of morphologies through heterochrony

    Development 1994:135–142.

    • Google Scholar
19. 1. Fleig R
    2. Sander K

    (1988)

    Honeybee morphogenesis: embryonic cell movements that shape the larval body

    Development 103:525–534.

    • Google Scholar
20. 1. Foe VE

    (1989) Mitotic domains reveal early commitment of cells in Drosophila embryos

    Trends in Genetics 107:1–22.

    https://doi.org/10.1016/0168-9525(89)90120-0

    • Google Scholar
21. 1. Fox DT
    2. Peifer M

    (2007) Abelson kinase (Abl) and RhoGEF2 regulate actin organization during cell constriction in Drosophila

    Development 134:567–578.

    https://doi.org/10.1242/dev.02748

    • Google Scholar
22. 1. Félix MA
    2. Barkoulas M

    (2015) Pervasive robustness in biological systems

    Nature Reviews Genetics 16:483–496.

    https://doi.org/10.1038/nrg3949

    • Google Scholar
23. 1. Goltsev Y
    2. Fuse N
    3. Frasch M
    4. Zinzen RP
    5. Lanzaro G
    6. Levine M

    (2007) Evolution of the dorsal-ventral patterning network in the mosquito, Anopheles gambiae

    Development 134:2415–2424.

    https://doi.org/10.1242/dev.02863

    • Google Scholar
24. 1. Gompel N
    2. Prud'homme B
    3. Wittkopp PJ
    4. Kassner VA
    5. Carroll SB

    (2005) Chance caught on the wing: cis-regulatory evolution and the origin of pigment patterns in Drosophila

    Nature 433:481–487.

    https://doi.org/10.1038/nature03235

    • Google Scholar
25. 1. Grosshans J
    2. Wieschaus E

    (2000) A genetic link between morphogenesis and cell division during formation of the ventral furrow in Drosophila

    Cell 101:523–531.

    https://doi.org/10.1016/S0092-8674(00)80862-4

    • Google Scholar
26. 1. Handel K
    2. Basal A
    3. Fan X
    4. Roth S

    (2005) Tribolium castaneum twist: gastrulation and mesoderm formation in a short-germ beetle

    Development Genes and Evolution 215:13–31.

    https://doi.org/10.1007/s00427-004-0446-9

    • Google Scholar
27. 1. Häcker U
    2. Perrimon N

    (1998) DRhoGEF2 encodes a member of the Dbl family of oncogenes and controls cell shape changes during gastrulation in Drosophila

    Genes & Development 12:274–284.

    https://doi.org/10.1101/gad.12.2.274

    • Google Scholar
28. 1. Indjeian VB
    2. Kingman GA
    3. Jones FC
    4. Guenther CA
    5. Grimwood J
    6. Schmutz J
    7. Myers RM
    8. Kingsley DM

    (2016) Evolving new skeletal traits by cis-regulatory changes in bone morphogenetic proteins

    Cell 164:45–56.

    https://doi.org/10.1016/j.cell.2015.12.007

    • Google Scholar
29. 1. Jiménez-Guri E
    2. Wotton KR
    3. Gavilán B
    4. Jaeger J

    (2014) A staging scheme for the development of the moth midge Clogmia albipunctata

    PLoS One 9:e84422.

    https://doi.org/10.1371/journal.pone.0084422

    • Google Scholar
30. Book

    1. Johannsen OA
    2. Butt FH

    (1941)

    Embryology of Insects and Myriapods

    New York/London: McGraw-Hill Book Company, Inc.

    • Google Scholar
31. 1. Kalinka AT
    2. Tomancak P

    (2012) The evolution of early animal embryos: conservation or divergence?

    Trends in Ecology & Evolution 27:385–393.

    https://doi.org/10.1016/j.tree.2012.03.007

    • Google Scholar
32. 1. Kam Z
    2. Minden JS
    3. Agard DA
    4. Sedat JW
    5. Leptin M

    (1991)

    Drosophila gastrulation: analysis of cell shape changes in living embryos by three-dimensional fluorescence microscopy

    Development 112:365–370.

    • Google Scholar
33. 1. Kerridge S
    2. Munjal A
    3. Philippe JM
    4. Jha A
    5. de las Bayonas AG
    6. Saurin AJ
    7. Lecuit T

    (2016) Modular activation of Rho1 by GPCR signalling imparts polarized myosin II activation during morphogenesis

    Nature Cell Biology 18:261–270.

    https://doi.org/10.1038/ncb3302

    • Google Scholar
34. 1. Klomp J
    2. Athy D
    3. Kwan CW
    4. Bloch NI
    5. Sandmann T
    6. Lemke S
    7. Schmidt-Ott U

    (2015) Embryo development. A cysteine-clamp gene drives embryo polarity in the midge Chironomus

    Science 348:1040–1042.

    https://doi.org/10.1126/science.aaa7105

    • Google Scholar
35. 1. Krenn HW
    2. Aspöck H

    (2012) Form, function and evolution of the mouthparts of blood-feeding Arthropoda

    Arthropod Structure & Development 41:101–118.

    https://doi.org/10.1016/j.asd.2011.12.001

    • Google Scholar
36. 1. Kölsch V
    2. Seher T
    3. Fernandez-Ballester GJ
    4. Serrano L
    5. Leptin M

    (2007) Control of Drosophila gastrulation by apical localization of adherens junctions and RhoGEF2

    Science 315:384–386.

    https://doi.org/10.1126/science.1134833

    • Google Scholar
37. 1. Lecuit T
    2. Lenne PF

    (2007) Cell surface mechanics and the control of cell shape, tissue patterns and morphogenesis

    Nature Reviews Molecular Cell Biology 8:633–644.

    https://doi.org/10.1038/nrm2222

    • Google Scholar
38. 1. Lemke S
    2. Schmidt-Ott U

    (2009) Evidence for a composite anterior determinant in the hover fly Episyrphus balteatus (Syrphidae), a cyclorrhaphan fly with an anterodorsal serosa anlage

    Development 136:117–127.

    https://doi.org/10.1242/dev.030270

    • Google Scholar
39. 1. Leptin M
    2. Grunewald B

    (1990)

    Cell shape changes during gastrulation in Drosophila

    Development 110:73–84.

    • Google Scholar
40. 1. Leptin M

    (1991) twist and snail as positive and negative regulators during Drosophila mesoderm development

    Genes & Development 5:1568–1576.

    https://doi.org/10.1101/gad.5.9.1568

    • Google Scholar
41. 1. Leptin M

    (2005) Gastrulation movements: the logic and the nuts and bolts

    Developmental Cell 8:305–320.

    https://doi.org/10.1016/j.devcel.2005.02.007

    • Google Scholar
42. 1. Manning AJ
    2. Peters KA
    3. Peifer M
    4. Rogers SL

    (2013) Regulation of epithelial morphogenesis by the G protein-coupled receptor mist and its ligand fog

    Science Signaling 6:ra98.

    https://doi.org/10.1126/scisignal.2004427

    • Google Scholar
43. 1. Manning AJ
    2. Rogers SL

    (2014) The Fog signaling pathway: insights into signaling in morphogenesis

    Developmental Biology 394:6–14.

    https://doi.org/10.1016/j.ydbio.2014.08.003

    • Google Scholar
44. 1. Martin AC
    2. Kaschube M
    3. Wieschaus EF

    (2009) Pulsed contractions of an actin-myosin network drive apical constriction

    Nature 457:495–499.

    https://doi.org/10.1038/nature07522

    • Google Scholar
45. 1. Mathew SJ
    2. Rembold M
    3. Leptin M

    (2011) Role for Traf4 in polarizing adherens junctions as a prerequisite for efficient cell shape changes

    Molecular and Cellular Biology 31:4978–4993.

    https://doi.org/10.1128/MCB.05542-11

    • Google Scholar
46. 1. McMahon A
    2. Supatto W
    3. Fraser SE
    4. Stathopoulos A

    (2008) Dynamic analyses of Drosophila gastrulation provide insights into collective cell migration

    Science 322:1546–1550.

    https://doi.org/10.1126/science.1167094

    • Google Scholar
47. 1. Misof B
    2. Liu S
    3. Meusemann K
    4. Peters RS
    5. Donath A
    6. Mayer C
    7. Frandsen PB
    8. Ware J
    9. Flouri T
    10. Beutel RG
    11. Niehuis O
    12. Petersen M
    13. Izquierdo-Carrasco F
    14. Wappler T
    15. Rust J
    16. Aberer AJ
    17. Aspöck U
    18. Aspöck H
    19. Bartel D
    20. Blanke A
    21. Berger S
    22. Böhm A
    23. Buckley TR
    24. Calcott B
    25. Chen J
    26. Friedrich F
    27. Fukui M
    28. Fujita M
    29. Greve C
    30. Grobe P
    31. Gu S
    32. Huang Y
    33. Jermiin LS
    34. Kawahara AY
    35. Krogmann L
    36. Kubiak M
    37. Lanfear R
    38. Letsch H
    39. Li Y
    40. Li Z
    41. Li J
    42. Lu H
    43. Machida R
    44. Mashimo Y
    45. Kapli P
    46. McKenna DD
    47. Meng G
    48. Nakagaki Y
    49. Navarrete-Heredia JL
    50. Ott M
    51. Ou Y
    52. Pass G
    53. Podsiadlowski L
    54. Pohl H
    55. von Reumont BM
    56. Schütte K
    57. Sekiya K
    58. Shimizu S
    59. Slipinski A
    60. Stamatakis A
    61. Song W
    62. Su X
    63. Szucsich NU
    64. Tan M
    65. Tan X
    66. Tang M
    67. Tang J
    68. Timelthaler G
    69. Tomizuka S
    70. Trautwein M
    71. Tong X
    72. Uchifune T
    73. Walzl MG
    74. Wiegmann BM
    75. Wilbrandt J
    76. Wipfler B
    77. Wong TK
    78. Wu Q
    79. Wu G
    80. Xie Y
    81. Yang S
    82. Yang Q
    83. Yeates DK
    84. Yoshizawa K
    85. Zhang Q
    86. Zhang R
    87. Zhang W
    88. Zhang Y
    89. Zhao J
    90. Zhou C
    91. Zhou L
    92. Ziesmann T
    93. Zou S
    94. Li Y
    95. Xu X
    96. Zhang Y
    97. Yang H
    98. Wang J
    99. Wang J
    100. Kjer KM
    101. Zhou X

    (2014) Phylogenomics resolves the timing and pattern of insect evolution

    Science 346:763–767.

    https://doi.org/10.1126/science.1257570

    • Google Scholar
48. 1. Morize P
    2. Christiansen AE
    3. Costa M
    4. Parks S
    5. Wieschaus E

    (1998)

    Hyperactivation of the folded gastrulation pathway induces specific cell shape changes

    Development 125:589–597.

    • Google Scholar
49. Book

    1. Nielsen C

    (2012)

    Animal Evolution

    Oxford: Oxford University Press.

    • Google Scholar
50. 1. Nikolaidou KK
    2. Barrett K

    (2004) A Rho GTPase signaling pathway is used reiteratively in epithelial folding and potentially selects the outcome of Rho activation

    Current Biology 14:1822–1826.

    https://doi.org/10.1016/j.cub.2004.09.080

    • Google Scholar
51. 1. Panfilio KA
    2. Roth S

    (2010) Epithelial reorganization events during late extraembryonic development in a hemimetabolous insect

    Developmental Biology 340:100–115.

    https://doi.org/10.1016/j.ydbio.2009.12.034

    • Google Scholar
52. 1. Parks S
    2. Wieschaus E

    (1991) The Drosophila gastrulation gene concertina encodes a G alpha-like protein

    Cell 64:447–458.

    https://doi.org/10.1016/0092-8674(91)90652-F

    • Google Scholar
53. 1. Pers D
    2. Buchta T
    3. Özüak O
    4. Wolff S
    5. Pietsch JM
    6. Memon MB
    7. Roth S
    8. Lynch JA

    (2016) Global analysis of dorsoventral patterning in the wasp Nasonia reveals extensive incorporation of novelty in a regulatory network

    BMC Biology 14:63.

    https://doi.org/10.1186/s12915-016-0285-y

    • Google Scholar
54. 1. Pritchard DK
    2. Schubiger G

    (1996) Activation of transcription in Drosophila embryos is a gradual process mediated by the nucleocytoplasmic ratio

    Genes & Development 10:1131–1142.

    https://doi.org/10.1101/gad.10.9.1131

    • Google Scholar
55. 1. Prud'homme B
    2. Gompel N
    3. Rokas A
    4. Kassner VA
    5. Williams TM
    6. Yeh SD
    7. True JR
    8. Carroll SB

    (2006) Repeated morphological evolution through cis-regulatory changes in a pleiotropic gene

    Nature 440:1050–1053.

    https://doi.org/10.1038/nature04597

    • Google Scholar
56. Book

    1. Raff RA

    (1996)

    The Shape of Life

    Chicago: University of Chicago Press.

    • Google Scholar
57. 1. Rafiqi AM
    2. Lemke S
    3. Schmidt-Ott U

    (2011) Megaselia abdita: fixing and devitellinizing embryos

    Cold Spring Harbor Protocols 2011:prot5602.

    https://doi.org/10.1101/pdb.prot5602

    • Google Scholar
58. 1. Rafiqi AM
    2. Park CH
    3. Kwan CW
    4. Lemke S
    5. Schmidt-Ott U

    (2012) BMP-dependent serosa and amnion specification in the scuttle fly Megaselia abdita

    Development 139:3373–3382.

    https://doi.org/10.1242/dev.083873

    • Google Scholar
59. 1. Ratnaparkhi A
    2. Zinn K

    (2007) The secreted cell signal Folded Gastrulation regulates glial morphogenesis and axon guidance in Drosophila

    Developmental Biology 308:158–168.

    https://doi.org/10.1016/j.ydbio.2007.05.016

    • Google Scholar
60. 1. Rauzi M
    2. Krzic U
    3. Saunders TE
    4. Krajnc M
    5. Ziherl P
    6. Hufnagel L
    7. Leptin M

    (2015) Embryo-scale tissue mechanics during Drosophila gastrulation movements

    Nature Communications 6:8677.

    https://doi.org/10.1038/ncomms9677

    • Google Scholar
61. 1. Rice P
    2. Longden I
    3. Bleasby A

    (2000) EMBOSS: the European Molecular Biology Open Software Suite

    Trends in Genetics 16:276–277.

    https://doi.org/10.1016/S0168-9525(00)02024-2

    • Google Scholar
62. 1. Ritter R

    (1890)

    Z Wiss Zool

    408–427, Die Entwicklung der Geschlechtsorgane und des Darmes bei Chironomus, Z Wiss Zool, 50.

    • Google Scholar
63. 1. Roth S

    (2004)

    Gastrulation

    Gastrulation in other insects, Gastrulation, New York, Cold Spring Harbor Laboratory Press.

    • Google Scholar
64. 1. Sauer E

    (1954) Keimblätterbildung und Differenzierungsleistungen in Isolierten Eiteilen der Honigbiene

    Wilhelm Roux' Archiv FÜR Entwicklungsmechanik Der Organismen 147:302–354.

    https://doi.org/10.1007/BF00575364

    • Google Scholar
65. 1. Seher TC
    2. Leptin M

    (2000) Tribbles, a cell-cycle brake that coordinates proliferation and morphogenesis during Drosophila gastrulation

    Current Biology 10:623–629.

    https://doi.org/10.1016/S0960-9822(00)00502-9

    • Google Scholar
66. 1. Seher TC
    2. Narasimha M
    3. Vogelsang E
    4. Leptin M

    (2007) Analysis and reconstitution of the genetic cascade controlling early mesoderm morphogenesis in the Drosophila embryo

    Mechanisms of Development 124:167–179.

    https://doi.org/10.1016/j.mod.2006.12.004

    • Google Scholar
67. 1. Shapiro MD
    2. Marks ME
    3. Peichel CL
    4. Blackman BK
    5. Nereng KS
    6. Jónsson B
    7. Schluter D
    8. Kingsley DM

    (2004) Genetic and developmental basis of evolutionary pelvic reduction in threespine sticklebacks

    Nature 428:717–723.

    https://doi.org/10.1038/nature02415

    • Google Scholar
68. 1. Solnica-Krezel L
    2. Sepich DS

    (2012) Gastrulation: making and shaping germ layers

    Annual Review of Cell and Developmental Biology 28:687–717.

    https://doi.org/10.1146/annurev-cellbio-092910-154043

    • Google Scholar
69. Book

    1. Sommer C
    2. Straehle C
    3. Kothe U
    4. Hamprecht FA

    (2011)

    Ilastik: Interactive Learning and Segmentation Toolkit

    IEEE.

    • Google Scholar
70. 1. Stappert D
    2. Frey N
    3. von Levetzow C
    4. Roth S

    (2016) Genome-wide identification of Tribolium dorsoventral patterning genes

    Development 143:2443–2454.

    https://doi.org/10.1242/dev.130641

    • Google Scholar
71. 1. Stauber M
    2. Taubert H
    3. Schmidt-Ott U

    (2000) Function of bicoid and hunchback homologs in the basal cyclorrhaphan fly Megaselia (Phoridae)

    PNAS 97:10844–10849.

    https://doi.org/10.1073/pnas.190095397

    • Google Scholar
72. 1. Strutt  DI
    2. White RA

    (1994) Characterisation of T48, a target of homeotic gene regulation in Drosophila embryogenesis

    Mechanisms of Development 46:27–39.

    https://doi.org/10.1016/0925-4773(94)90035-3

    • Google Scholar
73. 1. Waterhouse RM
    2. Tegenfeldt F
    3. Li J
    4. Zdobnov EM
    5. Kriventseva EV

    (2013) OrthoDB: a hierarchical catalog of animal, fungal and bacterial orthologs

    Nucleic Acids Research 41:D358–365.

    https://doi.org/10.1093/nar/gks1116

    • Google Scholar
74. 1. Wiegmann BM
    2. Trautwein MD
    3. Winkler IS
    4. Barr NB
    5. Kim JW
    6. Lambkin C
    7. Bertone MA
    8. Cassel BK
    9. Bayless KM
    10. Heimberg AM
    11. Wheeler BM
    12. Peterson KJ
    13. Pape T
    14. Sinclair BJ
    15. Skevington JH
    16. Blagoderov V
    17. Caravas J
    18. Kutty SN
    19. Schmidt-Ott U
    20. Kampmeier GE
    21. Thompson FC
    22. Grimaldi DA
    23. Beckenbach AT
    24. Courtney GW
    25. Friedrich M
    26. Meier R
    27. Yeates DK

    (2011) Episodic radiations in the fly tree of life

    PNAS 108:5690–5695.

    https://doi.org/10.1073/pnas.1012675108

    • Google Scholar
75. 1. Wittkopp PJ
    2. True JR
    3. Carroll SB

    (2002a)

    Reciprocal functions of the Drosophila yellow and ebony proteins in the development and evolution of pigment patterns

    Development 129:1849–1858.

    • Google Scholar
76. 1. Wittkopp PJ
    2. Vaccaro K
    3. Carroll SB

    (2002b) Evolution of yellow gene regulation and pigmentation in Drosophila

    Current Biology 12:1547–1556.

    https://doi.org/10.1016/S0960-9822(02)01113-2

    • Google Scholar
77. 1. Yin C
    2. Shen G
    3. Guo D
    4. Wang S
    5. Ma X
    6. Xiao H
    7. Liu J
    8. Zhang Z
    9. Liu Y
    10. Zhang Y
    11. Yu K
    12. Huang S
    13. Li F

    (2016) InsectBase: a resource for insect genomes and transcriptomes

    Nucleic Acids Research 44:D801–807.

    https://doi.org/10.1093/nar/gkv1204

    • Google Scholar
78. 1. Zusman SB
    2. Wieschaus EF

    (1985) Requirements for zygotic gene activity during gastrulation in Drosophila melanogaster

    Developmental Biology 111:359–371.

    https://doi.org/10.1016/0012-1606(85)90490-7

    • Google Scholar

Author details

1. Silvia Urbansky

   Centre for Organismal Studies Heidelberg, Heidelberg University, Heidelberg, Germany

   Contribution

   SU, Conception and design, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Contributed equally with

   Paula González Avalos

   Competing interests

   The authors declare that no competing interests exist.

2. Paula González Avalos

   Centre for Organismal Studies Heidelberg, Heidelberg University, Heidelberg, Germany

   Contribution

   PGA, Conception and design, Analysis and interpretation of data, Drafting or revising the article

   Contributed equally with

   Silvia Urbansky

   Competing interests

   The authors declare that no competing interests exist.

3. Maike Wosch

   Centre for Organismal Studies Heidelberg, Heidelberg University, Heidelberg, Germany

   Contribution

   MW, Acquisition of data, Contributed unpublished essential data or reagents

   Competing interests

   The authors declare that no competing interests exist.

4. Steffen Lemke

   Centre for Organismal Studies Heidelberg, Heidelberg University, Heidelberg, Germany

   Contribution

   SL, Conception and design, Analysis and interpretation of data, Drafting or revising the article

   For correspondence

   steffen.lemke@cos.uni-heidelberg.de

   Competing interests

   The authors declare that no competing interests exist.

   "This ORCID iD identifies the author of this article:" 0000-0001-5807-2865

• 2,548

  views

• 331

  downloads

• 23

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.