BIOLOGICAL OVERVIEW

Vein is a component of the epidermal growth factor receptor (EGF-R) pathway that is involved in signaling through the EGF receptor (EGF-R) during development of the ventrolateral epidermis and the wing. In contrast to Spitz, which must be processed to an active ligand, Vein is constitutively active. However, the activity of Vein is intrinsically weaker than that of Spitz. This is reflected by its reduced capacity to induce activated MAP kinase in cells and embryos, and the limited ability to induce ectopic expression of Egfr target genes (Schnepp, 1998 and Yarnitzky, 1998). Before discussing Vein in particular, a general word on EFG-R is in order.

The Drosophila Epidermal growth factor receptor is a ubiquitously expressed receptor tyrosine kinase. Activation of EGF-R signaling is achieved by spatially restricted extracellular factors such as ligands. For example, Drosophila Gurken is an oogenesis-specific EGF-like ligand involved in establishing egg polarity. grk transcripts are localized to the anterior-dorsal region of the oocyte and a local source of GRK is thought to activate EGF-R only in adjacent cells. Besides ligands, spatially localized cofactors can mediate the local activation of EGF-R. For example, Rhomboid and Star, which along with EGF-R are members of the spitz group responsible for structuring the ventrolateral epidermis of the embryo, modify the function of EGF-R to spatially restrict EGF-R signaling.

Vein is a candidate EGF-R ligand and is expressed in a spatially restricted manner. Vein has both an EGF domain and an immunoglobulin domain, and therefore is most closely related to the vertebrate neuregulins, which carry out their signaling function through vertebrate receptors that are related to EGF-R. The Ig-like domain of Vein, in other systems known to be involved in protein-protein interaction, may facilitate protein dimerization or interaction with other external proteins.

What role does VN play in ventrolateral signaling and in wing development? Unlike the EGF-R ligand Spitz, Vein is expressed in a localized pattern, thereby limiting Vein-EGF-R signaling to cells in the vein expression domain. In the embryo, vn and spi are expressed in overlapping domains; genetic data suggest they function together to achieve the level of EGF-R activation required for normal development of ventrolateral cells (Schnepp, 1996).

The situation is different in the wing. Vein is expressed primarily in developing interveins, while the genes that modulated Spitz/EGF-R signaling (rhomboid, Star and argos) are expressed in developing veins. EGF-R expression is down regulated in prevein cells, so that in pupal wings, EGF-R expression colocalizes with Vein in intervein regions. Therefore, it has been suggested that Spitz functions as the major ligand for activating EGF-R in vein territories, and that Vein is the major ligand for activating EGF-R in intervein regions. Spitz is broadly expressed in developing wing discs, but its active form, secreted SPI may be localized to the presumptive veins by the activity of the spitz-group members Star and Rhomboid, which are expressed in developing veins. This EGR-R signal may be activated in developing veins by SPI and may be limited to a short duration by two mechanisms: inhibition by the Spitz competitor, Argos, which is also restricted to presumptive veins, and down regulation of EGF-R transcript in the presumptive vein regions. This dampening of EGF-R signaling may be required for prevein cells to start their differentiation program (Simcox, 1996 and references).

That Vein acts to engender cell proliferation is made clear from studies of mutant vein clones. Where they appear, large vein clones reduce the size of intervein regions by an average of 28%. Effects on cell proliferation are not only locally restricted. vein mutant clones grow, but they also nonautonomously reduce the size of other compartments and of the wing as a whole. The larger the clones, the stronger the effects: a greater number of compartment boundaries and stretches of veins are affected. The wings carrying mutant clones for either vein or Rhomboid are smaller than those of control sibling flies, while the wings of double mutants are smaller still. In addition to their small size, the wings of double mutants lack all longitudinal and transverse veins. These synergistic effects indicate a close functional relationship between both vein and Rhomboid during proliferation and differentition (García-Bellido, 1994).

How do veins on the dorsal and ventral surfaces of a fly's wing line up in exact opposition to one another? The adult Drosophila wing consists of two wing surfaces, apposed by their basal membranes, which first come into contact at metamorphosis, following wing disc eversion. Contact is crucial. Veins normally appear in these surfaces in a dorsal-ventral symmetric pattern, but when contact between the two surfaces is prevented, the dorsal-ventral pattern of venation takes on a 'corrugated,' asymmetric appearance (vein cells are more compacted and more pigmented). Dorsal-ventral contact apposition was prevented during wing imaginal disc morphogenesis by implanting fragments of discs into metamorphosing hosts. In these implants, longitudinal veins differentiate in both surfaces, but exhibit wider than normal corrugation. These results and those of genetic mosaics for mutants that remove veins or cause ectopic veins, reveal mutual dorso-ventral induction/inhibition at work to modulate the final vein differentiation pattern and/or corrugation. While clones of mutants causing a lack of veins (the single mutant rhomboid, the double mutant rhomboid vein, and the triple mutant extramachrochaetaeAch rhomboid vein) are autonomous in both wing surfaces, the vein phenotype is partially rescued by wild type cells from the opposite surface. When apposed to a lack of vein differentiation in the dorsal surface, vein differentiation fails in distal vein territories, preferentially in the ventral surface. Conversely, extra differentiation as in plexus, extramachrochaetae, HS-rhomboid 27B, Notch and Delta mutants, causes extra veins in clones, not only in the wing surface of the clone, but also in the opposite wing surface. Non-autonomous effects are observed in the same wing surface, a phenomenon called 'connectivity'. Genetic mosiacs of plexus72 and emc HS-rho27B cause neighboring non-mutant cells on the same surface to differentiate extra veins, connecting them. Cis (planar) and trans (vertical) effects may be operationally related, inducing contacting cells to differentiate vein histotypes. Vein cells induce vein differentiation in neighbouring cells, either on the same surface by planar cell-cell communication, or on the opposite surface through signals along the basal membrane of the apposing epithelium. Thus, although the vein pattern is surface-autonomously generated, inhibitory (negative) and inductive (positive) signals take place between both dorsal and ventral wing surfaces in order to refine the final vein pattern with the corresponding dorso-ventral wing surfaces (Milan, 1997).

Inductive interactions between cells of distinct fates underlie the basis for morphogenesis and organogenesis across species. In the Drosophila embryo, somatic myotubes form specific interactions with their epidermal muscle attachment (EMA) cells. The establishment of these interactions is a first step toward further differentiation of the EMA cells into elongated tendon cells containing an organized array of microtubules and microfilaments. The molecular signal for terminal differentiation of tendon cells is the secreted Drosophila neuregulin-like growth factor Vein, produced by the myotubes. Although Vein mRNA is produced by all of the myotubes, Vein protein is secreted and accumulates specifically at the muscle-tendon cell junctional site. In loss-of-function vein mutant embryos, muscle-dependent differentiation of epidermal tendon cells, measured by the level of expression of specific markers (Delilah and beta1 tubulin) is blocked. When Vein is expressed in ectopic ectodermal cells, it induces the ectopic expression of these genes. These results favor the possibility that the Drosophila EGF receptor DER/Egfr expressed by the EMA cells functions as a receptor for Vein. Vein/Egfr binding activates the Ras pathway in the EMA cells leading to the transcription of the tendon-specific genes stripe, delilah, and beta1 tubulin. In Egfr1F26 mutant embryos lacking functional Egfr expression, the levels of Delilah and beta1 Tubulin are very low. The ability of ectopic Vein to induce the expression of Delilah and beta1 Tubulin depends on the presence of functional Egfrs. Activation of the Egfr signaling pathway by either ectopically secreted Spitz, or activated Ras, leads to the ectopic expression of Delilah. These results suggest that inductive interactions between myotubes and their epidermal muscle attachment cells are initiated by the binding of Vein, to the Egfr on the surface of EMA cells (Yarnitzky, 1997).

The tissue-specific regulation of Vn signaling was investigated by examining vn transcriptional control and Vn target gene activation in the embryo and the wing. The results show a complex temporal and spatial regulation of vn transcription involving multiple signaling pathways and tissue-specific activation of Vn target genes. In the embryo, vn is a target of Spi/Egfr signaling mediated by the ETS transcription factor PointedP1 (PntP1). This establishes a positive feedback loop in addition to the negative feedback loop involving Aos. The simultaneous production of Vn provides a mechanism for dampening Aos inhibition and thus fine-tunes signaling. In the larval wing pouch, vn is not a target of Spi/Egfr signaling but is expressed along the anterior-posterior boundary in response to Hedgehog (Hh) signaling. Repression by Wingless (Wg) signaling further refines the vn expression pattern by causing a discontinuity at the dorsal-ventral boundary. The potential for vn to activate Egfr target genes correlates with its roles in development: vn has a minor role in embryogenesis and does not induce Egfr target genes such as aos and pntP1 in the embryo. Conversely, vn has a major role in wing development and Vn/Egfr signaling is a potent inducer of Egfr target genes in the wing disk. Spi also has the potential to induce Egfr target genes in the wing disk. However, the ligands appear to evoke specific responses that result in different patterns of target gene expression. Other factors modulate the potential of Vn so that induction of Vn/Egfr target genes in the wing pouch is cell specific (Wessells, 1999).

Differences between Vn and Spi are apparent in the patterns of target gene induction resultant from the ectopic expression of Vn and soluble Spitz (sSpi) in the wing pouch. Effects have been noted for three Egfr target genes: aos, pntP1, and kekkon-1 (kek1). In each case, a different response to the ligands is seen. Both ligands induce ectopic aos expression but Vn does so in a broader domain than sSpi, however, neither induces aos in the L3/L4 intervein region. In the embryo, the Egfr target gene pntP1 mediates aos induction by sSpi. Likewise in the wing, ectopic sSpi induces pntP1 expression in cells that also expressed aos. However, following ectopic Vn no detectable change in pntP1 expression is seen using in situ hybridization and only very weak induction of pntP1-lacZ is seen in a domain that does not correspond fully with ectopic aos expression. This suggests either that another transcription factor mediates the induction of aos in response to Vn or that PntP1 is capable of inducing aos, even when changes in its own expression level are too low to be detected by current methods. Both Vn and sSpi induce ectopic expression of kek1, but predominantly in posterior cells rather than throughout the domain of their ectopic expression. Thus the action of both ligands appears limited by the presence or absence of some other factor in anterior cells. There is also a difference in the level of induction: Vn, which functions to induce kek1 expression in normal development, is a potent inducer of high levels of ectopic kek1 expression, whereas sSpi induces low levels of ectopic kek1 expression and appears to reduce expression of endogenous kek1 (Wessells, 1999).

cis-regulatory architecture of a short-range EGFR organizing center in the Drosophila melanogaster

This study characterized the establishment of an Epidermal Growth Factor Receptor (EGFR) organizing center (EOC) during leg development in Drosophila melanogaster. Initial EGFR activation occurs in the center of leg discs by expression of the EGFR ligand Vn and the EGFR ligand-processing protease Rho, each through single enhancers, vnE and rhoE, that integrate inputs from Wg, Dpp, Dll and Sp1. Deletion of vnE and rhoE eliminates vn and rho expression in the center of the leg imaginal discs, respectively. Animals with deletions of both vnE and rhoE (but not individually) show distal but not medial leg truncations, suggesting that the distal source of EGFR ligands acts at short-range to only specify distal-most fates, and that multiple additional 'ring' enhancers are responsible for medial fates. Further, based on the cis-regulatory logic of vnE and rhoE many additional leg enhancers were identified, suggesting that this logic is broadly used by many genes during Drosophila limb development (Newcomb, 2018).

The EGFR signaling pathway is widely used in animal development, and is frequently a target in human disease and developmental abnormalities. Yet despite its importance in animal biology, many questions remain about how this pathway functions. Among these questions is whether secreted ligands that activate this pathway can induce distinct cell fates in a concentration-dependent manner. This study tests this idea by specifically eliminating a single source of EGFR ligands from the center of the Drosophila leg imaginal disc, which fate maps to the distal-most region of the adult leg. One plausible scenario is that this single source of secreted EGFR ligands, which is referred to as the EOC, activates distinct gene expression responses at different distances from this source. Alternatively, eliminating ligands secreted from the EOC might only affect gene expression locally, close to or within the EOC. Taken together, the current data are most consistent with the second scenario. This conclusion is largely supported by the observations that CRM deletions that eliminate vn and rho expression from the EOC have mild developmental consequences, both in the L3 leg imaginal discs and adult legs. These phenotypes are significantly weaker than those generated when the entire EGFR pathway is compromised using a temperature sensitive allele of the EGFR receptor. The difference between these two phenotypes is most likely explained by removing only a single source of EGFR ligands in the enhancer deletion experiments versus affecting EGFR signaling throughout the leg disc in the Egfr^tsla experiments. This explanation is further supported by the observation that there are indeed additional CRMs, some of which were defined in this study, that drive EGFR ligand production in more medial ring-like patterns during the L3 stage (Newcomb, 2018).

One possible caveat to these conclusions is that there are a total of seven rho-like protease genes in the Drosophila genome that could, in principle, play a role in distal leg development. This study focused on rho and roughoid ru, based on previous results showing that triple rho ru vn clones generate severe leg truncations that phenocopy strong Egfr^tsla truncations. In addition, it is noted that if other rho family proteases were active in the EOC, leg truncations and patterning defects would not be expected in the leg discs of the rho^rhoE-Df vn^vnE-Df double mutant, because those proteases should be able to produce active Spi. These observations suggest that the remaining five rho-like protease genes play a minor (or no) role in leg development. However, this conclusion will ultimately benefit from further genetic and expression analysis of these additional rho-like genes (Newcomb, 2018).

An additional previous observation that contrasts with the suggestion that EOC activity has only a limited role in specifying distal leg fates is the partial rescue of the PD axis when only a small number of distal leg cells were wild type in legs containing large rho ru vn clones. However, it is noted that even in these 'rescued' legs, medial defects in PD patterning were apparent. It is also noteworthy that in these earlier experiments, only adult legs were examined. When the same experiment was repeated, but L3 discs were analyzed, it was found that rho ru vn clones generated phenotypes that were very similar to those produced by the double vnE rhoE enhancer deletions. Taken together, these observations suggest that timing must be considered in the interpretation of these experiments. When assayed at the late L3 stage, both enhancer deletion and rho ru vn clone experiments argue that EOC activity is limited to specifying only the most distal fates, marked by the expression of al and C15. Starting in mid L3, and perhaps continuing into pupal development, there are additional sources of EGFR ligands that, when compromised, can affect adult leg morphology. Nevertheless, at least at the L3 stage, these data suggest that EGFR ligands produced from the EOC have a limited and local role in specifying distal leg fates (Newcomb, 2018).

Integration of inputs from signaling pathways and organ selector genes at CRMs in order to execute distinct developmental programs is a recurrent theme during animal development. This study identified two leg EGFR ligand CRMs that integrate the inputs from the Wg and Dpp signaling pathways and the leg selector genes Dll and/or Sp1 in a manner that is very similar to a previously characterized leg enhancer DllLT. In addition, when the same regulatory logic was applied to the whole genome, a battery of leg enhancer elements was identified. Interestingly, each of these enhancers drives expression in a specific manner with slightly different timing despite the fact that many of the inputs are shared. It is conceivable that the different expression patterns directed by these enhancers are in part a consequence of additional inputs and/or the difference in the TF binding site grammar. In support of this idea, vnE and rhoE differ in the number of binding sites for many inputs and vnE requires Sp1 while rhoE does not. Both of these differences may contribute to the earlier onset of vnE expression compared to rhoE. The remaining enhancer elements identified in this study direct a plethora of PD-biased leg expression patterns -- ranging from ubiquitous, to central and 'ring' patterns (see Genome-wide analysis of combinatorial inputs of Dll, Sp1, Wg, and Dpp in leg discs), which likely integrate inputs in addition to the ones described here. Future studies of these CRMs would help reveal the complex network of regulation that orchestrates leg development in the fruit fly. Such detailed understanding of the cis-regulatory architecture of fly leg development would likely give insights into organogenesis and evolution in other animals as well (Newcomb, 2018).

The EGFR signaling pathway has tremendous oncogenic potential and understanding the various mechanisms regulating its activation is not only interesting from the point of view of animal development but also has important practical implications. While the core components of the EGFR pathway have been thoroughly studied because of their potent tumorigenic capability in humans, little is known about the transcriptional regulation of EGFR ligands that bind the receptor and activate the pathway. The reiterative use of EGFR signaling in many developmental processes implies that different cis-regulatory elements are likely utilized by each EGFR ligand in different organs and tissues in order to correctly read the diverse cues in any specific developmental context. It is conceivable that genomic variation in EGFR pathway CRMs might lead to a predisposition to different types of EGFR-dependent tumors in humans, since such CRMs may respond to potent growth-promoting signaling pathways, such as Wnt and BMP (Newcomb, 2018).

This study has characterized in detail two Drosophila EGFR CRMs, vnE and rhoE, and showed how they integrate the cues from two transcription factors, Dll and Sp1, and two signaling pathways, Wg and Dpp, in order to execute a leg patterning developmental program. Analogous EGFR CRMs are likely to exist in mammals, especially because complex interactions between BMP, Wnt, Shh, multiple Dlx paralogs and other factors, are implicated in the induction of FGF signaling in mammalian limb development. Consistent with this idea, specific single nucleotide polymorphisms (SNPs) in humans in non-coding loci of genes encoding EGFR ligands have been shown to be associated with different types of cancer. Such loci may be enhancer elements analogous to vnE and rhoE. It is also noted that the regulatory logic uncovered in this study is likely to be relevant to many CRMs and genes that share spatial and temporal expression programs. Exploiting this regulatory logic in other systems might streamline the identification of enhancer elements that will aid in the discovery of mechanisms that are relevant to EGFR-related human disease and developmental birth defects (Newcomb, 2018).

Egfr signaling is a major regulator of ecdysone biosynthesis in the Drosophila prothoracic gland

Understanding the mechanisms that determine final body size of animals is a central question in biology. In animals with determinate growth, such as mammals or insects, the size at which the immature organism transforms into the adult defines the final body size, as adult individuals do not grow. In Drosophila, the growth period ends when the immature larva undergoes the metamorphic transition to develop the mature adult. This metamorphic transition is triggered by a sharp increase of the steroid ecdysone, synthetized in the prothoracic gland (PG), that occurs at the end of the third instar larvae (L3). It is widely accepted that ecdysone biosynthesis in Drosophila is mainly induced by the activation of tyrosine kinase (RTK) Torso by the prothoracicotropic hormone (Ptth) produced into two pairs of neurosecretory cells that project their axons onto the PG. However, the fact that neither Ptth nor torso-null mutant animals arrest larval development but only present a delay in the larva-pupa transition mandates for a reconsideration of the conventional model. This study shows that Egfr signaling, rather than Ptth/torso, is the major contributor of ecdysone biosynthesis in Drosophila. Egfr signaling was found to be activated in the PG in an autocrine mode by the EGF ligands spitz and vein, which in turn are regulated by the levels of ecdysone. This regulatory positive feedback loop ensures the production of ecdysone to trigger metamorphosis by a progressive Egfr-dependent activation of MAPK/ERK pathway, thus determining the animal final body size (Cruz, 2020).

In contrast to the developmental delay phenotype observed in larvae with reduced Ptth or torso, this study foudn that specific depletion of Drosophila homolog transducers ras(ras85D), Raf oncogene (Raf), and ERK, the core components of the MAPK/ERK pathway, in the prothoracic gland (PG) using the phmGal4 driver (phm>) induced developmental arrest at L3. This result suggests that additional RTKs might play important roles in ecdysone production. To study this possibility, all known Drosophila RTKs in the PG were knocked down and found that only depletion of Egfr phenocopied L3 arrested development observed in phm > ras85DRNAi larvae. Likewise, overexpression in the PG of a dominant-negative form of Egfr (Egfr^DN) or depletion of the transcription factor pointed (pnt), the principal nuclear mediator of the Egfr signaling pathway, also resulted in arrested L3 larvae. The same results were obtained upon inactivation of Egfr or different components of the MAPK/ERK pathway using an alternative PG specific driver, amn^c651Gal4. Consistent with the observed phenotypes, overexpression of a constitutively activated form of either Egfr (Egfr^act) or Pnt (PntP2^VP16) in the PG induced premature pupariation and reduced pupal size. These results are in agreement with a previous report showing that overexpression of a constitutively activated form of Ras (Ras^V12) in the PG produced the same phenotype. Furthermore, overexpression of Ras^V12 in Egfr-depleted larvae rescued the developmental arrest phenotype and forced premature pupation. These results strongly suggest that Egfr signaling in the PG is required for the synthesis of the ecdysone pulse that triggers metamorphosis. Confirming this hypothesis, ecdysone titers in larvae depleted of either Egfr or pnt in the PG were dramatically reduced. Accordingly, Hr3 and Hr4 expression, two direct target genes of the hormone that have been used as readouts for ecdysone levels, was completely abolished in phm > EgfrRNAi and phm > pntRNAi L3 larvae compared to control animals. Moreover, addition of the active form of ecdysone, 20-hydroxyecdysone (20E), to the food rescued the developmental arrest phenotype induced by inactivation of Egfr signaling in the PG. Altogether, these results indicate that Egfr signaling in the PG endocrine cells is required for the production of the ecdysone pulse that triggers pupariation and fixes adult body size (Cruz, 2020).

Since Egfr signaling is involved in cell proliferation and survival, this study analyzed whether the above-described phenotype was due to compromised viability of PG cells. Although reduced activation of Egfr signaling diminished cell size, PG cell number and viability were not affected. Interestingly, ecdysone synthesis has been recently shown to correlate with endocycle progression and therefore cell size of PG cells. PG cells undergo three rounds of endoreplication during larval development resulting in chromatin values (C values) of 32-64 C by late L3. Remarkably, a clear reduction was observed in the C value of PG cells of phm > EgfrRNAi larvae at 120 h AEL, with most cells at 8-16 C, indicating that Egfr activation is also required to promote polyploidy in the PG cells (Cruz, 2020).

This result raised the possibility that Egfr signaling regulates ecdysone production by determining the size of the PG. To analyze this hypothesis, the effect of Egfr signaling in ecdysone production was examined. Steroidogenesis in the PG cells depends on the timely expression of ecdysone biosynthesis enzyme-encoding genes that mediate the conversion of cholesterol to ecdysone. To analyze whether Egfr signaling controls ecdysone synthesis by regulating the expression of these genes, qRT-PCR was performed in early (72 h after egg laying [AEL]), mid (96 h AEL), and late (120 h AEL) phm > EgfrRNAi and phm > pntRNAi L3 larvae. Whereas expression of the six ecdysone biosynthetic genes increased gradually from mid to late L3 in control larvae, correlating with the production of the high-level ecdysone pulse that triggers metamorphosis, inactivation of the Egfr pathway in the PG resulted in a dramatic reduction in the expression levels of neverland (nvd), spook (spo), shroud (sro), and phantom (phm) in late L3 larvae. In contrast, the expression of disembodied (dib) and shadow (sad) was not significantly reduced in Egfr-depleted larvae, which suggests that compromising Egfr signaling in the PG does not result in a general reduction in the transcriptional activity by its minor C value, as previously shown, but rather by a specific transcriptional effect. Further confirming this point, the overexpression of CycE in Egfr-depleted PGs was unable to restore normal expression of ecdysteroid biosynthetic genes nor induced proper pupariation of these animals, indicating that Egfr signaling is required for proper expression of ecdysone enzyme-encoding genes independently of promoting polyploidy of PG cells (Cruz, 2020).

As the levels of circulating ecdysone are influenced by the rates of hormone production and release, whether Egfr signaling also regulates ecdysone secretion was studied. Recently, it has been shown that ecdysone secretion from the PG cells is mediated by a vesicular regulated transport mechanism. After its synthesis, ecdysone is loaded through an ATP-binding cassette (ABC) transporter, Atet, into Syt1-positive secretory vesicles that fuse to the cytoplasmic membrane for release of the hormone in a calcium-dependent signaling. To analyze the role of Egfr signaling in this process, secretory vesicles were visualized in PG cells of phm > EgfrRNAi and phm > pntRNAi L3 larvae by expressing eGFP-tagged Syt1 (Syt-GFP) in these glands. Whereas Syt-GFP vesicles accumulate at the plasma membrane with a small number of vesicles in the cytoplasm in wild-type L3 larval PGs, a dramatic accumulation of Syt-GFP vesicles in the cytoplasm was observed in PGs with reduced Egfr signaling. Similar results were obtained when the subcellular localization of the ecdysone transporter Atet-GFP was analyzed. Consistently, overexpression of ras^V12 in PGs of phm > EgfrRNAi larvae restored the subcellular localization of both Syt and Atet-GFP. Furthermore, mRNA levels of several genes involved in vesicle-mediated release of ecdysone, including Syt and Atet, were dramatically downregulated in the PG of phm > EgfrRNAi and phm > pntRNAi larvae. Therefore, the results show that Egfr signaling is also required for the vesicle-mediated release of ecdysone from PG cells. Interestingly, direct effects of Egfr signaling on the endocytic machinery have been already described in Drosophila tracheal cells as well as in human cells (Cruz, 2020).

The next question was to determine which of the EGF ligands were responsible for the Egfr pathway activation in the PG. In Drosophila, Gurken (Gur), Spitz (Spi), Keren (Krn), and Vein (Vn) serve as ligands for Egfr. Expression analysis of the four ligands revealed that only vn and spi were expressed in the PGs. Consistently, the intramembrane protease rhomboid (rho), which is necessary for the proteolytic activation of Spi, was also expressed in the PG cells. A temporal expression pattern of staged L3 PGs revealed that rho expression progressively increased during the last larval stage, while the expression of spi and vn increased sequentially, with vn upregulated at mid L3 and spi at late L3. Consistent with the expression of the ligands, mRNA levels of Egfr also showed a clear upregulation by late L3. Likewise, a specific expression of PntP2 isoform was also observed in the PG of late L3 larvae. Altogether, these results suggest that Vn and Spi might activate Egfr signaling in an autocrine manner to induce ecdysone production (Cruz, 2020).

To determine the functional relevance of each ligand, vn, spi, or both simultaneously were knocked down in the PG. As in the case of phm > EgfrRNAi, depletion of spi, vn, or both ligands at the same time caused developmental arrest in L3, although Spi appeared to have a minor effect as around 40% of phm > spiRNAi larvae underwent delayed pupariation. The attenuated effect of spi-depleted animals was probably due to a weaker effect of the spiRNAi lines as depletion of the Spi-processing protease rho in the PG resulted in all phm > rhoRNAi animals arresting development at L3. Importantly, ecdysteroid levels in mid and late L3 were significantly reduced in animals depleted of either vn or spi. Consistent with their role in controlling ecdysone production, overexpression of either Vn or an active-cleaved form of Spi in the PG induced precocious pupariation and smaller pupae. Altogether, these findings show that spi and vn act in an autocrine manner as Egfr ligands in the PG to induce ecdysone biosynthesis during the last larval stage. In fact, the correlation between vn and spi expression with the occurrence of increasing levels of ecdysteroids points to a possible positive-feedback loop regulation with 20E inducing vn and spi expression. Consistent with this possibility, vn and spi mRNA levels were reduced in PGs of ecdysteroid deficient larvae that were generated by depleting spo (phm > spoRNAi) or by overexpressing a dominant-negative form of the ecdysone receptor (phm > EcRDN). Moreover, staged PGs were cultured for 6 h ex vivo in presence or absence of 20E, and vn and spi mRNA levels were found to be significantly upregulated in the presence of the hormone. Altogether, these observations demonstrate that ecdysone exerts a positive-feedback effect on PG cells amplifying its own synthesis by inducing the expression of vn and spi. This result is consistent with a previous proposed model of ecdysone regulation in an autonomous mechanism by a positive feedback and biogenic amines. Thus, a model is proposed in which increasing levels of ecdysone promote the expression of vn and spi in the PG cells, which, in turn, increases Egfr signaling in this gland in an autocrine manner to further promote the production of ecdysone. Interestingly, it has been already shown that expression of Spi and Vn in midgut cells of Drosophila depends on ecdysone activity during metamorphosis. In addition, in vertebrates, other hormones have been postulated to control Egfr activity, such as Thyrotropin-releasing hormone, which induced the phosphorylation and activation of the Egf receptor, leading to specific transcriptional events in GH3 pituitary cells. Likewise, the Growth Hormone modulates Egfr trafficking and signaling by activating ERKs (Cruz, 2020).

Thus far, the results above show that MAPK/ERK pathway is a central regulatory element in the control of ecdysone biosynthesis in the PG, with Egfr signaling chiefly contributing to its activity. However, since Ptth/torso signaling operates through the same MAPK/ERK pathway the relative contribution of this signaling pathway in the overall activity of the PG was investigated. The fact that inactivation of Egfr signaling in the PG did not affect the mRNA expression levels of either Ptth or torso points to a minor contribution of Ptth/torso signaling in the overall MAPK/ERK activity. To analyze this possibility, the levels were compared of dpERK, a readout of MAPK/ERK activity, in PGs of phm > EgfrRNAi and phm > torsoRNAi larvae. A dramatic reduction of dpERK levels was observed in PGs of phm > EgfrRNAi larvae. Importantly, dpERK levels were also reduced in phm > torsoRNAi PGs, although to a significant lesser extent when compared to phm > EgfrRNAi larvae. Similar results were observed when nuclear accumulation of dpERK was analyzed in both larvae. Consistently, the level of activity of the MAPK/ERK pathway in phm > pntRNAi and phm > torsoRNAi larvae correlated very well with expression of the biosynthetic enzyme phm and the ecdysone-responsive genes Hr3, Hr4, and Broad-Complex (BrC), although the levels of ecdysone were significantly reduced in both cases. The different level of activation of dpERK by Egfr and Ptth/torso signaling was also consistent with the respective accumulation of Syt-GFP and Atet-GFP vesicles at the cytoplasm and the reduction of the C value of PG cells. Finally, it is important to note that the level of activity of the MAPK/ERK pathway correlated with the respective phenotypes upon inactivation of each pathway, with phm > EgfrRNAi larvae arresting development at L3 and phm > torsoRNAi larvae presenting only a delay in the pupariation time. In line with this, whereas over-activation of Egfr pathway in the PG of phm > torsoRNAi larvae induced a significant advancement in pupariation, the expression of a constitutively activated form of Torso (torsoD4021 mutants) in PGs with depleted Egfr (EgfrRNAi; torso D4021) was not able to induce precocious pupariation (Cruz, 2020).

Overall, these results show that the Egfr signaling pathway plays the main role in the biosynthesis of ecdysone by activating the MAPK/ERK pathway in the PG during mid-late L3, whereas Ptth/torso signaling acts synergistically only to increase the MAPK/ERK pathway activity thus accelerating developmental timing. In this regard, it is possible that the different strength of MAPK/ERK activation by the two signaling pathways might underline this distinct requirement of each pathway. Furthermore, temporal expression of the Egfr and Torso ligands may also contribute to the difference strengths of MAPK/ERK activation, as EGF ligands vn and spi are highly expressed during L3, whereas Ptth is only upregulated at a specific developmental time, the wandering stage. Taken together, these data suggest a model in which the increasing circulating levels of ecdysone during the last larval stage are induced by a progressive Egfr dependent activation of MAPK/ERK in the PG, whereas Ptth/torso signaling further regulates ecdysone production by integrating different environmental signals such as nutritional status, crowding conditions, and light. It is important to note that, in addition to the Egfr and Ptth/torso pathways, ecdysone biosynthesis is also regulated by the insulin/insulin-like growth factor signaling (IIS)/target of Rapamycin (TOR) signaling pathway. However, in contrast to the major role of Egfr controlling ecdysteroid levels during mid-late L3, including the strong ecdysteroid pulse that triggers pupariation, the main effect of IIS/TOR pathway is to control the production of the small ecdysteroid peak that is associated to the nutrition-dependent critical weight checkpoint that occurs at the very early L3. Thus, decreasing the IIS/TOR activity in the PG delays the critical weight checkpoint, slowing development and delaying pupariation, while increasing IIS/TOR activity in the gland induces precocious critical weight and accelerates the onset of metamorphosis. Nevertheless, it is conceivable that the increasing levels of ecdysone at the critical weight checkpoint might initiate the expression of the Egf ligands, that in turn activates the ecdysone production during mid-late L3 (Cruz, 2020).

Finally, since no role of Ptth/torso signaling has been characterized in hemimetabolous insects, it is postulated that Egfr signaling might be the ancestral ecdysone biosynthesis regulator, whereas Ptth/torso signaling has probably been co-opted in holometabolous insects during evolution to fine-tune the timing of pupariation in response to changing environmental cues. Consistent with this view, depletion of Gb-Egfr in the hemimetabolous insect Gryllus bimaculatus, where no Ptth/torso has been described, results in arrested development by the last nymphal instar. Therefore, this double regulation in holometabolous insects might provide developmental timing plasticity contributing to an appropriated adaptation to a time-limited food supply (Cruz, 2020).

GENE STRUCTURE

Type 1 and type 2 cDNAs differ by the inclusion of an intron between exons 4 and 5 in the processed mRNA of type II messenger, resulting in an extra 1782 bp in the type II messenger (Schnepp, 1996).

Bases in 5' UTR - 2054

Exons - 5

Bases in 3' UTR - 1131

PROTEIN STRUCTURE

Amino Acids - 622 (type I) and 621 (type II)

Structural Domains

The vein cDNAs encode two proteins that differ only in their extreme C-terminal regions. These proteins are similar to residue 606 (the end of exon 4) where they diverge because of a 1782 bp region in the larger cDNA. The type 1 protein ends with 16 amino acids that differ from the terminal 15 amino acids in the type 2 protein. There is an N-terminal signal sequence, and the protein is secreted. A PEST region is present, found in proteins with short half lives, as well as an opa (polyglutamine) region, an Ig-like domain and an EGF-like domain. There are a number of residues that are potential N-linked glycosylation sites (Schnepp, 1996).

The EGF-like domain in VN is 43 amino acids long and has the six invariant cysteines and highly conserved glycine and arginine residues characteristic of the motif. The cysteines are thought to form disulfide bonds, thereby producing a looped structure. The EGF motif in VN is between 30% and 44% identical to other EGF-like ligands. It shares 37% identity with Spitz and 33% identity with Gurken. Amino-terminal to the EGF domain is an Ig-like domain of the C2 type that includes nonimmunological proteins. The IG-like domain in VN has the two invariant cysteine residues typical of the domain. The Ig region in VN shares between 33 and 35% identity with the Ig region in the neuregulins and is 32% identical with the Ig region in NCAM (Schnepp, 1996).

date revised: 5 February 2000  

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.