Early dpp expression pattern in the ectoderm is dynamic, consisting of three phases. Phase I, in which dpp is expressed in a broad dorsal domain, depends on elements in the dpp second intron that interact with the Dorsal transcription factor to repress transcription ventrally. In contrast, in phases II and III, dpp is expressed first in broad longitudinal stripes (phase II) and subsequently in narrow longitudinal stripes (phase III) (Schwyter, 1995). dpp is also expressed in the visceral mesodermal midgut, where it regulates formation of caeca. DPP has a major role in compartment formation between visceral segments and is expressed there during the process of segmentation (Manek, 1994).

dpp expression in the gut, at least some of which is presumably endodermal, includes the presumptive pharynx, a portion of the presumptive exophagus, the primordia of the gastric caeca, the parasegment 7 region of the midgut, and a portion of the presumptive hindgut. As these expression patterns are generated from promoter fragments, the finding of DPP expression in the endoderm should remain controvertial until confirmed (Jackson, 1994).

The Drosophila visceral mesoderm (VM) is a favorite system for studying the regulation of target genes by Hox proteins. The VM is formed by cells from only the anterior subdivision of each mesodermal parasegment (PS). The VM itself acquires modular anterior-posterior subdivisions similar to those found in the ectoderm. Mesodemal cells located just under the engrailed-expressing cells in the posterior ectodermal compartment have been called the mesodermal "P domain." The dorsal-most cells of the mesodermal P domain in each PS express the homeobox gene bagpipe (bap); they detach from the mesodermal fold and move inward toward the center of the embryo. These bap-expressing cells form the VM progenitor groups. The VM subdivisions, and the metameric expression of Connectin, form in response to ectodermal production of secreted signals encoded by the segment polarity genes hedgehog and wingless and are independent of Hox gene activity. A cascade of induction from ectoderm to mesoderm to endoderm thus subdivides the gut tissues along the A-P axis. Induction of VM subdivisions may converge with Hox-mediated information to refine spatial patterning in the VM. Con patches align with ectodermal engrailed stripes, so the VM subdivisions correspond to PS 2-12 boundaries in the VM. The PS boundaries demarcated by Connectin in the VM can be used to map expression domains of Hox genes and their targets with high resolution. The resultant map suggests a model for the origins of VM-specific Hox expression in which Hox domains clonally inherited from blastoderm ancestors are modified by diffusible signals acting on VM-specific enhancers (Bilder, 1998b).

Since Con expression marks the imprint of ectodermal PS boundaries on the VM, Con patches can be used to precisely map the domains of Hox gene transcription in relation to Con patches. teashirt is expressed in two domains. The anterior midgut domain extends from visceral mesoderm segment (VS) 4 to mid-VS 6, where it shares a posterior boundary with Antennapedia; the central midgut domain extends several cells to either side of the VS 8 boundary. dpp is also expressed in two domains: at the gastric caeca, it is found in the A domain of VS2 and the P domain of VS 3, while in the central midgut it extends from the A domain of VS 6 to terminate just anterior to the VS 8 boundary. wg is expressed just anterior to the VS 8 boundary, with some cells after stage 12 lying in VS 8. pnt is expressed throughout VS 8, although expression is not seen until early stage 13. At stage 13, the two domains of odd paired (opa) expression extend from the P domain of VS 4 to the VS 6 boundary and from VS 9 through VS 11 (Bilder, 1998b).

The formation of the tracheal network in Drosophila is driven by stereotyped migration of cells from the tracheal pits. No cell divisions take place during tracheal migration and the number of cells in each branch is fixed. This work examines the basis for the determination of tracheal branch fates, prior to the onset of migration. The EGF receptor pathway is activated by localized processing of the ligand Spitz in the tracheal placodes and is responsible for the capacity to form the dorsal trunk and visceral branch. Prominent double phosphorylation of Erk (Rolled) is detected in the tracheal placodes at stage 10-11. This pattern is Efgr-dependent and is abolished in rhomboid mutants. The double phosphorylated Erk domain is broader than the region of rhomboid expression. Since Rhomboid is known to regulate Spitz processing, this pattern probably reflects the diffusion of the secreted form of Spitz originating within the rhomboid-expressing cells, in the central part of the placode. In mutants for Egfr, tracheal pits appear normal, although certain tracheal branches fail to develop: specifically, the dorsal trunk and visceral branch are missing or incomplete. spalt mutants show specific defects in the migration of dorsal trunk cells, pointing to an important role for spalt in subdivision of tracheal fates. The Dpp pathway is induced in the tracheal pit by local presentation of Dpp from the adjacent dorsal and ventral ectodermal cells. This pathway patterns the dorsal and lateral branches. Elimination of both Dpp and Egfr pathways blocks migration of all tracheal branches. Antagonistic interactions between the two pathways are demonstrated. The opposing activities of two pathways may refine the final determination of tracheal branch fates. Egfr-dependent activation of Erk (Rolled) in the tracheal placode precedes the activation of the same pathway by Breathless. Only after Egfr induction is diminished, does a new double phosphorylated Erk pattern appear, induced by Breathless. It is proposed that two opposing gradients of Dpp and Spitz are operating within the placode. the cells in the center of the tracheal placode encounter high concentrations of secreted Spitz, and low or negligable levels of Dpp. Conversely, the cells located at the dorsal and ventral domains of the placode encounter high concentrations of Dpp and low levels of secreted Spitz. Therefore, induction by the Egfr and Dpp pathways creates three subsets of cells: dorsal, central and ventral (Wappner, 1997).

Wing and leg precursors of Drosophila are recruited from a common pool of ectodermal cells expressing the homeobox gene Dll. Induction by Dpp promotes this cell fate decision toward the wing and proximal leg. The receptor tyrosine kinase Egfr antagonizes the wing-promoting function of Dpp and allows recruitment of leg precursor cells from uncommitted ectodermal cells. By monitoring the spatial distribution of cells responding to Dpp and Egfr, it has been shown that nuclear transduction of the two signals peaks at different positions along the dorsoventral axis when the fates of wing and leg discs are specified and that the balance of the two signals assessed within the nucleus determines the number of cells recruited to the wing. Differential activation of the two signals and the cross talk between them critically affect this cell fate choice (Kubota, 2000).

In a screen for genes expressed in the embryonic limb primordia, rhomboid was found to be transiently expressed in the central part of Dll-expressing limb primordia in stage 11 embryos. rho transcription is the rate-limiting step of the activation of an EGFR ligand Spitz. As expected from the role of rho as a stimulator of Egfr, a transient expression of an activated, phosphorylated form of MAPK (dpMAPK) is detected in the nucleus of limb primordial cells surrounding the rho-expressing cells. The dpMAPK expression starts after the initiation of Dll transcription and diminishes before the separation of the wing and leg disc primordium. The dpMAPK expression is undetectable in null mutants of rho or Egfr. The peak of dpMAPK expression is located ventrally to the cells expressing dpp. The results suggest that rho-mediated stimulation of Egfr and MAPK occurs at the time of cell fate specification of wing and leg discs (Kubota, 2000).

The spatial distribution of cells responding to Dpp and its relationship to Egfr signals was studied. To this end, an antibody specific to phosphorylated C-terminal sequence of Mad was produced. The phosphorylated sequence corresponds to the site at which the type I BMP receptor phosphorylates SMad1. The antibody detects an antigen distributed in a pattern similar to, but broader than, that of DPP mRNA. This immunoreactivity is dependent on Dpp signaling, as it is absent in stage 11 mutants of thick veins encoding type I Dpp receptor and in dpp mutants. This indicates that other extant TGFbeta-related signaling molecules present in Drosophila embryos do not substitute for Dpp to induce this immunoreactivity. Conversely, ectopic expression of Dpp results in high accumulation of this immunoreactivity. These results suggest that the antibody detects a Dpp-specific signaling event, most likely the phosphorylation and nuclear transport of Mad. Hereafter, the immunoreactivity detected by this antibody is called pSSVS (Kubota, 2000).

pSSVS is found mainly localized in the nucleus and distributed in regions a few cells wider in diameter than those of dpp-expressing cells. These properties are consistent with the previous findings that Mad transduces the Dpp signal to the nucleus. Double labeling of pSSVS and DLL mRNA shows that pSSVS expression is higher in the dorsal region of Dll-expressing cells. Combined with the double-labeling results of dpMAPK and Dll or dpp, it is concluded that cells responding to Dpp and Egfr overlap, but the peak of the responses are shifted. Such differential distribution of the two signals results in an arrangement of cells responding to a different strength of Dpp and Egfr along the dorsoventral axis (Kubota, 2000).

To study the role of Egfr at the stage of wing and leg cell fate determination, specific marker gene expression was examined in Egfr signaling mutants. DLL mRNA is expressed in the entire limb primordium at stage 11 and becomes restricted to distal leg cells at stage 15. Esg protein expression was used to detect both wing and proximal leg cells. In rho mutants, the size of limb primordia at stage 11 is the same as the control, but the later development of leg discs is abnormal. The number of leg disc cells expressing Dll and/or Esg at stage 15 is reduced, and these cells no longer show the circular arrangement typical of leg disc precursors. Amorphic mutation of Egfr cause a ventral expansion of limb primordia as a result of a loss of the early function of Egfr, but the expression of leg markers is severely reduced at stage 15. A similar phenotype is observed in mutants lacking both maternal and paternal copies of Dsor1, which encodes a MAP kinase kinase. In all cases described above, Esg-expressing cells at the dorsal part of leg discs are most frequently lost, suggesting that the development of dorsoproximal leg cells is most sensitive to the loss of Egfr activity. In contrast, wing and leg disc development is normal in vein mutants, suggesting the putative ligand of Egfr encoded by this gene is dispensable. These results suggest that MAPK activation induced by Rho and Egfr is essential for normal leg development (Kubota, 2000).

The temporal requirement for Egfr was studied by the temperature-sensitive allele Egfrf1. When the temperature is increased to the restrictive temperature at 5 hours after egg laying (AEL) prior to the induction of the limb primordium, the expression of Dll is expanded to the ventral midline, as was also observed with the strong Egfr mutants. When the temperature is increased at 6 hour AEL, the initial Dll expression is not altered, but the leg disc development is severely affected. Only mild defect is found in leg discs when the temperature is increased at 7 hours AEL, suggesting that Egfr must function between 6 and 7 hours AEL to correctly specify the leg cell fate. This is the time when the transient activation of MAPK is observed. Furthermore, whether Egfr is required autonomously in limb primordial cells was examined by expressing a dominant-negative form of Egfr using Dll-Gal4. Expression of this driver starts in the limb primordium at stage 11 and persists in a subset of wing discs and in entire leg discs at stage 15 because of the persistence of Gal4 activity. Imaginal disc-specific inhibition of Egfr interfers with leg disc development, while leaving the wing disc intact. These results demonstrated that a transient activation of Egfr in stage 11 limb primordia is essential for the leg disc development (Kubota, 2000).

In contrast to the severe defects in leg discs, none of the mutations in Egfr signaling interfer with wing disc formation. In these mutants, wing primordia consistently express Esg and another wing disc marker Vestigial, and invaginate to form discs. However, an increase in the number of wing disc cells has been noted in Egfr signaling mutants. This effect was analyzed in rho mutants; unlike Egfr mutants, in rho mutants the number of limb primordial cells at stage 11 is the same as the control. The number of Esg-expressing wing disc cells in rho mutants is increased compared to the control, while the number of the proximal leg disc cells is severely reduced. It is concluded that Egfr signaling is required to limit wing disc cell differentiation in limb primordial cells that are not yet fully committed. It is inferred that a subset of prospective leg cells that do not receive a sufficient amount of Egfr signaling fail to differentiate as proximal leg and instead adopt a wing fate (Kubota, 2000).

The increase in the number of wing disc cells in rho mutants resembles the overexpression phenotype of Dpp and raises a possibility that Egfr might prevent wing disc development by negatively regulating Dpp signaling. Such a cross talk could occur at several levels including the following: (1) regulation of dpp transcription, (2) signal transduction from Dpp receptors to the nucleus, and (3) transcriptional regulation of downstream target genes. The analyses excluded the first two possibilities for two reasons. (1) The expression pattern of DPP mRNA is unaffected by the mutation of rho. A previous report showing an expansion of dpp expression in Egfr mutants probably reflects the global patterning role of Egfr in the earlier stage. (2) pSSVS expression around limb primordia does not change in rho mutants. Conversely, the expression pattern of dpMAPK is not changed by a null mutation of tkv. These results suggest that the differential distribution of cells responding to Dpp and Egfr is set up independently of each other's activity (Kubota, 2000).

Dpp and Egfr were found to antagonize each other after signal transduction into the nucleus. Hyperactivation of Egfr by an ectopic expression of an Egfr ligand Spitz causes a great accumulation of dpMAPK. As expected from the negative effect of Egfr on the wing development, this treatment completely eliminates wing disc formation and, in addition, causes a malformation of the leg disc. Since it was found that cells migrating out of the leg primordium express dpMAPK, it is unlikely that the failure in wing disc formation is due to the prevention of cell migration or to cell death. It has been suggested that hyperactivation of Egfr prevents limb primordial cells from adopting the wing cell fate. It is likely that those cells adopt the epidermal fate instead. Overexpression of Dpp causes an accumulation of pSSVS and an increase in the number of wing disc cells. Coexpression of Dpp with Spi partially restores the development of both wing and leg discs, suggesting that wing disc development overcomes the negative effect of Egfr if provided with a sufficient amount of Dpp. The restored wing primordia migrate with high levels of pSSVS and dpMAPK, further supporting the notion that Dpp and Egfr signals are transduced independently of one another (Kubota, 2000).

dad is an immediate transcriptional target gene of Dpp, the expression of which closely parallels that of pSSVS expression in embryos and is inducible by Dpp. dad expression is not affected in Egfr or rho mutants. Furthermore, elevated dad expression induced by Dpp is not affected by sSpi, suggesting that at least one of the immediate transcriptional responses to Dpp is unaffected by elevated Egfr signaling (Kubota, 2000).

The antagonism between Dpp and Egfr during wing disc development raises a question: what is the default state of the wing and leg primordia in the absence of the two signals? Double mutant phenotypes of Dpp and Egfr signaling were examined. tkv mutants lack wing discs and their leg discs are malformed. This phenotype reflects a disc cell autonomous requirement for Dpp signaling, because the phenotype is reproduced by the disc-specific inhibition of Dpp signaling by dad, which inhibits Mad. The phenotype of either tkv;rho or tkv;Egfr double mutants is a simple addition of each mutation, in which wing discs are lost completely and leg discs are severely reduced. Since Dll-expressing limb primordial cells are present in tkv;Egfr double mutants in stage 11, it has been concluded that these cells fail to differentiate as wing discs and their ability to differentiate as leg discs is also compromised. A few Esg-positive cells remain at the position of the leg, and it is speculated that this reflects the presence of a second leg-inducing signal. These results suggest that Dpp is absolutely required for wing disc development irrespective of the activity of Egfr (Kubota, 2000).

Egfr affects the choice of wing vs. leg developmental options differently; it promotes leg development while it inhibits wing development. These two activities of Egfr are the earliest of known events of leg specification, and occur prior to the establishment of proximodistal axis in the leg. In the absence of late functions of Dpp and Egfr, limb primordia are specified but fail to differentiate into wing disc and most of leg disc. Thus it is proposed that early limb primordium at stage 11 consists of cells not yet fully committed to either wing or leg disc fate, and the cells are exposed to different amounts of Dpp and Egfr signaling according to their dorsoventral location. Dpp recruits the cells to the wing disc fate. Egfr antagonizes the cellular response to the wing-inducing function of Dpp and allows the development of wing discs only in the dorsal region. Thus the dorsoventral difference in Dpp and Egfr signaling in the limb primordium provides key information to the separation and differentiation of the wing and leg discs. In contrast to the opposing roles of Dpp and Egfr on wing disc development, leg discs requires both signals. The effect of the loss of Egfr activity on leg disc development is not compensated for by a simultaneous loss of Dpp signaling, indicating that Egfr has an additional activity to promote leg development separately from its role to antagonize Dpp. Because dorsal and ventral limb primordial cells respond to Egfr differently, it is speculated that at least one additional dorsoventral factor influences leg disc formation at stage 11. This idea is consistent with the fact that residual proximal leg cells can still be induced in the almost complete absence of Egfr and Dpp activity. One candidate for the factor is Wg, which is expressed in the limb primordium (Kubota, 2000).

The nuclear transduction of the Dpp signal, as visualized by the distribution of pSSVS and expression of dad, is unaffected by Egfr. The results suggest that the antagonistic effect of Egfr on Dpp signaling occurs after transduction into the nucleus. Therefore, the mechanism of SMad inhibition by direct phosphorylation by MAP kinase does not play a major role in this case (Kubota, 2000).

The finding that Egfr is activated in the limb primordium and prevents wing disc formation suggests that Egfr is a key factor in the diversification of the wing and leg fate. It is proposed that the differential activation of Dpp and Egfr, and the dorsal cell migration brings a subset of limb primordial cells out of the range of Egfr signaling, and thereby allows Dpp to induce wing development. It follows that dorsally migrating cells acquire the wing cell identity only after the separation from leg-promoting signals. Consistent with this idea, expression of wing-specific markers Vg and Sna, start only after the separation of the two primordia. Mechanisms that promote the dorsal cell migration remain to be identified. Given that the basic genetic components for the induction of the wing and leg have been identified in the model organism Drosophila, it can now be asked how the genetic mechanism of wing and leg specification has evolved by comparing the expression and function of these genes in limb primordial cells of primitive insects (Kubota, 2000).

Stepwise formation of a SMAD activity gradient during dorsal-ventral patterning of the Drosophila embryo

Genetic evidence suggests that the Drosophila ectoderm is patterned by a spatial gradient of bone morphogenetic protein (BMP). Patterns have been compared of two related cellular responses - signal-dependent phosphorylation of the BMP-regulated R-SMAD, MAD, and signal-dependent changes in levels and sub-cellular distribution of the co-SMAD Medea. Nuclear accumulation of Medea requires a BMP signal during blastoderm and gastrula stages. During this period, nuclear co-SMAD responses occur in three distinct patterns. At the end of blastoderm, a broad dorsal domain of weak SMAD response is detected. During early gastrulation, this domain narrows to a thin stripe of strong SMAD response at the dorsal midline. SMAD response levels continue to rise in the dorsal midline region during gastrulation, and flanking plateaus of weak responses are detected in dorsolateral cells. Thus, the thresholds for gene expression responses are implicit in the levels of SMAD responses during gastrulation. Both BMP ligands, DPP and Screw, are required for nuclear co-SMAD responses during these stages. The BMP antagonist Short gastrulation (Sog) is required to elevate peak responses at the dorsal midline as well as to depress responses in dorsolateral cells. The midline SMAD response gradient can form in embryos with reduced dpp gene dosage, but the peak level is reduced. These data support a model in which weak BMP activity during blastoderm defines the boundary between ventral neurogenic ectoderm and dorsal ectoderm. Subsequently, BMP activity creates a step gradient of SMAD responses that patterns the amnioserosa and dorsomedial ectoderm (Sutherland, 2003).

These in vivo studies validate the molecular model for signal-dependent nuclear accumulation of Medea. Nuclear accumulation of Medea requires both competence to oligomerize and MAD. Nuclear accumulation is signal dependent, requiring both BMP ligands, Dpp and Scw. Conversely, all cells accumulated nuclear Medea in the presence of constitutively active Tkv receptor. At these stages, any independent contribution from activin-like signals is below the detection limit (Sutherland, 2003).

Furthermore, levels of Medea determine the strength of BMP responses at these stages. Medea overexpression leads to expansion of the dorsal-most fate, with increased numbers of amnioserosa cells. Signal-dependence for nuclear accumulation is retained. Decreased Medea exacerbates loss of amnioserosa from reduced Dpp levels (Sutherland, 2003).

The intensity of Medea staining was surprisingly sensitive to signal activity. However, steady-state levels of Medea are unaffected by the level of BMP activity. The antibodies appear highly sensitive to a Medea conformation that is prevalent in the nucleus, most probably an active SMAD complex. This sensitivity makes nuclear Medea an excellent assay to distinguish spatial patterns of endogenous BMP activity (Sutherland, 2003).

In wild-type embryos, two transitions in the distribution of BMP activity are evident. Many cellular blastoderm embryos lack detectable levels of nuclear Medea, but a few have low levels of nuclear Medea in a broad dorsal domain, with little gradation. From the proportion of cellular blastoderm embryos with this pattern, the duration of nuclear Medea appears to be brief. These data parallel reports of broad, weak PMad staining during mid-cellularization, except that nuclear Medea is detected later and in a broader pattern. The time lag between the earliest reported detection of PMad and detection of nuclear Medea probably stems from a combination of technical differences and the time necessary for nuclear accumulation. In sum, initial BMP activity is weak and distributed broadly in dorsal regions. Low BMP activity at this phase is required to maintain the early phase of zen expression (Sutherland, 2003).

Onset of gastrulation is associated with a dramatic change in the domain of nuclear Medea, which narrows to a tight midline stripe of cells while staining levels intensify. PMad shows a similar transition to a narrower domain, but earlier. Thus, lateral SMAD responses became undetectable just as a steep activity gradient forms along the dorsal midline (Sutherland, 2003).

A third response pattern arises during mid-gastrulation: dorsolateral domains of cells exhibit low levels of nuclear Medea. Response levels remain high in the dorsal-most cells, even as they move laterally during gastrulation. Levels fall off rapidly over a few cells on either side, with a sharp transition to flanking plateaus of weak responses. The subcellular distribution of Medea is unchanging in ventral and ventrolateral cells. The full BMP response domain does not extend as far ventrally as it does during blastoderm, even though many dorsal cells move laterally during germband extension. Thus, the lateral-most cells with responses at blastoderm have decreased responses during gastrulation (Sutherland, 2003).

In sum, the dorsal midline stripe of SMAD responses corresponds to a steep BMP activity gradient, with thresholds that correlate with patterning markers. The edges of the Medea peak response correlates precisely with the position of dorsal cephalic markers during stage 8, the cycle 14 mitotic domains 1, 3 and 5. The second phase of zen expression occurs in cells with peak PMad responses at the end of stage 5. Flanking cells with lower PMad levels correlate with the broader expression domain for the BMP target genes tailup and u-shaped. The full Medea response domain correlates approximately with the expression domain for u-shaped and extends into the presumptive dorsomedial ectoderm. The sharp transitions in SMAD response levels predict expression boundaries for BMP-responsive genes (Sutherland, 2003).

Similarly, in the wing primordium, a BMP gradient creates sharp transitions in PMad levels, which match gene expression boundaries. However, BMP activity is modulated by different mechanisms in this tissue. dpp is expressed in a narrow stripe at the center, and ligand spreads to nearby cells over a period of hours. In contrast, the early embryonic BMP activity gradient forms rapidly, and is narrower than the expression domains for dpp and scw. Extracellular binding proteins form the embryonic BMP activity gradient (Sutherland, 2003).

The final width of the midline peak response is sensitive to gene dosage for both dpp and sog. It is broader when dpp dosage is increased, and narrower with only one copy of dpp. Similarly, the width of the stripe is broader, but more variable, when sog levels are reduced. The response domain is broadest in sog null embryos; however, the level of response is significantly reduced. This is distinct from the effect of increased dpp dosage, in which the response domain is broader, but normal SMAD response levels are achieved or exceeded (Sutherland, 2003).

The role of Sog as both a short-range inhibitor and a long-range potentiator of dorsal patterning has led to a proposal that Sog transports BMP ligands from lateral regions to the dorsal midline. Biochemical analyses suggest mechanisms for Sog-BMP binding and release. Computational analysis have defined conditions under which transport could occur with these mechanisms. The transition from weak, broad SMAD responses to narrow, strong responses is consistent with concentration of BMP activity at the dorsal midline, and the loss of this transition with loss of Sog is consistent with a Sog-dependent transport model. However, there are significant differences between the current results and the assumptions used to develop the computational model. These include the presence of a midline SMAD response in dpp–/+ embryos and the sensitivity to reduced sog dosage. It will be important to refine future computational models to fit the complete set of BMP response data (Sutherland, 2003).

Both BMP ligands, Dpp and Scw, are required to form the dorsal-midline gradient. However, scw mutant embryos retain a small amount of dorsal ectoderm, with concomitant expansion of ventral ectoderm. Surprisingly, the weak dorsolateral Medea response is lost in scw embryos. It is concluded that the full Medea response domain encompasses the cell fates that are lost in scw mutants, amnioserosa and dorsomedial ectoderm. It appears that dorsal cells can acquire a dorsolateral fate without gastrula BMP activity (Sutherland, 2003).

Mutants with expanded ventral ectoderm show reduced SMAD responses during the first phase of BMP activity. PMad was not detected in blastoderm tld embryos. Homozygotes for moderate dpp alleles have lower PMad levels during blastoderm. Conversely, sog embryos have a slightly expanded PMad response during blastoderm, and a slight expansion of dorsal ectoderm. Thus, BMP activity during blastoderm positions the boundary between dorsal and ventral ectoderm (Sutherland, 2003).

Mutations that shift the boundary between amnioserosa and dorsal ectoderm show altered SMAD responses in the third phase of BMP activity, the dorsal-midline gradient. dpp-/+ embryos have variable reductions in midline SMAD responses and in the number of amnioserosa cells. Strikingly, sog null embryos have little amnioserosa and a strong reduction in SMAD response levels during gastrulation. Thus, SMAD response levels during gastrulation are critical for amnioserosa specification (Sutherland, 2003).

Taken together, these data suggest a multi-step model for DV patterning of the embryonic ectoderm, incorporating aspects of the two previous models. In a previous gradient model, ectodermal fates are subdivided simultaneously by a continuous BMP gradient involving Dpp and Scw. In the successive cell-fate decision model, amnioserosa is specified by dorsal-midline Dpp+Scw activity, and the dorsal ectoderm by Dpp alone at stage 9 (Sutherland, 2003).

Instead, it is proposed that the blastoderm phase of weak BMP activity establishes a dorsal ectoderm domain. Mutations that shift the boundary between dorsal and ventral ectoderm also have altered SMAD responses at this stage. It is at this stage that SMADs compete with Brinker to regulate the first phase of zen expression. Furthermore, this early signal maintains BMP activity, for the late-blastoderm domain of dpp expression is set by competition between BMPs, Sog and Brinker. BMP activity subsequently maintains the dorsal boundary for brinker expression. Thus, BMP activity at blastoderm defines a dorsal domain where dpp is expressed and brinker is not (Sutherland, 2003).

After cellularization is complete, a step gradient of BMP activity subdivides the dorsal region into amnioserosa, dorsomedial ectoderm and dorsolateral ectoderm. Peak activity levels determine the amount of amnioserosa. Flanking shoulders of weak activity specify the dorsomedial ectoderm. It is proposed that the dorsolateral ectoderm experiences a transient BMP response during late blastoderm, but little or no response during gastrulation. In sum, the dorsal-midline gradient of BMP activity specifies at least three cell fates (Sutherland, 2003).

BMP activity in the dorsal ectoderm does not end with germband extension. During stage 9, PMad is detected throughout the dorsal ectoderm and amnioserosa, and might finalize determination of dorsal ectoderm fates. Dpp expression within the dorsal ectoderm contributes to combinatoral regulation of gene expression patterns in subsets of dorsal ectodermal cells. However, the ventral boundary of dpp expression in the stage 9 dorsal ectoderm must be defined by earlier events (Sutherland, 2003).

The step gradient of SMAD responses is maintained during the morphogenetic movements of gastrulation and germband extension. The peak response is maintained only in cells that initially reside at the dorsal midline, even though ventral ectoderm moves to a dorsal position during stages 7 and 8. The BMP activity gradient is thought to form by diffusion in the perivitelline fluid; however, dorsal cells 'remember' their BMP exposure as they move laterally. It is probable that the ligand distribution is established prior to the time that peak SMAD responses are detected, and activity persists through cell biological mechanisms. For example, ligand may bind to the extracellular matrix, so that it remains associated with dorsal cells. Alternatively, receptor-ligand complexes may continue to signal following endocytosis. Understanding the intracellular modulation of BMP responses will be important to understand how extracellular morphogen gradients are translated into a stable pattern of cell fates (Sutherland, 2003).

Establishment of the leading edge during embryonic development

The leading edge (LE) is a single row of cells in the Drosophila embryonic epidermis that marks the boundary between two fields of cells: the amnioserosa and the dorsal ectoderm. LE cells play a crucial role in the morphogenetic process of dorsal closure and eventually form the dorsal midline of the embryo. Mutations that block LE differentiation result in a failure of dorsal closure and embryonic lethality. How LE cells are specified remains unclear. To explore whether LE cells are specified in response to early dorsoventral patterning information or whether they arise secondarily, the extent of amnioserosa and dorsal ectoderm was altered genetically, and LE cell fate was assayed. No expansion of LE fate is observed in dorsalized or ventralized mutants. Furthermore, the LE fate arises as a single row of cells, wherever amnioserosa tissue and dorsal epidermis are physically juxtaposed. Taken together these data indicate that LE formation is a secondary consequence of early zygotic dorsal patterning signals. In particular, proper LE specification requires the function of genes such as u-shaped and hindsight, which are direct transcriptional targets of the early Decapentaplegic/Screw patterning gradient, to establish a competency zone from which LE arises. It is proposed that subsequent inductive signaling between amnioserosa and dorsal ectoderm restricts the formation of LE to a single row of cells (Stronach, 2001).

Using mutations that influence DV patterning, it is possible to alter the size and distribution of BMP target gene expression patterns, which indicate the extent of amnioserosa and dorsal ectodermal cell fates. If LE fate was specified directly by a particular threshold level of BMP signal, then one would expect LE fate to be perturbed in concert with amnioserosa and dorsal ectoderm fates in DV mutants. Mutations in genes such as dorsal, Toll, brinker and short gastrulation alter the size of BMP target gene expression domains; however, these mutants failed to alter specification of LE fate. Among these genotypes, brk and sog specifically modulate the shape of the BMP signaling gradient in a region where LE fate might arise, yet LE formation in these mutants is fundamentally normal. Furthermore, in dorsalized embryos, LE cells were observed regularly at the boundary between amnioserosa and dorsal ectoderm even when the morphology of these tissues was severely disrupted. Islands of amnioserosa cells within a field of ectoderm were consistently surrounded with a single row of LE cells, independent of the number of amnioserosa cells constituting the island. The converse situation also occurred; again, a single row of LE cells formed at the boundary between the ectoderm and amnioserosa (Stronach, 2001).

DV mutants were also analyzed to determine whether a decrease in BMP signaling activity converts amnioserosa to LE as predicted by a gradient patterning model. A range of ventralizing mutations (cactus, sog, screw, dpp) displaying progressive loss of amnioserosa tissue did not give rise to embryos with an expanded domain of LE cells. In fact, LE cells were not detected in the absence of amnioserosa. No situation was found in which an altered BMP gradient was associated with expanded LE fate, thus the prediction of a direct gradient response model does not explain LE fate specification (Stronach, 2001).

Notably, DV mutant embryos that perturb the BMP gradient, also perturb the expression domains of target genes, including u-shaped and hindsight, without altering LE specification. However, loss of ush and hnt function results in specific and distinct perturbations in LE formation. Thus, the interpretation that LE fate specification is not a direct early response to the BMP gradient is favored, but rather is a secondary consequence of the specification of dorsal fates through the action of BMP target genes like ush and hnt (Stronach, 2001).

Taken together, these results raise the possibility that amnioserosa may be required for LE formation. To address the function of amnioserosa for LE specification, puc enhancer expression was examined in several mutants of the U-shaped class, including u-shaped (ush) and hindsight (hnt). Incidentally, the dorsal expression domains of these genes are directly regulated by DV patterning signals. In these mutant embryos, the amnioserosa tissue is fated normally and begins to differentiate up to stage 11, but then degenerates prematurely. In both ush and hnt mutants, programmed cell death takes place over the course of a few hours, with elimination of amnioserosa cells by stage 13 -- the time when dorsal closure would normally commence. Unexpectedly, different patterns of expression were observed with the puc enhancer in the two mutants. In ush embryos, ß-gal-positive cells were not detected. In contrast, hnt mutant embryos displayed Puc-positive LE cells at the edge of the dorsal ectoderm, albeit with less uniform expression than normally observed. To confirm these observations, the accumulation of dpp mRNA in the LE was examined. Similar to puc enhancer expression, differential expression of dpp was observed in ush versus hnt mutant embryos. ush mutant embryos show a consistent and significant reduction in LE dpp expression, although residual dpp transcripts are seen. dpp expression appears relatively normal in hnt mutant embryos (Stronach, 2001).

In addition to the differential expression of two LE markers in the U-shaped mutants, ectopic expression of LE markers is observed only in hnt mutant embryos. ß-gal-positive cells were observed in the region of the amnioserosa in hnt mutants as early as stage 11, raising the possibility that this could be an example of expanded LE cell fates. These cells adopt only partial LE cell fate, for the following reasons. These cells do not express the LE marker Fasciclin III, but do express two other LE molecules, albeit aberrantly. puc, for example, is expressed precociously in these cells, preceding Fasciclin III expression in the ectoderm, and dpp is rarely but reproducibly expressed. Additionally, cells in this region express amnioserosa fate markers such as race, through stage 11. Thus, based on the possibility that these cells may co-express LE and amnioserosa markers during stage 11, their identity cannot be unequivocally determined. These results may indicate that these cells are of mixed fate. The presence of ectopic LE-like cells in hnt mutant embryos, coupled with the severe reduction of LE fate markers in ush mutants, suggest that the distinction between amnioserosa and LE is a secondary consequence of Hnt and Ush functions, not a direct result of specific BMP signaling thresholds (Stronach, 2001).

If LE cells are specified as a secondary consequence of DV patterning gradients, then what additional mechanisms are at work to define LE as a single row of cells? The data are consistent with several mechanisms. One possibility is that specification of the LE involves the combinatorial action of nested sets of transcriptional regulators, including Hnt dorsally and Ush in a broader domain. Accordingly, loss of Hnt function is predicted to result in a failure to differentiate amnioserosa, coupled with dorsal expansion of more lateral fates, such as the LE. Consistent with this model, hnt mutant embryos display Puc-positive cells with partial LE character in the region of the dying amnioserosa during stage 11. These results suggest that Hnt may be necessary to distinguish amnioserosa from LE fate at the time of extended germ band stage. This timing is late, relative to the timing of the early BMP threshold response, further supporting the notion that LE specification is a secondary consequence of initial BMP signaling (Stronach, 2001).

Ush may play a role in differentiation of more lateral fates adjacent to the amnioserosa and the Hnt expression domain. Indeed, Ush function is essential for LE development because LE does not form in ush mutant embryos. Based on these results, it is imagined that Ush could define a competency zone from which LE cells arise, or Ush could participate in generating or modulating a signal(s) for communication between the differentiating amnioserosa and dorsal ectoderm. Ush is related to mammalian zinc-finger protein family, Friend of GATA (FOG), which has been shown to participate as a cofactor with GATA transcription factors. Together, these protein complexes regulate cell fate determination multiple times during both mammalian and Drosophila development. Interestingly, FOG2, a mammalian homolog of Ush, appears to be required during an inductive signaling event between two distinct tissues in the mouse heart, suggesting that inductive processes in development may commonly use the function of Ush family members. It has not been determined whether the function of Ush in LE cell specification is localized to the amnioserosa, the dorsal ectoderm, or both. Experiments to replace Ush function in a tissue-specific manner should address this issue (Stronach, 2001).

Although transcriptional targets of BMP signaling, such as ush and hnt, among others, define at least three specific threshold responses, the size difference between the nested expression domains of these markers still fails to account for a cell fate defined by a single row of cells. An additional mechanism to explain the spatially restricted stripe of LE cells is through an inductive signaling event. From the analysis of dorsalized mutants, it is observed that LE forms as a result of the juxtaposition of amnioserosa tissue with dorsal ectoderm, which may provide spatially limited activation of the JNK pathway. Thus, restricted expression of JNK target genes, such as puc and dpp may be a direct result of a signal that specifies LE (Stronach, 2001).

Communication between the amnioserosa and the dorsal ectoderm during embryogenesis has been suggested in two cases recently: (1) Hnt expression in the amnioserosa is required nonautonomously for proper cell rearrangements in the dorsal ectoderm, associated with retraction of the embryonic germband; (2) the raw gene product appears to be expressed in the amnioserosa, though it influences the activity of the JNK pathway in the ectoderm during dorsal closure. As amnioserosa and ectoderm develop, they may acquire different cell affinities, which cause them to sort into separate domains or islands (in the case of dorsalized embryos), displaying smooth borders at their interface. A difference in cell adhesion at the boundary may be sufficient to generate signaling for LE specification similar to inductive mechanisms at work at the compartmental boundaries of larval imaginal discs. The challenge now will be to identify molecules that may participate in an inductive signal (Stronach, 2001).

These results suggest that a multistep process determines the LE as a single row of cells. LE does not form directly in response to discrete intermediate levels of BMP signaling activity, but forms secondarily by the action of transcriptional regulators that are themselves BMP target genes. Among these targets, Hnt and Ush define a LE competency zone that is expanded in hnt mutants and eliminated in ush mutants. It is proposed that from within the competency zone, LE fate is further refined to a single row by an unknown inductive signal generated by the physical juxtaposition of amnioserosa with dorsal ectoderm. This signal activates the JNK pathway that regulates localized expression of dpp and puc (Stronach, 2001).

Shark (SH2 domain ankyrin repeat kinase) is a Drosophila nonreceptor tyrosine kinase that contains from amino to carboxyl terminus, a Src homology 2 (SH2) domain (N-SH2), five ankyrin repeats, a second SH2 domain (C-SH2), a proline-rich and basic region, and a tyrosine kinase domain. Analysis of the phenotypes associated with a shark loss-of-function mutation demonstrate that Shark activity is essential for the migration of the dorsolateral epidermis of the embryo during dorsal closure (DC). Shark kinase functions in DC upstream of Dpp expression by leading edge (LE) cells (Fernandez, 2000).

Because no obvious genetic interactions were obtained between Shark and mutations of the JNK pathway, tests were performed to see whether constitutive activation of the JNK or the Dpp pathway could rescue the shark1 DC phenotype. When shark1 GLCs were generated in the background of flies carrying a shark1 chromosome with an inserted transposon expressing an activated form of c-Jun (hs-SEjunAsp), shark1 DC defects were completely rescued, in some cases, as determined by the decreased penetrance of embryonic lethality (~10% lower than the fully penetrant 50% observed without the expression of hs-SEjunAsp) and by the complete or partial enclosure observed in unhatched embryos. These results are consistent with the action of Shark upstream of bsk (JNK) in the JNK pathway in LE cells (Fernandez, 2000).

Graded maternal short gastrulation protein contributes to embryonic dorsal-ventral patterning by delayed induction

Establishment of the dorsal–ventral (DV) axis of the Drosophila embryo depends on ventral activation of the maternal Toll pathway, which creates a gradient of the NFkappaB/c-rel-related transcription factor Dorsal. Signaling through the maternal BMP pathway also alters the dorsal gradient, probably by regulating degradation of the IkB homologue Cactus. The BMP4 homologue decapentaplegic (dpp) and the BMP antagonist short gastrulation (sog) are expressed by follicle cells during mid-oogenesis, but it is unknown how they affect embryonic patterning following fertilization. This study provides evidence that maternal Sog and Dpp proteins are secreted into the perivitelline space where they remain until early embryogenesis to modulate Cactus degradation, enabling their dual function in patterning the eggshell and embryo. Metalloproteases encoded by tolloid (tld) and tolkin (tok), which cleave Sog, are expressed by follicle cells and are required to generate DV asymmetry in the Dpp signal. Expression of tld and tok is ventrally restricted by the TGF-α ligand encoded by gurken, suggesting that signaling via the EGF receptor pathway may regulate embryonic patterning through two independent mechanisms: by restricting the expression of pipe and thereby activation of Toll signaling and by spatially regulating BMP activity (Carneiro, 2006).

This study has shown that sog, dpp, and tld act during oogenesis to promote the formation of dorsal anterior structures of the eggshell and to establish the embryonic DV axis. According to a proposed model, Sog is produced in follicle cells and is processed into different forms depending on DV location and stored in the perivitelline space. These forms of Sog then persist until early stages of embryogenesis at which time they act by a delayed induction mechanism to alter signaling mediated by maternally derived Dpp. It is proposed that an asymmetric distribution of Sog peptides is produced through the action of the ventrally localized Tld and Tok metalloproteases. Different forms of Sog act locally to inhibit Dpp signaling ventrally (e.g., N-Sog) or diffuse over considerable distances to concentrate Dpp dorsally (e.g., full-length Sog or C-Sog). According to this model, a dorsal-to-ventral gradient of Dpp activity is formed in the perivitelline space that counteracts and sharpens the inverse gradient of nuclear dorsal (Carneiro, 2006).

An important finding in this study is that Sog protein produced by follicle cells is secreted into the perivitelline space where it persists until the end of oogenesis and early embryogenesis, prior to initiation of zygotic sog expression. One way maternal Sog fragments might influence DV patterning in the embryo is to modify zygotic Dpp signaling. However, maternal Dpp signaling is involved in establishing the relative positions of the ventral mesoderm versus the lateral neuroectodermal territories, while zygotic Dpp activity determines the relative positions of dorsal and lateral domains. These distinct phenotypes suggest that maternal Sog acts by modulating the maternal rather than the zygotic component of Dpp signaling (Carneiro, 2006).

This analysis also suggests that the Dpp synthesized by follicle cells is secreted into the perivitelline space and stored there until advanced stages of oogenesis. These maternally synthesized Sog and Dpp proteins may act on the embryo following fertilization when signaling through the Toll pathway is initiated. Several lines of evidence support this hypothesis. (1) Through epistatic analysis, it was shown that maternal Dpp does not act upstream of the Toll receptor. Therefore, genes expressed in the follicle cell epithelium that regulate DV patterning exclusively via the Toll pathway should not be targets of maternal Dpp signaling. Alternatively, undescribed non-Toll mediators of DV patterning could potentially be targets of maternal Dpp in the follicular epithelium. (2) Blocking Tkv receptor function or reducing maternal Dpp activity (by 8xhssog, in follicle cells has no effect on the pattern of pip expression. It has been shown that maternal dpp does not alter grk expression. Thus, no evidence was found that the embryonic effects here described in this study are due to alterations in patterning of the follicular epithelium. (3) Maternal dpp signaling increases the levels of Cactus protein in the embryo by a mechanism that is independent of Toll. Finally, inhibition of Tkv with tkvDN expressed with an early maternal driver alters the embryonic expression domains of ventral and lateral genes such as vnd and snail, which are targets of dorsal activation but not of zygotic BMP signaling. tkvDN expression also alters expression of DV genes in lateralized embryos, which lack dorsal ectoderm and early zygotic dpp expression. In aggregate, these various data support the view that maternal dpp and sog act by delayed induction on the embryo itself. The possibility cannot be ruled out, however, that the embryonic DV phenotypes described in this study result from the combined effects of direct and indirect maternal dpp signaling with the predominant effect being direct (Carneiro, 2006).

Delayed inductive activities have been proposed for a variety of proteins synthesized during oogenesis. For example, activation of the terminal system relies on delayed inductive activity of the secreted product of the torsolike gene (tsl), which is expressed by follicle cells at the two poles of the oocyte and associates with the vitelline membrane. ndl has a dual action on chorion integrity and embryonic patterning. The embryonic patterning function of ndl is thought to be mediated by Nudel protein that is secreted into the perivitelline space where it associates with the embryonic plasma membrane and initiates a proteolytic cascade. It is proposed that Sog and Dpp secreted by follicle cells also serve two roles. First, they contribute to patterning the follicle cell epithelium and chorion, and secondly, they are transferred to and stored in the perivitelline space where it is proposed that they function after fertilization to modify Toll patterning in the embryo (Carneiro, 2006).

During embryogenesis, Sog protein diffuses dorsally from the neuroectoderm and may carry Dpp dorsally in a complex with Tld, Tsg, and Scw, resulting in the generation of peak Dpp activity in the dorsal midline. The spatial distribution of maternal Sog, Dpp, Tld and Tok during oogenesis could also create asymmetric BMP activity. Since tld and tok are expressed only in ventral follicle cells, a ventral-to-dorsal gradient of Sog fragments is likely to be produced. Because cleavage of Sog by Drosophila Tld and Tok is dependent on the amount of Dpp, cleavage of Sog by Tld and Tok should be increased near the source of Dpp, generating an oblique gradient of Sog fragments in the egg chamber. The existence of such a gradient is supported by the greater staining seen in anterior ventral cells with the anti-Sog 8A antiserum during stage 10B. However, greater asymmetry may exist as a result of differential distribution of an array of Sog fragments throughout the egg chamber. Unfortunately, visualization of such asymmetry would be hard to achieve due to limitations in the ability to recognize several fragments by existing Sog antisera (Carneiro, 2006).

The analysis of marked sog− and tld− follicle cell clones suggests that the mobility of Sog fragments in the extracellular compartment may contribute to creating a maternal Dpp activity gradient. Such clones resulted in different Sog staining patterns in the perivitelline space adjacent to the clones depending on where they were located along the DV axis. The staining pattern observed with the 8A antibody suggests that ventrally generated N-Sog cleavage products may be less diffusible than intact Sog or than C-Sog and remain restricted to their site of production. In contrast, full-length Sog and C-Sog fragments appear to diffuse more readily (Carneiro, 2006).

Diffusion of Dpp may also contribute to patterning the eggshell. The expression of dpp in anterior follicle cells is consistent with its role in the formation of dorsal anterior chorionic structures. An anterior-to-posterior gradient of Dpp activity in dorsal regions of the egg chamber is suggested by the Dpp-dependent activation of the A359 enhancer trap and graded repression of bunched along the AP axis. In addition, BR-C expression is lost in mad− clones away from the source of Dpp. sog is likely to contribute to establishing this BMP gradient since ventral sog−clones act non-cell-autonomously to decrease the size of the operculum. Since ventral tld− clones also alter the extent and angle of the operculum, Tld may process Sog to generate a fragment that diffuses and carries Dpp to a dorsal anterior location, concentrating and thus enhancing Dpp activity. Further evidence that a fragment with such activity exists derives from the observation that overexpression of a C-terminal Sog fragment generates chorionic phenotypes that strongly resemble dpp overexpression (Carneiro, 2006).

A dorsally produced form of Sog also appears to participate in patterning the eggshell since sog− clones located dorsally result in fusion of dorsal appendages along the dorsal midline. DV positioning of the dorsal appendages depends on several factors, most critically on EGFR signaling. In contrast, mild overexpression of dpp generates fusion of the dorsal appendages. Considering the well-established role of Sog in modulating Dpp activity, the fused appendage phenotype generated by dorsal sog− clones most likely reflects the loss of Dpp antagonism exerted by Sog (Carneiro, 2006).

In addition to the activities described above, N-Sog fragments which remain ventrally restricted could exert Supersog-like activity, antagonizing BMPs while acquiring resistance to further cleavage and degradation by Tld. This ventrally restricted activity most likely patterns the embryo but does not affect dorsal positioning of eggshell structures, which depends on the combined activity of Dpp/BMPR signaling and dorsally generated Grk/EGFR signals (Carneiro, 2006).

The assortment of Sog fragments in egg chambers is very similar to that in the embryo. Full-length and processed forms of Sog generated by Tld during oogenesis might remain asymmetrically distributed during embryogenesis and exert distinct activities. This hypothesis is in agreement with the effect of tld− and sog− follicle cell clones on the embryo. In the majority of cases, tld− follicle cell clones result in ventralized cuticles, indicating that Tld generates some activity that synergizes with Dpp. Reciprocally, the great majority of sog− follicle cell clones result in dorsalized cuticles and embryos, indicating that Sog primarily acts by antagonizing Dpp. Since only ventral sog− clones generate cuticle defects, ventrally produced Sog presumably generates a ventralizing activity that blocks Dpp locally. In contrast, since in a minority of cases ventral shifts are observed in embryonic gene expression domains resulting from sog− clones, as well as a minority of dorsalized cuticles from tld− clones, there may also be a form of Sog that can enhance Dpp signaling. This positive BMP promoting activity could be generated ventrally, as suggested above in the case of chorion patterning (Carneiro, 2006).

A model depicting the proposed effects of different Sog forms on formation of the chorion and embryonic patterning is presented. According to this model, ventrally restricted Tld cleaves Sog near the Dpp source in ventral anterior follicle cells generating N-Sog and C-Sog. It is suggested that N-Sog fragments remain restricted near ventral anterior cells to antagonize Dpp, while C-Sog fragments diffuse dorsally concentrating Dpp in dorsal anterior cells that direct formation of the operculum. This asymmetric production of Sog molecules would generate a dorsal-to-ventral gradient of Dpp, with the highest levels dorsally near the anterior Dpp source. Although direct visualization of the predicted resulting Dpp gradient in the embryo is hard to achieve with the tools available, it is proposed that such a similarly oriented gradient persists until early embryogenesis based on the asymmetric pattern of Dpp-GFP distribution during late oogenesis and the observed alterations in embryonic gene expression domains resulting from modifications in maternal Dpp signaling (Carneiro, 2006).

The slope of the Dl nuclear gradient ultimately defines the extent of the mesoderm (Mes), neuroectoderm (NE), and dorsal ectoderm (DE). A uniform increase or decrease in nuclear Dl along the DV axis can only alter the extent of the Mes and DE and positioning of the NE, while a change in the slope of the gradient will modify the extent of NE territories such as the vnd expression domain. Under all conditions that Dpp signaling was altered, modifications were observed in the width of the vnd domain. This suggests that graded maternal Dpp signaling helps determine the slope of the dorsal gradient. Earlier studies suggested that Dpp inhibits Cactus degradation and as a consequence decreases Dl translocation into the nucleus. Increased Dpp signaling should result in more Dl retained in the cytoplasm, with consequent narrowing of the mesoderm and ventral shift in lateral and dorsal expression domains. Conversely, inhibition of Dpp signaling would result in increased levels of Dl becoming available for nuclear translocation. Considering the proposal that maternal Dpp is highest dorsally, and that Cactus may also act to prevent Dl diffusion along the DV axis, decreasing Dpp should lower Cactus levels in dorsal–lateral regions of the embryo and result in the redistribution of free Dl from ventral to lateral regions. As a consequence of this redistribution of Dl, there would be a slight decrease in Dl levels ventrally and an increase laterally that would have the net effect of flattening the gradient. Such a mechanism would require a certain degree of mobility of dorsal dimers in the syncytial blastoderm. In future studies, it will be interesting to determine the relative mobilities of Dl/Cactus complexes in the cytoplasm (Carneiro, 2006).

Maternal BMP signaling may also increase the robustness of dorsal patterning. The prevailing view of DV patterning is that signaling through the Toll pathway is sufficient to generate threshold-dependent activation of several dorsal target genes along the entire DV axis. Activation of Toll triggered by the ON/OFF pip expression pattern must be transformed into a ventrally centered gradient of Toll signaling. Several mechanisms may contribute to generate this gradient, based on autoregulatory feedback mechanisms. Although the Toll system may be internally robust, regulatory inputs from other signaling pathways could also contribute further to its stability, such as suggested for the wntD pathway and for maternal Dpp. While a significant body of evidence supports the standard view that establishment of the dorsal gradient through the Toll pathway is central to DV axis specification, the maternal Dpp pathway may constitute an important secondary mechanism that sharpens and ensures robustness and stability of the dorsal gradient in response to a rapidly changing embryonic environment (Carneiro, 2006).

The initiating event in maternal DV patterning is localized activation of the Grk/EGFR pathway in dorsal cells. Grk functions by restricting the expression of both pip and tld/tok, providing two potentially independent means for spatially regulating the activity of Toll and Dpp. This dual action of the Grk/EGFR pathway is consistent with analysis in which it was found that embryonic cuticles from gd−; grk−; Tl[3] mothers displayed a phenotype distinct from those collected from gd−; Tl[3] mothers. While cuticles from both genotypes had denticle belts surrounding the entire circumference of the embryo, cuticles from gd−; grk−; Tl[3] mothers were more elongated than those from gd−; Tl[3] mothers and exhibited a more ventral character. This suggests that grk provides an additional signal for asymmetry downstream or in parallel to gd. It is suggested that the hypothetical system proposed acts downstream of grk/EGFR and in parallel to Toll may be the Dpp pathway (Carneiro, 2006).

Advances in image acquisition and informatics technology have led to organism-scale spatiotemporal atlases of gene expression and protein distributions. To maximize the utility of this information for the study of developmental processes, a new generation of mathematical models is needed for discovery and hypothesis testing. A data-driven, geometrically accurate model has been developed of early Drosophila embryonic bone morphogenetic protein (BMP)-mediated patterning. Nine different mechanisms for signal transduction with feedback, eight combinations of geometry and gene expression prepatterns, and two scale-invariance mechanisms were tested for their ability to reproduce proper BMP signaling output in wild-type and mutant embryos. It was found that a model based on positive feedback of a secreted BMP-binding protein, coupled with the experimentally measured embryo geometry, provides the best agreement with population mean image data. The results demonstrate that using bioimages to build and optimize a three-dimensional model provides significant insights into mechanisms that guide tissue patterning (Umulis, 2010).

In many systems, spatially patterned cellular differentiation is regulated by signaling molecules called morphogens, which initiate spatiotemporal patterns of gene expression in a concentration-dependant manner. In early Drosophila embryos, a morphogen composed of a heterodimer of Decapentaplegic (Dpp) and Screw (Scw), two members of the bone morphogenetic protein (BMP) family. Unlike classical morphogen systems that rely on the slow spreading of a molecule from a localized source to establish a gradient, BMPs in the early Drosophila embryo are secreted from a broad region making up the dorsal-most 40% of the embryo circumference. Subsequently, they are dynamically concentrated into a narrow region centered about the dorsal midline that makes up only 10% of the embryo circumference (Umulis, 2010).

A number of extracellular regulators contribute to the dynamics and localization of BMP signaling. Laterally secreted Short gastrulation (Sog) and dorsally secreted Twisted gastrulation (Tsg) diffuse from their regions of expression and form a heterodimer inhibitor (Sog/Tsg) that binds to Dpp-Scw, preventing it from binding to receptors. The cell matrix may mediate the formation of this complex, as it has recently been shown that collagen can bind both BMPs and Sog, thereby facilitating their association (Wang, 2008). The extracellular binding reactions lead to a gradient of inhibitor-bound Dpp-Scw that is high laterally and low at the dorsal midline, and an opposing gradient of free Dpp-Scw that is high at the dorsal midline. The dorsally secreted metalloprotease Tolloid (Tld) processes Sog only when Sog is bound to BMP ligands, and the degradation of Sog by Tld further enhances both the gradient of inhibitor-bound Dpp-Scw and of free Dpp-Scw. Thus, extracellular Dpp-Scw is redistributed by a combination of binding to inhibitor, processing of this complex, and diffusion (Umulis, 2010).

Simultaneously, receptors and other surface-localized binding proteins compete with Sog to bind the available Dpp-Scw. Dpp-Scw activates signaling by binding to and recruiting the Drosophila type I receptors, Thickveins (Tkv) and Saxophone (Sax), into a high-order complex containing two subunits of the type II receptor Punt. The receptor complex phosphorylates Mad (pMad), a member of the Smad family of signal transducers, and phosphorlyated Mad binds to the co-Smad Medea, forming a complex that then accumulates in the nucleus, where it regulates gene expression in a concentration-dependent manner (Umulis, 2010).

Although complex formation and transport favor a net movement of ligand toward the dorsal midline of the embryo, positive feedback in response to pMad signaling is needed to further concentrate the surface-localized Dpp-Scw at the dorsal midline. A loss of extracellular BMP regulators or positive feedback impedes the attenuation of pMad laterally as well as the accumulation of pMad signaling at the dorsal midline. Although feedback, extracellular transport, and signal transduction each provide a specific mode of Dpp-Scw signal regulation, it is the dynamic interaction of these regulatory mechanisms that patterns the dorsal surface of Drosophila embryos. Not only does the mechanism work under optimal laboratory conditions, but dorsal surface patterning appears to be remarkably resilient to nonideal conditions such as temperature fluctuations, reductions in the level of regulatory factors such as Tsg, ectopic gene expression, and other perturbations. These issues illustrate the complexity of the problem and suggest that it is not possilbe to rely solely on genetic and biochemical data to fully explain this rather simple patterning problem (Umulis, 2010).

To address a number of unanswered questions about Dpp-Scw-mediated patterning and to take full advantage of the available data on Drosophila development, a methodology was developed that seamlessly integrates biological information in the form of prepatterns, geometry, mechanisms, and training data into an organism-scale model of the blastoderm embryo that is based on a reaction-diffusion description of patterning. The mathematical model is simulated by using the widely available computational frameworks Comsol and Matlab, which makes extensive use of the model and methodology feasible (Umulis, 2010).

An image analysis protocol was developed to obtain model training and initial condition data and to calculate population statistics for patterns of pMad signaling in wild-type (wt) and mutant D. melanogaster. Both the mean and variability of pMad signaling along the dorsal-ventral (DV) axis depends on anterior-posterior (AP) position and the specific choice of threshold. Using mutations previously considered robust, differences could be detected between mutant and wild-type pMad signaling patterns, which provided an information-rich data set for model training and for testing the contributions of diverse positive-feedback mechanisms and of proteins that concentrate BMPs at the cell surface. Unexpectedly, it was found that geometry also has a large impact on the predicted patterns of BMP-bound receptors, whereas the prepatterned expression of receptors and other modulators of signaling did not greatly affect model-data correspondence. It was found that if the embryo geometry is perturbed slightly in the model, then including the prepattern information greatly enhanced the model's ability to fit the observed pMad patterns, which suggests that the prepatterns may mitigate the effects of slightly misshapen embryos. Conditions in the model were identified that improve the scale invariance of patterning and tested the model predictions by staining for pMad in different species of Drosophila. These studies demonstrate that building a model based on image data and training the three-dimensional (3D) model against multidimensional expression data provide insights into the properties of several important developmental principles, including positive feedback, biological robustness, and scale invariance (Umulis, 2010).

Response to the BMP gradient requires highly combinatorial inputs from multiple patterning systems in the Drosophila embryo

Pattern formation in the developing embryo relies on key regulatory molecules, many of which are distributed in concentration gradients. For example, a gradient of BMP specifies cell fates along the dorsoventral axis in species ranging from flies to mammals. In Drosophila, a gradient of the BMP molecule Dpp gives rise to nested domains of target gene expression in the dorsal region of the embryo; however, the mechanisms underlying the differential response are not well understood, partly owing to an insufficient number of well-studied targets. This study analyzed how the Dpp gradient regulates expression of pannier (pnr), a candidate low-level Dpp target gene. It was predicted that the pnr enhancer would contain high-affinity binding sites for the Dpp effector Smad transcription factors, which would be occupied in the presence of low-level Dpp. Unexpectedly, the affinity of Smad sites in the pnr enhancer was similar to those in the Race enhancer, a high-level Dpp target gene, suggesting that the affinity threshold mechanism plays a minimal role in the regulation of pnr. The results indicate that a mechanism involving a conserved bipartite motif that is predicted to bind a homeodomain factor in addition to Smads and the Brinker repressor, establishes the pnr expression domain. Furthermore, the pnr enhancer has a highly complex structure that integrates cues not only from the dorsoventral axis, but also from the anteroposterior and terminal patterning systems in the blastoderm embryo (Liang, 2012).

Most blastoderm genes are regulated primarily on either the DV or AP axis. For example, the gap genes are expressed in one or two domains of expression along the AP axis and, although some of them may exhibit regulation along the DV axis, they are nonetheless considered AP genes. pnr represents an interesting case because although it was originally reported as a DV gene, closer inspection of its expression pattern in wild-type and mutant embryos and detailed dissection of its cis-regulatory enhancers revealed that pnr is highly regulated by both AP and DV genes. Its pattern is a composite of two superimposed patterns that each exhibit AP and DV spatial regulation: a dorsal patch and six AP stripes, which are limited to the dorsal 30% of the embryo. The patch domain, but not the stripes, disappeared in dpp mutants, whereas both the patch and stripes expand ventrally in the absence of Brk. The stripes are more sensitive to Brk repression because activation of the patch domain is limited to the region where Dpp is present dorsally, whereas the stripes can be activated along the entire DV axis. Brk in the ventrolateral region and Sna in the ventral-most region repress stripe expression. Since pnr specifies dorsomedial fates, restricting its expression to the dorsal 30% of the circumference is crucial. Ectopic expression of pnr ventrally causes transformation of ventral epidermis into dorsomedial epidermis (Liang, 2012).

Competition between Brk and Smads for binding to overlapping DNA sequences is likely to set the border of the patch domain. Two Smad sites are particularly important for patch expression, and one of these, the M3 site, is a composite site that binds both Brk and Smads, raising the possibility that the patch border is established by competition between activating inputs from Smads in the dorsal region and repressive inputs from Brk emanating from the ventral region. Competition between Brk and Smads for overlapping binding sites has been observed for several Dpp target enhancers (Liang, 2012).

Repression of the AP stripes ventrally requires both Brk sites B1 and B2. The two posterior stripes driven by P3 expand to a lesser degree than the four anterior stripes driven by P4. This can be explained by the fact that P4 lacks Brk site B1, which is a stronger Brk site. Loss of both Brk sites would likely result in expansion to the edge of the mesoderm, as seen in embryos that lack Brk protein. Repression by Sna is likely to involve the Sna binding sites in the pnr enhancer, as genome-wide binding studies have shown that the pnr enhancer is bound by Sna (Liang, 2012).

The positioning of the stripes, as well as of the patch, along the AP axis is regulated by the gap genes. The results suggest that Hb, Gt and Tll set the anterior edge of the pnr domain, whereas Tll sets the posterior, and that direct and indirect interactions among the gap proteins establish the stripe borders relative to one another, as has been observed for eve. For example, the broad central stripe seen in kni- could be explained by the lack of direct Kni repression. However, owing to the complex cross-regulatory interactions among the gap genes, it is difficult to predict which gap proteins regulate the pnr stripes directly, although genome-wide binding data of the gap factors support their direct binding to the pnr enhancer. Although Bcd does not appear to bind directly to the pnr enhancer, its effects are mediated through its targets Gt and Hb (Liang, 2012).

In depth studies of three genes with different boundary positions in the dorsal region, Race, C15 and pnr, indicate that complex combinatorial mechanisms are employed to establish their expression domains, with each gene having a unique regulatory network of its own. Although they all respond to Dpp signaling, their borders of expression are not set by a simple threshold response to the Dpp gradient that depends on differential binding site affinity (Liang, 2012).

The feature that has been shown to be important for high-level Dpp target expression is the feed-forward motif involving Dpp and Zen. High levels of Dpp/Smads first activate zen expression in the dorsal-most region, the presumptive amnioserosa, and then both Zen and Smads bind and activate the Race enhancer. The intermediate-level target C15 has a different enhancer structure than high-level targets, containing many Smad sites that act in a cumulative manner to drive expression in regions of intermediate Dpp levels. Mutation analysis has shown that the number of intact Smad binding sites, rather than their affinity, is important for the C15 response. Nevertheless, the enhancer structure of C15 might promote high levels of Smad binding in vivo, and this may increase the response to Dpp. Do all intermediate-level Dpp targets have a similar enhancer structure? The enhancer that drives expression of the intermediate-level Dpp target gene tup was examined for putative Smad binding sites (SBEs and GC-rich regions), and observed multiple Smad sites across the enhancer, similar to that seen with C15. Thus, the multiple Smad site signature might be necessary for response to lower than peak levels of Dpp. In addition, intermediate-leveltargets may utilize repression mechanisms to help establish their borders of expression, as was shown for C15 (Liang, 2012).

These studies have revealed that the pnr enhancer resembles that of a high-level target in Smad site organization and Smad binding site affinity. In fact, it was surprisingly easy to convert the low-level target enhancer into a high-level target by mutating a single Smad site. This result could be easily explained if the M3 site had a higher affinity for Smads than those in Race; however, comparison of the binding sites by gel shift showed they have similar affinities. Furthermore, replacing the M3 Smad site with a Race Smad site had little effect on the expression pattern. These results suggest that activation of pnr in its broad domain has little to do with Smad binding affinity. How then does pnr respond to low levels of Dpp? One possible mechanism involves the highly conserved AGCAATTAA site that lies adjacent to the Smad sites. In the absence of this site, the P3 enhancer could not respond to low-level Dpp. It is possible that this site, when bound, leads to greater Smad binding, which would then promote pnr activation (Liang, 2012).

What factor(s) might bind to the AGCAATTAA site? ATTA is the core binding site for Antp class HD proteins. Although Zen binds to the ATTA site in vitro, neither the endogenous pnr pattern nor P3-lacZ expression is significantly affected in zen mutants. To identify candidate factors, the TOMTOM tool at FlyFactor Survey was used, and the best match was to the HD protein Hmx, which binds CAATTAA. However, Drosophila Hmx is expressed only in an anterior region that does not overlap with pnr (see FlyBase). Likewise, although several Antp class HD proteins were predicted to bind to the ATTA core sequence, their timing or domains of expression do not overlap ideally with those of pnr (Liang, 2012).

It has been proposed that the AGCAATTAA site in the Msx2 enhancer might bind a factor in addition to an HD protein via the 5' half of the site, perhaps a transcriptional partner such as FAST1, which was previously shown to function with Smads. Although the search did not reveal any candidates, if this is the case for pnr then the bipartite motif could potentially bind four proteins: Smads, Brk, HD and 'partner X', The combination of these proteins in a given cell along the DV axis would determine pnr transcriptional activity. The fact that the bipartite motif is not present in the enhancers of Race or C15, or in the other pnr enhancers identified, demonstrates the versatility of how Dpp uses different partners to establish multiple target gene domains (Liang, 2012).

Is the structure of the pnr enhancer typical for low-level Dpp targets? This is difficult to address owing to the lack of candidate low-level Dpp targets. Brk is considered a low-level Dpp target in imaginal disc development; however, Dpp represses brk, giving rise to a reciprocal gradient of the Brk repressor. Target gene borders are thus established by competition between Smad and Brk for overlapping binding sites, as mentioned above for pnr. The brk enhancer contains multiple enhancer/silencer modules consisting of activator and repressor (Mad/Medea/Schnurri) binding sites, which contribute to threshold responses to the Dpp gradient, and thus it does not resemble the pnr enhancer. Although good progress has been made in understanding how pnr is expressed in regions with low levels of Dpp, learning the general rules that control broad dorsal patterns will require the analysis of more enhancer elements (Liang, 2012).

What rules do target genes for other morphogens follow? Long before the 'feed-forward' term was it was shown that both the Dl and Bcd morphogens interact with their high-level targets, Twi and Hb, respectively, to activate downstream; thus, combinatorial motifs are generally utilized. Moreover, as more target genes of Dl and Bcd were identified and studied, it became apparent that the affinity threshold model could not explain all cases of differential response to the gradient. For example, analysis of several enhancers that drive Bcd-dependent expression in anterior regions of the embryo revealed a poor correlation between Bcd binding site affinity and the AP limits of the pattern. Also, although Dl targets remain archetypal examples of genes that utilize the affinity threshold mechanism, it was found that genes expressed in the lateral region also require input from the Zelda (Vielfaltig - FlyBase) transcription factor for expression in regions of low-level Dl. Zelda binding sites are present in target enhancers, and it was proposed that Zelda boosts Dl binding to help activate the neuroectodermal genes (Liang, 2012).

Downstream target gene interactions also shape domains of expression, in particular cross-

repression among the targets. In both the Drosophila neuroectoderm and the vertebrate neural tube, morphogen targets are expressed in discrete domains rather than nested overlapping domains due to the repression of one target by another. This mechanism establishes sharp boundaries among the target genes (Liang, 2012).

Thus, it is clear that additional factors help morphogens set threshold responses. Given that the pnr enhancer could potentially interact with four different factors along the DV axis and at least four factors along the AP axis, several combinations of inputs could regulate other Dpp target genes. More generally, depending on the number of different factors that interact with the cis-regulatory regions of target genes, morphogen gradients could elicit multiple threshold responses, as has been seen for morphogens such as Dl in Drosophila, Activin in the Xenopus blastula and Shh in the vertebrate neural tube, where up to seven threshold responses have been described. Only by dissecting enhancers can it be fully understood how target genes integrate diverse inputs (Liang, 2012).

Dpp and the abdomen

The adult abdominal epidermis develops during the pupal stage from groups of cells called histoblast nests, which differ from imaginal discs in two important respects: (1) abdominal histoblasts do not invaginate during embryogenesis, but remain part of the larval epidermis and secrete larval cuticle, and (2) they do not proliferate during the larval stages. After pupariation, histoblasts multiply rapidly and migrate to replace the polyploid larval epidermal cells (LEC). As the individual nests grow and merge, LEC are destroyed only upon contact with histoblasts, so that the continuity of the pupal epidermis is maintained at all times. The replacement of LEC by histoblasts is completed by 40-42 hours after puparium formation (APF). The epidermis of each abdominal segment is produced by three bilateral pairs of histoblast nests: the anterior dorsal nests produce the tergite; the posterior dorsal nests form the flexible intertergal cuticle, and the ventral nests produce the sternite and pleura. In addition, a spiracular nest produces a small patch of epidermis around each spiracle. Dorsally, each segment is composed of a sclerotized, pigmented tergite and flexible, unpigmented intertergal cuticle that is normally folded underneath the tergite. All cells in the sternites, pleura and tergites secrete 3-4 trichomes per cell. The wide-based, curved trichomes secreted by pleural cells are distinct from the thinner, straighter sternal and tergal trichomes. In addition, sternites and tergites, but not the pleura, contain arrays of bristles. Dorsoventral patterning is also present within tergites, since the dark pigment band at the posterior edge of each tergite is wider medially than laterally. Some segments deviate from the typical pattern. For example, the first abdominal segment (A1) lacks a sternite. Also, in the male, A7 lacks both a sternite and a tergite, A6 lacks bristles on its sternite, and A5 and A6 have uniformly darkly pigmented tergites (Kopp, 1999 and references).

The adult abdominal epidermis is formed during the first 40-42 hours of pupal development. At pupariation, the abdominal epidermis is composed predominantly of the polyploid LEC, which are easily distinguishable from the much smaller, diploid histoblasts. At this stage, the anterior dorsal histoblast nest (aDHN) contains 13-19 cells; the posterior dorsal nest (pDHN) contains 5-8 cells, and the ventral nest (VHN) contains 9-13 cells. Histoblasts begin to proliferate and migrate to supplant the LEC soon after pupariation. At 18-20 hours APF, the aDHN and pDHN merge to form a single dorsal histoblast nest (DHN). The DHN merges with the VHN and the spiracular anlage between 20 and 28 hours APF. The spiracle, located at the lateral midline, marks the boundary between ventral and dorsal histoblasts, and eventually the boundary between the pleura and the tergite. The fusion of histoblast nests of consecutive segments begins at 28 hours APF and proceeds until 40-42 hours APF, when the formation of a continuous adult epidermis is completed by the fusion of contralateral nests at the dorsal and ventral midlines. Morphological differentiation of the epidermis into sternite, tergite and pleural territories becomes evident shortly thereafter. These regions can be distinguished at 45 hours APF by differences in the shape and arrangement of cells and by the pattern of developing adult muscles (Kopp, 1999).

The origin of the adult wg and dpp patterns can be traced to the early pupal stage. At the time of fusion of the aDHN and pDHN (18 hours APF), wg and dpp expression domains encompass both adult and larval cells, and are limited to the posterior region of the anterior compartment. Within this zone, the patterns of wg and dpp are largely complementary along the DV axis. wg is expressed in a dorsal-posterior sector of the aDHN and the adjacent dorso-lateral LEC, as well as in a ventral sector of the VHN and the adjacent ventro-lateral LEC. wg is not expressed in the dorsal part of the VHN, in the ventral part of the aDHN or in the lateral LEC between the two nests. wg expression is also weak or absent in the LEC near the dorsal and ventral midlines. dpp is expressed in a dorsal sector of the VHN and in a few cells at the dorsal margin of the aDHN. dpp expression is also seen in the lateral LEC between the VHN and aDHN, and in the dorsal LEC between contralateral dorsal nests. The pupal expression of wg evolves from the pattern present in the larva, where wg is expressed in a circumferential stripe along the AP compartment boundary. The early pupal pattern develops by gradual elimination of expression at the ventral, dorsal and lateral midlines. dpp is not expressed in the epidermis of third instar larvae. However, the expression of dpp in the pupa is reminiscent of the embryo, where it is expressed in mid-dorsal and ventro-lateral stripes. Thus, the pupal expression of dpp may reflect some memory of this embryonic pattern (Kopp, 1999 and references).

The patterns of wg and dpp expression established by 18 hours APF are maintained during the subsequent growth of the histoblast nests. At the time of fusion of the VHN and DHN (24-28 hours APF), wg expression is seen in sectors in the ventral third of the VHN and in the dorsal half of the DHN. dpp is expressed in a stripe in the dorsal two-thirds of the VHN and in a group of 30-40 cells at the dorsal DHN margin. dpp expression also extends transiently into the ventral DHN margin; this expression lasts for only a few hours, and encompasses about 15 cells at its peak (Kopp, 1999).

The complementary expression patterns of wg and dpp are retained in the newly formed adult epidermis at 40-42 hours APF. dpp is expressed in a transverse stripe in the presumptive pleura and in a wedge-shaped stripe along the dorsal midline of the tergite. The limits of pleural expression of dpp coincide precisely with the sternite-pleura and tergite-pleura boundaries. wg is expressed in the sternite and in the medial tergite, but is excluded from the dorsal midline. Neither gene is expressed in a large lateral region of the tergite. The expression of wg and dpp remains restricted to the posterior region of the anterior compartment, with sharply defined posterior and graded anterior boundaries. Double labelling with Engrailed shows that the posterior limit of dpp expression coincides with the compartment boundary. Based on morphological landmarks and on the pattern of lacZ expression in wg-lacZ/dpp-lacZ pupae, the same appears to be true for wg (Kopp, 1999).

The division into dorsal tergite, ventral sternite and ventro-lateral pleural cuticle is largely specified during the pupal stage by Wingless, Decapentaplegic and Egf receptor signaling. Expression of wg and dpp is activated at the posterior edge of the anterior compartment by Hedgehog signaling. Within this region, wg and dpp are expressed in domains that are mutually exclusive along the dorso-ventral axis: wg is expressed in the sternite and medio-lateral tergite, whereas dpp expression is confined to the pleura and the dorsal midline. Neither gene is expressed in the lateral tergite. Tergite and sternite cell fates are specified by Wg signaling. Egfr acts synergistically with Wg to promote tergite and sternite identities, and Wg and Egfr activities are opposed by Dpp signaling, which promotes pleural identity. Wg and Dpp interact antagonistically at two levels:(1) their expression is confined to complementary domains by mutual transcriptional repression and (2) Wg and Dpp compete directly with one another by exerting opposite effects on cell fate. Egfr signaling does not affect the expression of wg or dpp, indicating that it interacts with Wg and Dpp at the level of cell fate determination. Within the tergite, the requirements for Wg and Egfr function are roughly complementary: Wg is required mainly in the medial region, whereas Egfr is most important laterally. Dpp signaling at the dorsal midline controls dorso-ventral patterning within the tergite by promoting pigmentation in the medial region (Kopp, 1999).

The major conclusion of this report is that much of the dorso-ventral (DV) patterning of the adult abdomen is determined by antagonistic interactions between Dpp, which specifies pleural cell fate, and Wg and Egfr signaling, which together specify tergite and sternite fates. Expression of wg and dpp is activated at the posterior edge of the anterior compartment by Hh signaling. Within this zone, wg and dpp are expressed in complementary patterns along the DV axis: wg is expressed in the presumptive sternite and in the medio-lateral region of the tergite, whereas dpp is expressed in the presumptive pleura and at the dorsal midline of the tergite. Although the pattern of Egfr activation in the abdomen has not been determined, Egfr signaling is most important in the lateral tergite, a region where neither wg nor dpp are expressed (Kopp, 1999).

Dpp signaling and the dorsolateral peripheral nervous system

Transforming growth factor ß signaling mediated by Decapentaplegic and Screw is known to be involved in defining the border of the ventral neurogenic region in the fruitfly. A second phase of Decapentaplegic signaling occurs in a broad dorsal ectodermal region. The dorsolateral peripheral nervous system forms within the region where this second phase of signaling occurs. Decapentaplegic activity is required for development of many of the dorsal and lateral peripheral nervous system neurons. Double mutant analysis of the Decapentaplegic signaling mediator Schnurri and the inhibitor Brinker indicates that formation of these neurons requires Decapentaplegic signaling, and their absence in the mutant is mediated by a counteracting repression by Brinker. Interestingly, the ventral peripheral neurons that form outside the Decapentaplegic signaling domain depend on Brinker to develop. The role of Decapentaplegic signaling on dorsal and lateral peripheral neurons is strikingly similar to the known role of Transforming growth factor ß signaling in specifying dorsal cell fates of the lateral (later dorsal) nervous system in chordates (Halocythia, zebrafish, Xenopus, chicken and mouse). It points to an evolutionarily conserved mechanism specifying dorsal cell fates in the nervous system of both protostomes and deuterostomes (Rusten, 2002).

The second phase of Dpp signaling, covering most if not all the dorsal ectoderm, starts at stage 9 and lasts until stage 10/11 [3.40 to 5.20 hours after egg laying (AEL)]. Initially proneural clusters (PNCs) and later sensory organ precursors (SOPs), singled out within each PNCs, can be visualized by the expression of the proneural genes achaete (ac, 4.20-7.20 hours AEL), atonal (ato, 5-6.30 hours AEL) and amos (amo, 5.20-6 hours AEL). Thus, the second wave of Dpp signaling precedes and overlaps with the development of the PNCs and SOPs. The domain of Dpp signaling was examined using an enhancer trap lacZ line inserted in the gene daughters against dpp (dad), a target of Dpp. Double immunofluorescence staining shows that dorsally located Ac and Ato positive PNCs and SOPs originate inside the dad-lacZ positive region, suggesting that they have received, or still receive, Dpp signaling. However, a subset of PNCs and SOPs are ventral to the dad-lacZ domain. As the PNS neuronal precursors differentiate close to the position where they originate, it can be concluded that a part of the dorsal PNS forms within an active Dpp signaling region (Rusten, 2002).

The embryonic abdominal (A) PNS of Drosophila consists of three bilateral clusters of neurons (ventral, lateral and dorsal) per segment, which can be most especially appreciated in the serially homologous segments A1-A7. In order to investigate whether the second phase of Dpp signaling is necessary for patterning the PNS, mutant alleles for a gene involved in the Dpp signaling pathway, schnurri (shn), were examined. This gene encodes a zinc-finger transcription factor that is necessary for the transcription of some Dpp target genes and binds directly to the main Dpp mediator Mothers against Dpp (Mad). Unlike the zygotic mutants of dpp, scw, tolloid (tld) or mad, shn mutants have no effect on the initial dpp/scw governed dorsoventral patterning of the blastoderm. They express normally the early Dpp target genes, such as pannier (pnr, stage 7), dpp itself in the dorsal ectoderm (stage 9) and Krüppel (Kr) (which is a marker for the amnioserosa), showing that the dorsal ectoderm is correctly specified. By contrast, several Dpp target genes that are expressed following the second phase of Dpp signaling are affected in shn zygotic mutants: at stage 11, the expression of genes responsive to Dpp signaling, such as dad, pnr, spalt or dpp itself is reduced or lost. Thus, any failures in PNS formation, which are observed in shn mutant embryos, must originate from the second rather than the first phase of Dpp signaling and are likely to be mediated by Shn. PNS malformations were sought in strong shn zygotic mutant embryos using the ubiquitous PNS neuronal marker 22C10. Homozygous shn1 and shnk00401 fail to undergo dorsal closure and show severe defects of PNS development. A strong reduction in number of neurons is observed, especially in the dorsal and lateral PNS clusters, although it is difficult to determine exactly which neurons are affected because of the dorsal closure failure. Therefore, another allele, shnk04412, which does undergo dorsal closure, was also examined. In these embryos, position and identity of PNS neurons could be more clearly assigned. In homozygosity, as well as in transheterozygosity over shn1, this mutant shows a reduction in the number of dorsal and lateral neurons, similar to the other mutants analyzed. These results are consistent with a role for Shn-mediated Dpp signaling in the formation of the dorsal and lateral PNS (Rusten, 2002).

A different way to interfere with the second phase of Dpp signaling is to express specific inhibitors once the initial dorsoventral patterning is accomplished. Brk is a nuclear protein that negatively regulates Dpp-induced genes and is expressed ventrally in a complementary pattern to Dpp in the embryo. Sog is a secreted protein that can bind to Dpp and inhibit it from signaling, and Supersog (Ssog) is a hyperactive inhibitory fragment of Sog. In order to avoid interference with the first wave of Dpp signaling (stage 5 to 7, 2.10-3.10 hours AEL), brk and ssog were misexpressed from stage 8 (3.10 hours AEL) to interfere with the second phase (stage 9 to 10/11, 3.40-5.20 hours AEL). UAS-brk expression in segments T2-A3, which is driven by the Kr-Gal4 driver, and ubiquitous expression of ssog in the entire embryo, produced using a HS-ssog construct, leads to reduced number of neurons in the dorsal and lateral PNS. The effects are less severe for Ssog misexpression than for UAS-brk misexpression and notable for shn mutations. Approximately 20% of the embryos expressing ubiquitous ssog do not undergo dorsal closure, similar to the phenotype observed when strong alleles of shn are analyzed. The HS-ssog produces a manifest decrease in phosphorylated-Mad (p-Mad) in the dorsal region. This indicates a reduction in Dpp signaling responsible for the phenotype. The residual p-Mad staining observed in some embryos might be the reason why Ssog misexpression leads to less severe effects than UAS-brk misexpression or shn mutations (Rusten, 2002).

In all these mutant backgrounds the dorsal and lateral PNS clusters show a severe reduction in the number of neurons. No major differences are found depending on the neuronal type: the percentage of external sensory organ neurons lost is similar to the loss of neurons in the chordotonal organs. The penetrance of this effect, as measured in the differentiated PNS clusters, varies among abdominal segments. The average reduction in neuronal number ranges from 25% (HS-ssog) to 41% (shnk00401) in the dorsal cluster and 8% (HS-ssog) to 52% (shnk00401) in the lateral cluster. By contrast, the ventral cluster is less affected because it shows 2% (HS-ssog) to 18% reduction (shnk00401). The lateral pentascolopodial organ shows migration defects in these embryos, but the other sensory organs are located in their expected relative positions (Rusten, 2002).

The reduced number of neurons observed in the dorsal and lateral PNS when Dpp signaling is impeded could result from lack of proneural gene expression, which is known to be necessary for PNC and SOP formation. The expression of ato and ac was analyzed to examine the specification of progenitor cell subclasses in mutant backgrounds defective for Dpp signaling. The development of the serially homologous abdominal segments A1 to A7 is similar and very synchronous. Thus, in the wild type, whenever a specific number of PNCs and SOPs appear in one abdominal segment, a similar pattern is observed in the other abdominal segments as well. This is not true for shnk04412 mutants and for embryos expressing ubiquitous ssog, where the numbers of Ac and Ato positive SOPs and PNCs vary among the abdominal segments. This is consistent with the variably penetrant phenotypes observed in differentiated PNS among abdominal segments. In embryos expressing Kr-Gal4;UAS-brk, loss of Ato- and Ac-positive PNCs and SOPs was observed specifically in the abdominal segments A1-A3 where brk was misexpressed, when compared with abdominal segments A4-A7 that served as an internal reference. The reduced numbers of Ato- and Ac-positive neuronal progenitors appear to result from failure of PNC formation rather than an increase in cell death ratio: apoptosis does not appear to increase in segments expressing brk compared with the other abdominal segments. Taken together, these results suggest that reduction in the number of neurons is produced by failure in proneural gene expression (Rusten, 2002).

Hedgehog and Decapentaplegic instruct polarized growth of cell extensions in the Drosophila trachea

The trachea is a respiratory organ consisting of a network of tubular epithelia that delivers outside air directly to target organs. The tracheal primordium forms six primary branches that migrate towards specific target tissues expressing Branchless (Bnl), a Drosophila homolog of FGF. Bnl activates Breathless (Btl), an FGF receptor, at the tip of primary branches and cell process formation. The dorsal branch (DB) migrates toward the dorsal midline, where it fuses with another DB from the contralateral side. Two specialized cells are present at each DB tip. Fusion cells lead the migration and form anastomoses of tracheal tubules, whereas terminal cells extend a long cell process called a terminal branch. The terminal branch is also present in other branches such as visceral and ganglionic branches, and in all cases spreads over the surface of target tissues and serves as an interface for gas exchange by extending unicellular processes containing a dead-ended lumen, a structure known as the tracheole. Terminal branching in postembryonic stages is regulated by Bnl, which is induced as a hypoxic response. The regulation of terminal branch migration in the CNS uses the same molecules involved in axon guidance. Current knowledge is lacking, however, on the regulation of directed terminal branch growth over the epidermis, as well as the mechanism by which it is positioned over the epidermis to maximize oxygen transfer. The guidance roles of the morphogens Hedgehog and Decapentaplegic during directed outgrowth of cytoplasmic extensions in the Drosophila embryonic trachea was investigated. A subset of tracheal terminal cells adheres to the internal surface of the epidermis and elongates cytoplasmic processes called terminal branches. Hedgehog promotes terminal branch spreading and its extension over the posterior compartment of the epidermis. Decapentaplegic, which is expressed at the onset of terminal branching, restricts dorsal extension of the terminal branch and ensures its monopolar growth. Orthogonal expression of Hedgehog and Decapentaplegic in the epidermis instructs monopolar extension of the terminal branch along the posterior compartment, thereby matching the pattern of airway growth with that of the epidermis (Kato, 2004).

Terminal branches extend numerous cell processes that rapidly and repeatedly extend and retract in many directions. Although cell processes that extend anteriorly and dorsally are unstable, a subset of cell processes that extend ventrally along the posterior (P) segmental compartment becomes selectively stabilized. The behavior of terminal cells in the anterior compartment is strikingly different from that in the P compartment. In the anterior compartment, the number and size of the cell processes are much smaller, suggesting that the P compartment constitutes the preferred substrate for terminal cell spreading (Kato, 2004).

Hh is important for promoting terminal cell spreading. Hh is secreted from the P compartment and forms symmetrical gradients of cellular responses in both the anterior and posterior directions within the epidermis. It is proposed that Hh stimulates the adhesion of terminal cells to the epidermis by activating Ci. Because Hh signaling is submaximal in terminal cells, terminal branch filopodia that extend randomly would be preferentially stabilized near the source of Hh. Thus, terminal cell bodies are placed at the point of highest Hh concentration and terminal branches are stabilized at the apex of the Hh concentration gradient. Because terminal cell growth continues while the level of Hh signaling remains below maximum within terminal cells, terminal branches would be expected to extend along the P compartment (Kato, 2004).

The mechanism(s) of terminal branch guidance by Hh through regulation of Ci-dependent transcription differs from those that guide the behavior of the growth cone, which is primarily regulated at the level of cytoskeletal motility. The latter mechanism has the advantage of maintaining a small cell-surface area receiving guidance cues to minimize the chances of making aberrant connections during synapse formation. Terminal cells, however, use the entire basal cell surface to receive a guidance signal and to stabilize their association with the epidermis. This mechanism of terminal branching by cell spreading meets the physiological requirement that the terminal branch serves as an interface for gas exchange (Kato, 2004).

Hh signaling has been implicated in another cell adhesion-related process, namely cell sorting behavior at the AP compartmental boundary in the wing imaginal disc. Because this behavior is regulated transcriptionally by Ci, there may be a common downstream target of Ci that acts in both wing disc cells and terminal cells (Kato, 2004).

Terminal branch extension is limited to the AP compartmental border, suggesting that there is an additional mechanism that shifts the terminal branch to the anterior side of the P compartment. Bnl was expressed as short stripes in the dorsal epidermis (DE) at the time of terminal branching. It has been reported that mesodermal cells also contribute to correct patterning of dorsal branch (DB). It will be interesting to address guidance functions of those components on terminal branch outgrowth (Kato, 2004).

Dpp is expressed at the dorsal edge of the DE during dorsal closure. This corresponds to the time when terminal branch outgrowth in the DB starts, suggesting that Dpp affects the initial stage of terminal branch outgrowth. Prolonged activation of Dpp signaling in terminal cells by expression of Dpp prevents its elongation. These observations suggest that in normal development Dpp prevents dorsally directed terminal branch extension at the onset of terminal branching, thereby shunting terminal branch extension towards the ventral direction. It is proposed that Dpp converts the initial bipolar shape of a terminal branch into one that is monopolar. Once terminal branch extension is initiated, its direction may be maintained by localized Bnl expression at the AP compartment border. Whether this inhibitory effect of Dpp is mediated by direct signaling to cytoskeletons at the cell periphery or mediated by a nuclear transduction of the signal, remains to be determined (Kato, 2004).

Terminal cells undergo an enormous increase in cell volume and surface area during terminal branching, which continues throughout embryonic and post-embryonic stages. FGF signaling promotes this process in two ways, first by activating the target gene SRF, which is required for terminal branch growth, and second by stimulating rapid filopodial movement. These two FGF signaling effects seem to be independent of Hh and Dpp. The FGF ligand Bnl is expressed in epidermal cells beneath terminal cells during terminal branching. Because this expression is limited to a relatively small region, it is considered unlikely that FGF signaling is sufficient to provide vectorial information for terminal branching. It is suggested that FGF-driven growth and the motility of terminal branches are restricted to the P compartment by Hh signaling, wherein they are further limited by Dpp to establish a monopolar growth pattern (Kato, 2004).

Hh functions as both a cell fate determinant and a guidance molecule for terminal branch extension. But how are these two distinct Hh functions coordinated? Inactivation of Hh after initiation of terminal branching via a temperature shift of hhts2 mutant embryos causes a loss of terminal cells, suggesting that maintenance of tracheal cell fate also depends on a late Hh function. Thus, the striped expression of Hh in the epidermis is used simultaneously for epidermal patterning, tracheal cell fate determination and terminal branch guidance, exemplifying a simple strategy to coordinate the patterning of complex organs having multiple tissue types (Kato, 2004).

Proximodistal subdivision of Drosophila legs and wings: the elbow-no ocelli gene complex serves as a mediator of the function of the Wg and Dpp signaling systems

Appendages are thought to have arisen during evolution as outgrowths from the body wall of primitive bilateria. In Drosophila, subsets of body wall cells are set aside as appendage precursors through the action of secreted signaling proteins that direct localized expression of transcription factors. The Drosophila homeodomain protein Distal-less is expressed in the leg primordia and required for formation of legs, but not wings. The homeodomain protein Nubbin is expressed in the wing primordia and required for formation of wings, but not legs. Given that insect legs and wings have a common developmental and evolutionary origin, attempts were made to identify genes that underlie the specification of all appendage primordia. Evidence is presented that the zinc-finger proteins encoded by the elbow and no ocelli genes act in leg and wing primordia to repress body wall-specifying genes and thereby direct appendage formation (Weihe, 2004).

Evidence suggests that the el and noc genes serve as mediators of the function of the Wg and Dpp signaling systems in specification of the appendage field within the imaginal discs. El and Noc are induced by Wg and Dpp and are required to repress the proximally expressed proteins Hth and Tsh. Previous work had identified Dll as a gene required for appendage formation in leg and antenna, and nub as a gene required for wing. This report identifies El and Noc as a pair of zinc-finger proteins that function in both ventral and dorsal appendages. However, there are interesting differences in the way that they do so, when examined in detail (Weihe, 2004).

Dll expression is required for the formation of all leg and antenna elements in the ventral (leg) discs, and until this work Dll was the earliest known marker for the distal region leg disc. Previous work has shown that repression of Hth and Tsh by Dpp and Wg was not required for expression of Dll in the leg, nor could Dll repress Hth and Tsh. Thus an essential mediator of the effects of Wg and Dpp was missing. The current results present evidence that El and Noc serve this function, since their removal leads to ectopic expression of Hth and Tsh. Removal of El and Noc does not cause loss of Dll expression, so it is concluded that Wg and Dpp act independently to induce El and Noc expression and Dll to define the distal region of the leg disc (Weihe, 2004).

The situation differs slightly in the wing. Repression of Tsh is the earliest marker for specification of the distal wing region, preceding the onset of Hth repression or of Nub induction. Loss of Tsh and Hth are required to allow Nub expression. Ectopic expression of Hth and Tsh and loss of Nub is observed in clones lacking El and Noc activity. Thus in the wing, expression of the distal marker Nub cannot be demonstrated to be independent of El and Noc (because ectopic Hth can repress Nub, but not Dll). The vestigial gene is also important for wing development and has been proposed to be a wing specifying gene. However, Vestigial is expressed all along the DV boundary of the wing, both in the wing primordium and in the body wall. This led to the suggestion that while Vestigial is essential for wing development, its expression cannot be taken as a molecular marker for wing identity per se, particularly at early stages. For this reason analysis of the relationship between El, Noc and Vestigial was not performed in this study (Weihe, 2004).

Is the repression of trunk genes needed to specify appendage, as opposed to body wall, in wing and leg discs? In the wing disc the answer appears to be yes; repression of 'trunk genes' like hth is necessary to make the remaining part of the disc competent to form the appendage. However, in the leg the situation is more complex. Coexpression of Dll and Hth does not disrupt proximal-distal axis formation, but leads to homeotic transformation of leg tissue into antennal tissue. Hth is not repressed and limited to proximal areas in the antenna. However, loss of el and noc activities in the leg disc leads to loss of distal leg tissue without any evident transformation into antennal tissue. Thus, El and Noc may regulate the expression of other 'trunk genes', whose restricted expression is required to make the remaining leg and antenna disc competent to form the appendage (Weihe, 2004).

The regional requirements for El and Noc highlight another interesting difference between leg and wing disc development. el noc double mutant cells are excluded from contributing to the tarsal region of the leg but not from contributing to the femur and tibia. Lineage tracing has shown a considerable net flux of cells from the proximal (Tsh-expressing domain) into femur and tibia. While there is no boundary of lineage restriction separating these domains, cells must be able to change from expressing the proximal marker Hth to expressing the distal marker Dll in order to move from one territory to the other. The wing in contrast does not appear to normally exhibit this large net flux of cells from proximal to distal and the el noc double mutant cells are excluded from contributing to the entire wing region. Clonal analysis has suggested that el noc double mutant cells attempt to sort out toward proximal territory, or if that fails, they can be lost from the disc, apparently by sorting out perpendicular to the epithelium. These observations suggest that El and Noc activity may contribute to the production of proximal-distal differences in cell affinities and thereby may help to maintain segregation of these cell populations during development (Weihe, 2004).

Embryonic development of the Drosophila corpus cardiacum

The development of the Drosophila neuroendocrine gland, the corpus cardiacum (CC) was investigated, along with the role of regulatory genes and signaling pathways in CC morphogenesis. CC progenitors segregate from the blastoderm as part of the anterior lip of the ventral furrow. Among the early genetic determinants expressed and required in this domain are the genes giant (gt) and sine oculis (so). During the extended germ band stage, CC progenitor cells form a paired cluster of 6–8 cells sandwiched in between the inner surface of the protocerebrum and the foregut. While flanking the protocerebrum, CC progenitors are in direct contact with the neural precursors that give rise to the pars intercerebralis, the part of the brain whose neurons later innervate the CC. At this stage, the CC progenitors turn on the homeobox gene glass (gl), which is essential for the differentiation of the CC. During germ band retraction, CC progenitors increase in number and migrate posteriorly, passing underneath the brain commissure and attaching themselves to the primordia of the corpora allata (CA). During dorsal closure, the CC and CA move around the anterior aorta to become the ring gland (see Image). Signaling pathways that shape the determination and morphogenesis of the CC are decapentaplegic (dpp) and its antagonist short gastrulation (sog), as well as hedgehog (hh) and heartless (htl; a Drosophila FGFR homolog). Sog is expressed in the midventral domain from where CC progenitors originate, and these cells are completely absent in sog mutants. Dpp and hh are expressed in the anterior visceral head mesoderm and the foregut, respectively; both of these tissues flank the CC. Loss of hh and dpp results in defects in CC proliferation and migration. Htl appears in the somatic mesoderm of the head and trunk. Although mutations of htl do not cause direct effects on the early development of the CC, the later formation of the ring gland is highly abnormal due to the absence of the aorta in these mutants. Defects in the CC are also caused by mutations that severely reduce the protocerebrum, including tailless (tll), suggesting that additional signaling events exist between brain and CC progenitors. The parallels between neuroendocrine development in Drosophila and vertebrates are discussed (De Velasco, 2004).

In the larva, the ring gland forms a large and conspicuous structure located anterior to the brain and connected to the brain by a pair of tracheal branches and the paired nerve of the corpus cardiacum (NCC). Three different glands, the corpus allatum (CA; dorsally), prothoracic gland (laterally), and corpus cardiacum (CC; ventrally) form part of the ring gland. By far, most of its volume is taken up by the prothoracic gland whose cells, the source of ecdysone, grow in size and number as larval development progresses, whereas the cells of the CC remain small and do not appear to proliferate. Both the CC and CA, as well as axons innervating the ring gland, are FasII positive from the late embryonic stage onward. Labeling of the CC is stronger and starts earlier (stage 11) than that of the CA (stage 15), which makes it easy to distinguish between the two structures in the embryo. Another convenient marker of the CC is adipokinetic hormone (AKH), which is expressed exclusively in the CC from late embryonic stages onward (De Velasco, 2004).

The ring gland of the mature embryo is situated posterior to the brain hemispheres. The CC and CA occupy their positions ventral and dorsal to the aorta, respectively. The prothoracic gland cannot yet be recognized as a separate entity, possibly due to the fact that its precursors are small and few in number. Cells of the CC number around eight on each side and are arranged in a U-shape around the floor of the aorta. All cells are spindle shaped and send short processes ventromedially where they meet and form a bundle attached to the ventral wall of the aorta (subaortic processes) (De Velasco, 2004).

Several signaling pathways, notably Shh, BMP, and BMP antagonists, Wnt and FGF, specify the fate map of the head in vertebrates and also control later morphogenetic events shaping head structures. The same signaling pathways are active at multiple stages in Drosophila head development, and the pattern of activity and requirement of these pathways in regard to CC development was therefore investigated (De Velasco, 2004). .

The first signal acting zygotically in the Drosophila head is the BMP homolog Dpp, which forms a dorsoventral gradient across the blastoderm. The homolog of the BMP antagonist Chordin, short gastrulation (Sog), is expressed in the ventral blastoderm, overlapping with the ventral furrow. Loss of sog results in the absence of the CC, while the SNS is still present, which reflects ventral origin of the CC. Sog seems to be the only signal, of those tested, required for CC determination, since mutation of all other pathways does not eliminate the CC but merely effects its size, shape, or location (De Velasco, 2004).

Following its early widespread dorsal expression, Dpp becomes more confined during gastrulation to a narrow mid-dorsal stripe and an anterior cap that corresponds to parts of the anlagen of the esophagus and epipharynx. From this domain segregates the most anterior population of head mesoderm cells that give rise to the visceral muscle of the esophagus and which maintain Dpp expression. The visceral mesoderm of the esophagus flanks both CC and SNS. Loss of Dpp causes absence of the SNS; the CC is still present and expresses AKH but does not migrate posteriorly (De Velasco, 2004).

Activity of the MAPK signaling pathway is widespread in the Drosophila head from gastrulation onward. Beside a wide anterior and posterior domain traversing the lateral and dorsal domain of the head ectoderm, the primordia of the foregut, including the SNS, and head mesoderm show a dynamic MAPK activity. At least two RTKs, EGFR and FGFR/heartless, drive the MAPK pathway in the embryonic head. EGFR is responsible for activation in the ectoderm and foregut. Loss of EGFR causes widespread cell death in the head and the absence of the SNS. The CC is still present, although reduced in size. Activation of MAPK by Heartless (Htl) occurs in a narrow anterior domain of head mesoderm that gives rise to the dorsal pharyngeal muscles. The foregut, SNS, and CC develop rather normally in htl mutants. However, the CC shows variable defects in shape and location, which are most likely due to the absence of the aorta and CA, both of which are derivatives of the dorsal mesoderm, which is defective in htl loss of function and to which the CC is normally attached (De Velasco, 2004).

Dpp signalling orchestrates dorsal closure by regulating cell shape changes both in the amnioserosa and in the epidermis

During the final stages of embryogenesis, the Drosophila embryo exhibits a dorsal hole covered by a simple epithelium of large cells termed the amnioserosa (AS). Dorsal closure is the process whereby this hole is closed through the coordination of cellular activities within both the AS and the epidermis. Genetic analysis has shown that signalling through Jun N-terminal Kinase (JNK) and Decapentaplegic (Dpp), a Drosophila member of the BMP/TGF-β family of secreted factors, controls these activities. JNK activates the expression of dpp in the dorsal-most epidermal cells, and subsequently Dpp acts as a secreted signal to control the elongation of lateral epidermis. This analysis shows that Dpp function not only affects the epidermal cells, but also the AS. Embryos defective in Dpp signalling display defects in AS cell shape changes, specifically in the reduction of their apical surface areas, leading to defective AS contraction. These data also demonstrate that Dpp regulates adhesion between epidermis and AS, and mediates expression of the transcription factor U-shaped in a gradient across both the AS and the epidermis. In summary, this study shows that Dpp plays a crucial role in coordinating the activity of the AS and its interactions with the LE cells during dorsal closure (Fernández, 2007).

Several studies have implicated Dpp signalling in the process of Drosophila dorsal closure. The expression of dpp in the LE cells and the observation that a failure of Dpp signalling leads to defects in the dorso-lateral epidermal cells led to the suggestion that Dpp acts as a secreted factor regulating epidermal morphogenesis during dorsal closure. During dorsal closure zygotic mutant embryos lacking Dpp receptor activity (tkv) display specific defects in the epidermal cells that can be rescued in a tissue autonomous manner. The LE cells fail to elongate properly in a coordinated manner and display aberrant morphologies. Moreover, the LE cells do not organise microtubule bundles in the dorsoventral axis as in the wild-type. The more lateral epidermal cells begin to elongate but eventually this movement fails. Together these defects in Dpp signalling mutants prevent the zipping of the epidermal fronts at both ends, confirming and extending previous suggestions of a role for Dpp signalling in the epidermal cells (Fernández, 2007).

In addition to the defects in the cytoskeleton of the epidermal cells, this study shows that tkv mutant embryos display severe defects in the behaviour of the AS cells that can be rescued by activation of Dpp signalling in a tissue autonomous manner. The importance of the AS for germ band retraction and dorsal closure has been demonstrated by laser ablation experiments and by cell ablation using tissue specific drivers to express toxins in the AS. However, the signals that control the AS movements still remain elusive. The current results show that one of these signals is likely to be Dpp coming for the LE cells. In this regard, it is notable that during vertebrate wound healing, a process similar to dorsal closure, TGFβ elicits a paracrine response in both the epidermis cells adjacent to the LE as well as in the mesenchymal cells underlying the wound (Fernández, 2007).

In addition to tissue autonomous defects in the AS and the epidermis, it was observed that the adherens junctions between these two tissues are defective in tkv mutants as reflected in low levels of Armadillo staining. Consistent with this, PTyr levels, which are largely localised to the adherens junctions, are downregulated in Dpp signalling mutants at the late stages of dorsal closure. Eventually this interface breaks up and the tissues separate from each other resulting in a dorsal hole. It is difficult to clearly distinguish if the defects in the adherens junctional markers are a cause or a consequence of the detachments that were observe between the AS and the epidermis in tkv mutant embryos. However, it is clear that Dpp signalling is involved in maintaining the AS-LE interface integrity. Very recent work (Wada, 2008) has confirmed this possibility showing that Dpp can regulate integrin activity at the interface between the AS and the LE (Fernández, 2007).

Together these results identify several discrete requirements for Dpp signalling during dorsal closure and suggest that Dpp is acting as an orchestrator of morphogenetic movements to ensure that the elongation of the epidermis and the contraction of the AS occur in a coordinated manner. How Dpp signalling controls these morphogenetic processes is not clear? Two studies have shown that Dpp is involved in ensuring the correct architecture of epithelial cells in the wing disc. In this epithelium, tkv mutant cells lose their normal columnar organisation, round up and are extruded from the tissue. Additionally, these cells display abnormal microtubule polarity similar to that of the LE cells of tkv mutants. It is possible that Dpp signalling has an effect on microtubule organisation by activating the localised expression of regulatory proteins, but these targets still remain to be identified (Fernández, 2007).

Previous studies have shown that Dpp signalling acts downstream of JNK signalling during dorsal closure. This conclusion is mainly based on two observations: the expression of dpp in the LE cells during dorsal closure is absent in JNK mutants and ectopic activation of Dpp signalling can rescue the defects of JNK mutants. However, the phenotype of the epidermal cells is different in JNK and Dpp signalling mutants, and more significantly it was shown that the contraction of the AS is not compromised in the absence of JNK signalling. This suggests that Dpp signalling is acting independently of JNK signalling in the AS, and that it is only its later function within the LE that is dependent on JNK activation. This hypothesis is consistent with the observation that only the LE expression of dpp is absent in JNK mutants (Fernández, 2007).

The requirement for Dpp signalling in the AS correlates with earlier patterns of dpp expression, in particular with the broad dorsal epidermal expression in the extended germ band, which is not strongly disrupted in JNK mutants. Cytoskeletal rearrangements of the AS cells are required for germ band retraction and tkv mutants show variable but clear defects in this process. It is possible that early Dpp signalling acts on the AS to regulate germ band retraction and that this early activity also initiates cellular activities that are required later for the contraction of the AS during dorsal closure (Fernández, 2007).

The data suggest a model in which Dpp mediates temporally and spatially separable functions during the process of dorsal closure. The results show that dpp expression at the extended germ band stage is required to regulate the behaviour of the AS, in a JNK independent manner. Subsequently, during dorsal closure stages, JNK drives the expression of dpp at the LE cells, which is necessary for epidermal morphogenesis and appears to contribute to the adhesive integrity of the interface between the LE cells and AS. During the early phase of closure the main force that drives the process seems to be provided by the AS which pulls the epidermal cells in tow. While the epidermis elongates dorsally the AS cells actively constrict their apical surface. In a tkv mutant where the AS cells are not actively constricting the process can be overcome by overexpressing a constitutively active form of Tkv in the epidermis. Also, if the epidermis is maintained mutant and the AS cells express the constitutively active form of the receptor, the dorsal gap is again closed. In both cases closure is achieved at a slower rate and the final pattern is imperfect compared to the wild-type situation. These results suggest that, closure is accomplished by the cooperative and active contribution of both tissues regulated by Dpp signalling. The fact that tkv mutants can be rescued when an activated form of Tkv is expressed either in the AS cells or in the ectodermal cells also suggests that Dpp signalling may not be required in a completely cell autonomous manner. Dpp could act through a relay mechanism in the nearby tissue to induce a diffusible signalling factor required for dorsal closure. However, this possibility seems unlikely as such putative factor and the corresponding signalling cascade have not been identified. An alternative explanation is based on the regulation of adhesion at the LE/AS interface by Dpp; activation of the Dpp pathway on either side of the interface may be sufficient to strengthen the adhesion and rescue the tkv mutant phenotype, at least partially. In any case, Dpp seems to be acting as a coordinator of dorsal closure to ensure that the cell shape changes in the epidermis and in the AS result in the desired final pattern (Fernández, 2007).

Leading edge-secreted Dpp cooperates with ACK-dependent signaling from the amnioserosa to regulate myosin levels during dorsal closure

Dorsal closure of the Drosophila embryo is an epithelial fusion in which the epidermal flanks migrate to close a hole in the epidermis occupied by the amnioserosa, a process driven in part by myosin-dependent cell shape change. Dpp signaling is required for the morphogenesis of both tissues, where it promotes transcription of myosin from the zipper (zip) gene. Drosophila has two members of the Activated Cdc42-associated Kinase (ACK) family: DACK and PR2. Overexpression of DACK (Ack) in embryos deficient in Dpp signaling can restore zip expression and suppress dorsal closure defects, while reducing the levels of DACK and PR2 simultaneously using mutations or amnioserosa-specific knock down by RNAi results in loss of zip expression. ACK function in the amnioserosa may generate a signal cooperating with Dpp secreted from the epidermis in driving zip expression in these two tissues, ensuring that cell shape changes in dorsal closure occur in a coordinated manner (Zahedi, 2008).

The results on the regulation of zip expression by Dpp are consistent with the model of Fernandez and colleagues (2007) in which two rounds of dpp expression in the DME cells regulate dorsal closure. In the first round of dpp expression, before completion of germband retraction, Dpp signals from the DME cells to the amnioserosa. This signaling can be visualized by pMad staining in the amnioserosa, which is obvious during germband retraction but fades away by the beginning of dorsal closure, a pattern paralleled by zip expression. In the second round of signaling, occurring during dorsal closure, Dpp signals to the dorsal epidermis, as demonstrated by robust pMad in this tissue and high zip levels in the DME cells. Dpp signaling and ACK function are both necessary but not sufficient for zip expression in the embryo during dorsal closure. Activation of either Dpp signaling or ACK function in prd stripes does not lead to ectopic zip expression, indicating that in each case additional inputs are required. DACK is able to elevate zip expression only on the dorsal side of the embryo in regions where zip is normally expressed, indicating that required additional inputs are present; it is proposed that one such input is Dpp signaling. When DACK is overexpressed in the amnioserosa, either in prd stripes or throughout the whole tissue, its effects on zip expression are non–cell-autonomous, leading to up-regulation of zip expression throughout the dorsal side of the embryo. With regard to the zip expression pattern, seen with prd-Gal4-driven DACK transgenes, a diffusible signal emitted from prd stripes in the amnioserosa could attain a fairly uniform distribution over the dorsal side of the embryo. It is proposed that Dpp secreted from the DME cells cooperates with a diffusible signal from the amnioserosa (regulated by ACK in a kinase-independent manner) to drive coordinated zip expression in these two tissues. ACK appears to make the larger input into zip expression as DACK overexpression results in a clear elevation in zip levels on the dorsal side of the embryo, but this is not seen with excessive Dpp signaling. Simultaneously activating Dpp signaling and overexpressing DACK with prd-Gal4 is not sufficient to promote ectopic zip expression (for example in the ventral epidermis), indicating that other components are required for zip expression, consistent with DACK operating through downstream signaling events (Zahedi, 2008).

It is well established that communication between the amnioserosa and the epidermis is critical for embryonic morphogenesis, and this study has identified the zip locus as one target of such crosstalk, with zip transcription in both tissues dependent on signals secreted by both tissues. A diffusible signal from the amnioserosa to the epidermis has been proposed in the regulation of germband retraction by Hindsight, a transcription factor that is member of the U-shaped group of genes expressed in the amnioserosa, and preliminary results indicate that the U-shaped group is involved in the regulation of zip expression in both the amnioserosa and the epidermis. How could ACK tie into transcriptional regulation of a diffusible signal from the amnioserosa to the epidermis? There is evidence that ACK functions in clathrin-mediated receptor endocytosis in a kinase-independent manner, and is possible that ACK regulates by receptor endocytosis a pathway in the amnioserosa that leads to a transcriptional response. These data suggest that the kinase activity of DACK may actually impair its ability to drive zip expression, as KD-DACK promoted higher zip levels than wild-type DACK. Interestingly the binding of mammalian ACK1 to SNX9/SH3PX1, a member of the sorting nexin family of proteins involved in the sorting of proteins in the endosomal pathway, is inhibited by ACK1 kinase activity . The interaction of ACK with a sorting nexin is conserved in flies, where Drosophila SH3PX1 binds to DACK (Zahedi, 2008).

In addition to the transcription of myosin in the DME cells being dependent on input from the amnioserosa, assembly of the actomyosin contractile apparatus in the DME cells requires an expression border for the adhesion molecule Echinoid between the amnioserosa and the epidermis. The juxtaposition in epithelia of cells expressing Ed with those not expressing Ed triggers actomyosin cable assembly. Ed is expressed in the epidermis but not the amnioserosa and this provides a means, in addition to ACK-mediated signaling, by which the amnioserosa 'communicates' with the epidermis in regulating the actomyosin contractile apparatus (Zahedi, 2008).

Previous studies have found that global activation of Cdc42 signaling in tkv mutant embryos could suppress the dorsal closure defects caused by a reduction in Dpp signaling, and the present results indicate a major route of action for this suppression is ACK. The activation of Cdc42 throughout the embryo leads to increased expression of DACK specifically in the amnioserosa, and this study has shown that overexpressing DACK in this tissue can suppress tkv dorsal closure defects. The results indicate a tissue-specific regulation of DACK levels by Cdc42 that may be part of a sophisticated signaling network enabling the coordinated morphogenesis of tissues in the embryo. DACK does not bind Cdc42 but PR2 does and Cdc42 may also regulate ACK function during dorsal closure through direct interaction with PR2. The serine/threonine kinase dPak, an effector for Rac/Cdc42, may also be a component of Cdc42-mediated communication between the amnioserosa and the epidermis during dorsal closure. It has been shown that dPak expression in the amnioserosa is regulated by Cdcd42, but in the opposite direction from DACK in that impairment of Cdc42 signaling leads to elevated dpak transcription in this tissue. Impairing dPak kinase activity through amnioserosa-specific expression of the dPak autoinhibitory domain leads to defects in head involution and dorsal closure (Zahedi, 2008).

Head involution defects and germband retraction failures are seen in ACK-deficient embryos and loss of DACK can suppress the head involution defects and germband retraction failure caused by overexpression of Dpp. These results suggest that Dpp signaling and ACK cooperate in the regulation of these morphogenetic events in addition to dorsal closure. A recent review has highlighted the parallels between dorsal closure and head involution in terms of morphogenetic events and the genes required, with both involving epithelial sheet migration, zip expression and Dpp signaling. DACK overexpression leads to excessive zip levels in the head and it is likely that ACK and Dpp signaling work together to provide myosin for head involution. That signaling from the amnioserosa is involved in regulating head involution is supported by an earlier finding that impairing dPak function in the amnioserosa causes failures in this process (Zahedi, 2008).

Does ACK impact Dpp signaling other than at the level of zip transcription? Homozygosity for a DACK allele suppresses the ectopic wing vein phenotype caused by excessive Dpp signaling. It is likely that this phenotype is caused by something other than misregulation of zip expression, and ACK could be regulating the expression of a subset of Dpp target genes (other than dad or salm) or may be interacting with the Dpp pathway at another level (Zahedi, 2008).

TGF-β family signaling is a central regulator of dorsal closure and other epithelial fusions, but how Dpp controls dorsal closure has not been well-defined. We have shown that regulation of zip expression in cooperation with the Drosophila ACKs constitutes a major route of action of Dpp during dorsal closure. These findings may be relevant to vertebrate wound healing, in which closure of the wound involves both epithelial movement and TGF-β–dependent contraction of connective tissue in the wound (Zahedi, 2008).

Larval and pupal stages (part 1/2)

Dpp and the Eye-antennal Disc

The eye-antennal imaginal discs of Drosophila melanogaster form the head capsule, the eyes and the antenna of the adult fly. Unlike the limb primordia, each eye-antennal disc gives rise to morphologically and functionally distinct structures. As a result, these discs provide an excellent model system for determining how the fates of primordia are specified during development. An investigation has been carried out of how the adjacent primordia of the compound eye and dorsal head vertex are specified. Subdivision of the eye-antennal disc is not based on compartmentalization: this is in contrast to the basis for subdivision in the wing and leg discs. Therefore, selector gene-mediated division of the disc into compartments, mediated by engrailed and invected, as in the wing disc for example, is not likely to be the basis for regionalization within the antennal primordium. Instead, in this region, the genes wingless and orthodenticle are expressed throughout the entire second instar eye-antennal disc, conferring a default fate of dorsal vertex cuticle. Mutations that decrease dpp expression in the eye primordia lead to the formation of severely reduced eyes. Similarly, the loss of otd or wg function in the vertex primordia causes the elimination of dorsal head structures (Royet, 1997).

Transplantation experiments show that the eye primordium occupies most of the posterior half of the eye-antennal disc (the so-called 'eye disc'). The head vertex forms from the dorsomedial region of the disc, while the antenna develops from the anterior half of the disc (the so-called 'antennal disc'). During the early third instar stage (70-80 hours after egg laying), dpp is expressed in a horseshoe-shaped domain along the ventral, posterior and dorsal periphery of the eye disc. Dorsal dpp expression does not extend as far anteriorly as ventral expression, but instead ends at the vertex primordium. At this stage, otd expression covers the vertex primordium and extends along the edge of the antennal disc. The posterior boundary of otd expression in the vertex anlage coincides, approximately, with the anterior boundary of the dpp domain. At the same stage of disc development, wg is expressed in two regions of the eye disc. One region corresponds to the future gena (the lateral part of the head capsule bounded above by the eye) and the other to the head vertex (Royet, 1997).

dpp expression prevents dorsal head development in the eye primordium. Flies homozygous for the dppd-blk allele that reduces dpp activity in the eye primordium greatly reduces the compound eye giving rise to an eye with only a few residual ommatidia. In these mutants the eyes are largely replaced by frons cuticle, which normally appears only on the dorsal areas of the head. This ectopic frons lies between the orbital cuticle and the remaining ommatidia, and to the anterior, between the shingle cuticle and the ommatidia. In other eye loss mutants, such as sine oculis or eyes absent, the eyes are completely lost but are not replaced by ectopic frons. This suggests that dorsal head cuticle does not result simply from loss of the eyes, but is caused instead by loss of dpp function. Clones of Mothers against dpp, coding for a protein involved in transmission of the Dpp signal, likewise transform ommatidia into frons (Royet, 1997).

Activation of decapentaplegic expression in the posterior eye disc eliminates wg and otd expression, thereby permitting eye differentiation. In dppd-blk mutants, the otd domain expands toward the anlagen of the shingle cuticle and the compound eyes, consistent with the location of ectopic frons cuticle on dppd-blk mutant heads. wg expression also expands in these mutant discs. Ectopic activation of the wingless pathway (the result of the generation of clones mutant for shaggy/zeste-white 3) in the eye primordium induces otd expression and vertex formation. Loss of shaggy function results in constitutively activated wg signaling and ectopic otd expression. This suggests that otd expression in the vertex primordium is normally activated or maintained by wingless. Early activation of dpp depends on hedgehog expression in the eye anlage prior to morphogenetic furrow formation. Loss of hh activity during the second instar larval stage eliminates dpp expression along the posterior and lateral margins of the eye disc and in the antennal primordium. This loss of dpp expression is associated with a dramatic expansion of the otd expression domain. wg expression also expands into the eye primordium (Royet, 1997).

Unlike the limb discs, which derive from single trunk segments, each eye-antennal disc arises from multiple embryonic head segments. Divisions between segment primordia within the disc could contribute to certain aspects of regional specification. It is proposed that wg and otd expression in the eye-antennal discs are inherited from the embryo, where the two genes are expressed in segments from which these discs are derived. The almost ubiquitous expression of these two genes serves to program the early disc for a vertex fate. Later, hh expression in the posterior region of the future eye disc induces dpp expression along the margins of the eye primordium. dpp represses wg, permitting the formation of the eye primordium (Royet, 1997).

Development of the Drosophila retina occurs asynchronously. The leading edge of differentiation, its front marked by the morphogenetic furrow, progresses across the eye disc epithelium over a 2 day period. The mechanism by which this front advances suggest that developing retinal cells behind the furrow drive the progression of morphogenesis utilizing the products of the hedgehog and decapentaplegic genes. Analysis of hh and dpp genetic mosaics indicates that the products of these genes act as diffusible signals in this process. Expression of dpp in the morphogenetic furrow is closely correlated with the progression of the furrow under a variety of conditions. HH, synthesized by differentiating cells, induces the expression of dpp, which appears to be a primary mediator of furrow movement (Heberlein, 1993).

Pattern formation in the eye imaginal disc is initiated at the posterior edge and moves in a wave toward the anterior; the front of this wave is called the morphogenetic furrow (MF). DPP is required for proliferation and initiation of pattern formation at the posterior edge of the eye disc. It has also been suggested that DPP is the principal mediator of Hedgehog function in driving progression of the MF across the disc. This paper shows that ectopic DPP expression is sufficient to induce a duplicated eye disc with normal shape, MF progression, neuronal cluster formation and direction of axon outgrowth. Induction of ectopic eye development occurs preferentially along the anterior margin of the eye disc. Ectopic DPP expressing clones, situated away from the margins, induce neither proliferation nor patterning. The DPP signaling pathway is shown to be under tight transcriptional and post-transcriptional control within different spatial domains in the developing eye disc. DPP expressing clones located in the middle of the eye disc (i.e. in a region competent to induce dpp and morphogenetic furrows in response to ectopic HH) neither induce ectopic morphogenetic furrows nor endogeneous dpp. DPP positively controls its own expression, as evidenced by the absence of DPP expression in Mad mutant clones. Also, DPP suppresses wingless transcription. In contrast to the wing disc, DPP does not appear to be the principal mediator of Hedgehog function in the eye. Whereas eye tissue away from the margins can respond to HH, it is not competent to respond to ectopic DPP. DPP, WG and HH control proliferation and patterning in essentially all imaginal discs studied. However, their relationship to one another differs from tissue to tissue. In contrast to the developing wing, in the eye disc DPP does not fall strictly under the control of HH, nor is it the principal mediator of HH function. The antagonistic relationship between DPP and WG in the eye disc is similar to that in the leg, however it differs from that in the wing disc (Pignoni, 1997).

Decapentaplegic is expressed at the disc's posterior margin prior to initiation. Under the control of hh, it is expressed in the furrow, during MF progression. While dpp has been implicated in eye disc growth and morphogenesis, its precise role in retinal differentiation has not been determined. To address the role of dpp in initiation and progression of retinal differentiation, the consequences of reduced and increased dpp function have been analyzed during eye development. dpp is not only required for normal MF initiation, but is sufficient to induce ectopic initiation of differentiation. Inappropriate initiation is normally inhibited by wingless. Loss of dpp function is accompanied by expansion of wg expression, while increased dpp function leads to loss of wg transcription. In addition, dpp is required to maintain, and sufficient to induce, its own expression along the disc's margins. It is thought that dpp autoregulation and dpp-mediated inhibition of wg expression are required for the coordinated regulation of furrow initiation and progression. In the later stages of retinal differentiation, reduction of dpp function leads to an arrest in MF progression (Chanut, 1997a).

The progression of retinal morphogenesis in the Drosophila eye is controlled to a large extent by Hedgehog (HH), a signaling protein emanating from differentiating photoreceptor cells. Adjacent, more anterior cells in the morphogenetic furrow respond to HH by expressing dpp, suggesting that the relationship between HH and DPP might be similar to that in the limb imaginal discs where DPP mediates the organizing activity of HH. This study contradicts that suggestion. Analysis of somatic clones of cells lacking the DPP receptors Punt or Tkv reveals that DPP plays only a minor role in furrow progression and no critical role in subsequent ommatidial development. Within tkv and punt clones traversing the furrow at the time of dissection, neuronal differentiation, as shown by ELAV staining, is somewhat retarded, especially in the middle of large clones. The function of DPP in this context must be nonessential or redundant as the furrow is only slightly slowed, but not stopped. Normal ommatidial development occurs in the complete absence of DPP. In contrast, HH-independent dpp expression around the posterior and lateral margins of the first and second instar eye discs is important for the growth of the eye disc and for initiation of the morphogenetic furrow at these margins. Tkv and Punt are absolutely required for cell proliferation in the early developing eye imaginal disc. tkv clones are severly restricted in their ability to grow, implying a strong requirement for the DPP signal for cell proliferation in the early eye disc. There is a posterior requirement for punt function in eye development, which suggests a role for DPP signaling in the initiation of the furrow at the posterior margin Adult eyes containing predominantly punt mutant tissue are regularly observed, but such eyes always have some wild-type tissue at the posterior margin. Both punt and tkv clones cause local overproliferation and block neural differentiation. The tissue in these marginal clones must die, as loss of head cuticle and eye structures is observed in eyes containing mutant clones (Burke, 1996).

The posteriorly expressed signaling molecules Hedgehog and Decapentaplegic drive photoreceptor differentiation in the Drosophila eye disc, while at the anterior lateral margins Wingless expression blocks ectopic differentiation. Mutations in axin prevent photoreceptor differentiation and leads to tissue overgrowth; both these effects are due to ectopic activation of the Wingless pathway. In addition, ectopic Wingless signaling causes posterior cells to take on an anterior identity, reorienting the direction of morphogenetic furrow progression in neighboring wild-type cells. Signaling by Dpp and Hh normally blocks the posterior expression of anterior markers such as Eyeless. Wingless signaling is not required to maintain anterior Eyeless expression and in combination with Dpp signaling can promote Ey downregulation, suggesting that additional molecules contribute to anterior identity. Along the dorsoventral axis of the eye disc, Wingless signaling is sufficient to promote dorsal expression of the Iroquois gene mirror, even in the absence of the upstream factor pannier. However, Wingless signaling does not lead to ventral mirror expression, implying the existence of ventral repressors (Lee, 2001).

Loss of axin function at the posterior margin results in outgrowths from the disc, over-riding the normal control of organ size. axin mutant clones also form smooth borders with surrounding cells, suggesting that their ability to adhere to wild-type cells is decreased. Growth control requires the formation of normal junctions between cells, so it is possible that the outgrowth results from this loss of adhesion. Because the posterior margin is the site of dpp expression prior to initiation, the outgrowth observed could also require Dpp signaling; in the leg disc, overlap between dpp and wg promotes the extension of a proximal-distal axis. However, punt;axin double mutant clones show a similar degree of overgrowth, suggesting that Dpp signaling does not contribute to this (Lee, 2001).

Although Wg signaling is sufficient to establish but not necessary to maintain anterior identity in the eye disc, Dpp signaling has complementary properties; it is sufficient to promote but not essential to maintain posterior identity. Ectopic Dpp can downregulate ey in the anterior, but loss of components of the Dpp pathway has a variable effect on ey expression and does not lead to hairless expression or induce reorientation of the furrow in neighboring cells. It is possible that this is due to redundancy with Hh signaling, which also has a weak and variable effect on ey expression. Clones that are mutant for cell-autonomous components of both the Hh and Dpp pathways do not differentiate, while loss of one or the other pathway only delays differentiation; however, even loss of both pathways does not reorient adjacent wild-type cells (Lee, 2001).

Downregulation of ey and photoreceptor differentiation appear to be independent events, since smo clones have a weaker effect on ey expression than Mad clones, despite their stronger effect on differentiation. Anterior Dpp can also downregulate ey over a much longer range than that over which it promotes photoreceptor differentiation, and loss of slimb downregulates ey without leading to ectopic differentiation (Lee, 2001).

The complementary effects of Wg and Dpp on AP polarity appear to be independent of one another. The effect of Dpp is not simply due to its repression of wg, and posterior Wg signaling can upregulate ey even when the Dpp pathway is also activated. However, activation of both pathways reveals the existence of another mechanism that distinguishes anterior from posterior, since the anterior of the disc is more sensitive to the effects of Dpp and the posterior is more sensitive to the effects of Wg. This leads to a striking repolarization of the eye disc when both pathways are activated, resulting in initiation of an ectopic morphogenetic furrow from the anterior margin as well as reduction or elimination of the normal posterior furrow. Since normal development requires anterior cells to be gradually converted to posterior by the action of Dpp and Hh, it is important for Dpp to overcome the effects of Wg in this region. This also underscores the importance of establishing the early expression patterns of wg and dpp (Lee, 2001).

It is not clear how differential sensitivity to the two pathways is controlled; it may require other localized factors such as Hth or Eyegone. Because the differential sensitivity is observed even when both pathways are activated at the intracellular level, it is not likely to be due to regulation of receptor levels as occurs in the wing disc. However, the intracellular dpp antagonist daughters against dpp (dad) is induced by Dpp signaling and could be present at lower levels in the anterior of the eye disc, making this region more sensitive to ectopic Dpp. No intracellular antagonist of Wg has yet been shown to be induced by the Wg pathway (Lee, 2001).

eyeless (ey) is a key regulator of the eye development pathway in Drosophila. Ectopic expression of ey can induce the expression of several eye-specification genes (eya, so, and dac) and induce eye formation in multiple locations on the body. However, ey does not induce eye formation everywhere where it is ectopically expressed, suggesting that Ey needs to collaborate with additional factors for eye induction. Ectopic eye induction by Ey has been examined in the wing disc; eye induction was spatially restricted to the posterior compartment and the anterior-posterior (A/P) compartmental border, suggesting a requirement for both Hh and Dpp signaling. Although Ey in the anterior compartment induces dpp and dac, these are not sufficient for eye induction. Coexpression experiments show that Ey needs to collaborate with high level of Hh and Dpp to induce ectopic eye formation. Ectopic eye formation also requires the activation of an eye-specific enhancer of the endogenous hh gene (Kango-Singh, 2003).

These results indicated that Ey needs to collaborate with high levels of Hh and Dpp for eye induction. Since Hh and Dpp are secreted molecules and can act over long range, the requirement for their high levels restricts the site of eye development. At the time of MF initiation, hh and dpp are expressed at the posterior margin of the eye disc, and ey is expressed throughout the eye disc. So the coexistence of Ey, high Hh, and high Dpp occurs only at the posterior margin to induce MF initiation. After MF initiation, ey is downregulated in the developing photoreceptor cells posterior to the MF, where hh is expressed (Kango-Singh, 2003).

dpp is expressed only at the MF. Thus, the only location where Ey, high Hh, and high Dpp levels coexist is just anterior to the MF, allowing the MF to progress anteriorly. These results clearly show that even when ey, hh, and dpp are all provided, eye induction still does not occur everywhere, suggesting that additional factors are required. It is proposed that the coexpression of the eye-specification genes, eya, so, dac, ey, toy, and eyg, occurring first in the second instar eye disc, specifies the eye fate (Kango-Singh, 2003).

In leg and antennal discs, the posterior compartment is maintained by engrailed, while the anterior compartment becomes asymmetric in the D/V axis, with decapentaplegic expression defining dorsal anterior leg, and wingless expression defining ventral anterior leg. Both dpp and wg are maintained in the anterior compartment by Hedgehog signaling. Unlike the wing disc, dpp is not expressed at the A/P compartment boundary. In addition, unlike wing discs, no D/V compartment has been demonstrated in legs or antennae. How are the dorsal anterior and ventral anterior territories defined and maintained? wg inhibits dpp expression and dpp inhibits wg expression in leg and eye/antennal discs. Loss of DPP signaling leads to ectopic wg expression. DPP signaling is transduced by its receptor Punt, and punt mutation results in an expansion of wg expression into the dorsal region of leg discs. The antennal portion of the eye-antennal disc is analogous to the leg disc, but inverted in the D/V and A/P axes. Thus wg, which is expressed in the ventral region of the leg disc, is expressed in the dorsal region of the wild-type antennal disc. In punt mutants, wg expression expands into the ventral domain of the antenna to form a continuous stripe along the A/P boundary from dorsal to ventral. In addition, wg expression expands from its normal location at the periphery of the eye anlage into the morphogenetic furrow. Likewise, loss of Wg signaling leads to ectopic activation of dpp transcription. This mutual repression provides a mechanism for maintaining separate regions of wg and dpp expression in a developing field. The term 'territory' is proposed to describe regions of cells that are under the domineering influence of a particular morphogen. Territories differ from compartments in that they are not defined by lineage but are dynamically maintained by continuous morphogen signaling. Thus it is thought that the anterior compartment of the leg disc is divided into dorsal and ventral territories by the mutual antagonism between WG and DPP signaling (Theisen, 1996).

The Decapentaplegic and Notch signaling pathways are thought to direct regional specification in the Drosophila eye-antennal epithelium by controlling the expression of selector genes for the eye (Eyeless/Pax6, Eyes absent) and/or antenna (Distal-less). The function of these signaling pathways in this process has been investigated. Organ primordia formation is indeed controlled at the level of Decapentaplegic expression but critical steps in regional specification occur earlier than previously proposed. Contrary to previous findings, Notch does not specify eye field identity by promoting Eyeless expression but it influences eye primordium formation through its control of proliferation. Analysis of Notch function reveals an important connection between proliferation, field size, and regional specification. It is proposed that field size modulates the interaction between the Decapentaplegic and Wingless pathways, thereby linking proliferation and patterning in eye primordium development (Kenyon, 2003).

This paper analyzes the role of Dpp and Notch in the regional specification of the eye-antennal disc. This study makes four observations: (1) domains of regional identity emerge in a complex pattern starting early in L2; (2) formation of eye and antenna primordia depend upon specific domains of dpp expression that emerge in early-L2 (eye) and mid-L2 (antenna); (3) neither Notch nor Dpp control the establishment of separate eye and antennal fields; (4) Notch can influence the establishment of an eye primordium through its control of proliferation in the eye field. Current models of regional specification have been evaluated based on these results and a new perspective on the emergence of regional identity in this tissue is presented (Kenyon, 2003).

It has been proposed that allocation of eye field and antennal field identity occurs in the latter half of L2 through the restriction of eye selectors, such as Ey, and antennal selectors, such as Dll, to distinct regions of the disc. However, two observations reported in this paper are not consistent with this interpretation: (1) Dll is not expressed ubiquitously at any time during disc development; (2) eye and antennal fields are clearly established by mid-L2 as evidenced by the restricted expression of Ey (eye field) and Cut (antennal field), and by distinct Dpp/Wg patterning centers within each field. These observations place the emergence of separate eye and antennal fields in the first half of L2 and not in the second half as previously proposed. Moreover, onset of Eya occurs in early-L2 and so is expressed by mid-L2. The beginning of eye primordium formation in early-L2, prior to the appearance of distinct fields, indicates that regional specification within this disc does not follow a two-step mechanism (i.e., establishment of separate fields followed by induction of organ primordia) but occurs in a more complex pattern. Further analysis of the transcription factors and signaling molecules active in the late-L1 and early-L2 disc is necessary to better understand how the establishment of eye field identity relates to eye primordium formation and the emergence of an antennal field (Kenyon, 2003).

Analyses of hypomorphic dpp alleles and tissue mutant for Mad implicate Dpp in the control of eya and Dll expression during late larval development. The onset of Eya and Dll expression correlate with specific changes in dpp expression during normal development. Using temperature shift experiments, it has also been established that Dpp signaling in L2 is required for the proper induction of both Eya and Dll in their respective fields. Gain-of-function analyses show that Dpp is also sufficient to induce Eya expression within the eye field and Dll expression within the antennal field. Clearly, though, Dpp must function in the context of selector factors such as Ey in order to produce two independent primordia within the eye-antennal epithelium. In the presence of Ey, Dpp signaling induces Eya expression as opposed to Dll. The absence of Ey in the antennal field at the time of Dpp signaling is of crucial importance to ensure the proper induction of Dll and subsequent formation of an antenna primordium. Indeed, as described in this study, the restriction of Ey to the eye field precedes the emergence of dpp and Dll in the antennal field during normal development (Kenyon, 2003).

In conclusion, Dpp, unlike Notch, functions as an inducer of tissue identity during specification of the eye-antennal disc, and the spatial and temporal aspect of organ primordia formation is controlled at the level of dpp transcription (Kenyon, 2003).

The effect of field size on eye primordium formation cannot be simply mediated by Dpp but is likely due to the influence of a third signaling system, the Wingless pathway. Wg functions as a negative regulator of eye development and is known to antagonize Dpp signaling in L3 discs. This antagonistic interaction occurs at least in part at the posttranscriptional level and is likely established earlier in development. At the time of onset of Eya expression, early in L2, the sources of dpp and wg are localized to opposing regions of the eye-antennal disc -- dpp along the posterior margin and wg across the dorsal anterior region. Hence, the relative concentration of Dpp and Wg experienced by disc cells likely depends on their location within and the size of the morphogenetic field. Since Dpp induces Eya expression and Wg antagonizes Dpp signaling, field size becomes a critical variable in determining the response to Dpp/Wg signaling and thus influences eye primordium formation (Kenyon, 2003).

This model readily accounts for the changes in Eya expression observed in the various genetic backgrounds. In discs expressing Notch antagonists, dpp and wg are still expressed; however, inhibition of cell proliferation results in a smaller disc and a smaller morphogenetic field. This reduction in size changes the balance between Dpp and Wg signaling resulting in a lack of Eya induction. In this context, stimulation of cell proliferation by CycE increases field size, thus restoring relative levels of Dpp and Wg signaling compatible with Eya induction. A simple prediction of this model is that modification of Dpp/Wg signaling in favor of Dpp should restore Eya expression in small ey-Gal4 UAS-SerDN discs. This was tested by removing one wild-type copy of the wg gene and thus lowering Wg signaling in SerDN-expressing discs. In agreement with the model, Eya expression is significantly rescued in late-L2/early-L3 wg+/+ey-Gal4 UAS-SerDN discs regardless of disc/field size (Kenyon, 2003).

Extracellular signals responsible for spatially regulated proliferation in the differentiating Drosophila eye

Spatially and temporally choreographed cell cycles accompany the differentiation of the Drosophila retina. The extracellular signals that control these patterns have been identified through mosaic analysis of mutations in signal transduction pathways. All cells arrest in G1 prior to the start of neurogenesis. Arrest depends on Dpp and Hh, acting redundantly. Most cells then go through a synchronous round of cell division before fate specification and terminal cell cycle exit. Cell cycle entry is induced by Notch signaling and opposed in subsets of cells by EGF receptor activity. Unusually, Cyclin E levels are not limiting for retinal cell cycles. Rbf/E2F and the Cyclin E antagonist Dacapo are important, however. All retinal cells, including the postmitotic photoreceptor neurons, continue dividing when rbf and dacapo are mutated simultaneously. These studies identify the specific extracellular signals that pattern the retinal cell cycles and show how differentiation can be uncoupled from cell cycle exit (Firth, 2005).

The EGFR holds R2-R5 cells in G1 phase and promotes G2/M progression of other cells during the second mitotic wave (SMW). Earlier regulation is now found to depend on longer-range signaling by the Hh, DPP, and N signals already known to drive the progression of the morphogenetic furrow. These studies exclude other models that show that Hh, Dpp, or N act indirectly by releasing other, cell cycle-specific signals from differentiating cells, or that patterned cell cycle withdrawal or reentry occur independent of extracellular signals, such as by synchronized growth. Instead, specific signals are necessary or sufficient for each aspect of cell cycle patterning (Firth, 2005).

G1 arrest ahead of the morphogenetic furrow depends on posterior-to-anterior spread of Hh and Dpp. Hh is secreted from differentiating cells, starting at column 0 in the morphogenetic furrow. Dpp is transcribed in ~6 ommatidial columns in the morphogenetic furrow in response to Hh. Cells accumulate in G1 about 16-17 cell diameters anterior to column 0, suggesting an effective range of ~13-17 cells for Hh and Dpp (Firth, 2005).

The contribution of Dpp to this cell cycle arrest is known already, but that of Hh was not suspected. Both Dpp and Hh signaling can promote proliferation in other developmental contexts (Firth, 2005).

S phase entry in the SMW depends on another signal, N. Expression of the N ligand Dl begins at the anterior of the morphogenetic furrow. The first S phase cells are detected 6-8 cell diameters more posteriorly, just behind column 0. The transmembrane protein Dl must act more locally or more slowly than the secreted Hh and Dpp proteins, to explain gaps between S phases (Firth, 2005).

Although N activity has been associated with growth through indirect mechanisms involving the release of other secreted growth factors and also regulates endocycles, this appears to be the first report of a specific role of N in G1/S in diploid Drosophila cells. Notably, deregulated N signaling contributes to at least two human cancers and is oncogenic in mice (Firth, 2005).

At the same time that N promotes S phase entry in the SMW, EGFR activity ensures that R2-R5 cells remain in G1. N is still required in the absence of EGFR, so N activity is a positive signal and is not required only to counteract EGFR activity. Instead, EGFR activity interferes with S phase entry in response to N (Firth, 2005).

Ligands for the EGF receptor are thought to be released from R8 precursor cells, although EGFR-dependent MAPK phosphorylation is detected one ommatidial column before the column where R8 precursor cells can be identified, which is in column 0. This means that EGFR activation begins after Dl expression but before S phase DNA synthesis starts. Later, ligands released from differentiating precluster cells activate EGFR in surrounding cells to permit SMW mitosis around columns 3-5 (Firth, 2005).

Hh and Dpp together promote expression of Dl and of EGFR ligands; in part, this occurs indirectly through Atonal and the onset of differentiation. EGF receptor activity also promotes Dl expression (Firth, 2005).

At least three genetic mechanisms arrest distinct retinal cells in G1. Arrest ahead of the morphogenetic furrow depends on Dpp and Hh. During the SMW, R2-R5 cells are held in G1 by EGFR, which counteracts the SMW-promoting N activity. In addition, R8 cells, which are defined by the proneural gene atonal, remain in G1 independent of EGFR. After the SMW, all cells remain in G1 indefinitely, independent of EGFR. Although cell cycle withdrawal roughly correlates with differentiation, many of the cells that arrest after the SMW are still unspecified (Firth, 2005).

Loss of rbf and dap together overcome all cell cycle blocks, even though cell differentiation continues. This redundancy indicates that Cyclin E/Cdk2 targets other than Rbf are needed for proliferation, consistent with many other studies. Dap may be regulated by EGFR in R2-R5 cells. If rbf regulates the normal SMW, where Cyclin E expression seems not to be limiting, then other E2F targets may be involved. Some cell cycle arrest can also be overridden by forced expression of Cyclin E, E2F/DP, dRef, and ORC1, or by mutation of the Cyclin A antagonist rux (Firth, 2005).

The results show that mechanisms that assure both short- and long-term arrest of retinal cells must operate upstream of (or parallel to) Rbf and Cyclin E activities. They might resemble the barriers to transformation and regeneration that exist in mammals (Firth, 2005).

Restricted teashirt expression confers eye-specific responsiveness to Dpp and Wg signals during eye specification in Drosophila

In Drosophila, the eye primordium is specified as a subdomain of the larval eye disc. The Zn-finger transcription factor teashirt (tsh) marks the region of the early eye disc where the eye primordium will form. Moreover, tsh misexpression directs eye primordium formation in disc regions normally destined to form head capsule, something the eye selector genes eyeless (ey) and twin of eyeless (toy) are unable to do on their own. Evidence suggests that tsh induces eye specification, at least in part, by allowing the activation of eye specification genes by the wingless (wg) and decapentaplegic (dpp) signaling pathways. Under these conditions, though, terminal eye differentiation proceeds only if tsh expression is transient (Bessa, 2005).

The specification of the eye primordium within the main epithelium (ME) of L2 eye discs correlates with tsh expression, suggesting that tsh might be involved in this specification. If this is the case, it would be expected that ectopic tsh expression will transform the overlaying squamous layer, the peripodial epithelium (PE) cells into an eye primordium, characterized by: (1) columnar morphology of the epithelial cells; (2) eye-specific gene expression, and (3) eye-specific response to key signaling pathways. Each of these points has been analyzed in turn by inducing the expression of tsh in marked clones of cells in the PE (Bessa, 2005).

Cells expressing tsh in the margin of the disc or in the PE overproliferate, adopt a columnar shape, with elongated nuclei, and are more densely packed than non-expressing cells. Some of these clones further show a sorting behavior, by which the tsh-expressing cells arrange themselves as hollow sacs with their apical sides pointing inwards, as monitored by expression of armadillo/ß-catenin, which localizes to adherens junctions. Such a sorting behavior is usually considered to be the consequence of the cells adopting a new identity (Bessa, 2005).

In order to test if tsh is sufficient to induce eye primordium identity in PE cells, the expression of the eye selector gene ey, as well as that of the early retinal genes eya and Dac, was examined in tsh-expressing clones. tsh-positive cells show increased Ey expression. In addition, PE tsh-expressing clones that lie close to the posterior margin activate eya and the eya target Dac, indicating that these cells adopt an eye primordium-like fate. PE clones overexpressing ey are not able to induce eya, neither are similar toy-expressing clones, in which ey expression is upregulated. In these PE clones, tsh expression is not induced. Therefore, it is concluded that neither ey upregulation nor the joint overexpression of toy and ey are able to re-specify the peripodial epithelium. In addition, overexpression of eya in PE clones do not turn Dac on either, which reinforces the idea that PE re-specification as eye primordium occurs only if tsh is expressed (Bessa, 2005).

Expression of tsh activates eya expression mostly in the center and posterior half of the PE, but not in the anterior half. Clones in this anterior region retain the expression of hth, which is normally expressed in all PE cells. Since dpp and wg are expressed in the domains of the posterior and anterior discs, respectively, it was reasoned that these differences in the response of tsh-expressing cells could be the result of these signaling pathways acting differently in anterior and posterior domains of the PE (Bessa, 2005).

To test this hypothesis, the response of normal PE cells to variations in both wg and dpp pathways was tested. Clones where the dpp pathway was hyperactivated through the expression of a constitutively active dpp-receptor, thick veins (tkvQD), or blocked by removing the signal transducer Mothers against dpp (Mad), showed no induction of eya expression or cell morphology changes. Neither did anterior clones expressing Axin, a negative regulator of the wg pathway or overexpressing wg. Nevertheless, when alterations in the dpp and wg pathways were performed in the presence of ectopic tsh, PE cells showed gene expression responses characteristic of the ME. Thus, whereas posterior tsh-expressing PE cells induce eya expression, tsh-expressing cells in which the dpp pathway has been blocked by removing Mad no longer express eya. Again, this is the behavior exhibited by tsh+ ME cells deprived of dpp signaling. Similarly, while anterior tsh-expressing PE cells retain hth expression, most clones expressing both tsh and Axin lose hth expression, as they do if Axin is expressed in the ME within the tsh domain. PE tsh+ tkv+ clones still fail to activate eya in anterior dorsal and anterior ventral regions, suggesting that even in these clones wg signaling can prevent PE re-specification. Clones of PE cells expressing tsh, tkvQD and Axin now activate eya anywhere in the disc, indicating that, in the presence of tsh, wg and dpp antagonize each other to regulate eya expression. It is noted, however, that the squamous to columnar cell shape change induced by tsh is independent of the activity of the wg and dpp pathways. These results suggest that tsh, when expressed in the PE, can reprogram this epithelial layer to respond to wg and dpp signals such that it develops in an eye primordium-specific manner (Bessa, 2005).

During the development of the eye disc, only cells of the ME will be specified as eye primordium. Although Wg and Dpp signals play essential roles during eye development, PE cells are relatively insensitive to these signaling pathways, as measured by cell survival, morphology, proliferation or gene expression changes. tsh starts being expressed in the ME around the time when the eye primordium is specified, and tsh has the potential to redirect eye disc PE cells towards eye development, an ability the eye selector genes toy and ey do not have on their own. These results indicate that the PE can be re-specified by tsh throughout most of the life of the larva. Thus, tsh-expressing clones induced during L1 and L2 induce eya and Dac expression. The transient expression of tsh during L2, or its induction by Gal4 drivers active during late-L2/L3, results in ectopic PE eyes (Bessa, 2005).

It is proposed that one way in which tsh might be involved in eye fate specification is by altering the response of eye disc cells to Dpp and Wg signals. The molecular mechanisms by which tsh might achieve this during eye development remain to be further investigated, but they might be similar to those already described during embryogenesis, where Tsh modulates wg and dpp pathways directly interacting with Armadillo, the wg signaling transducer, and with Brinker, a transcriptional repressor of the dpp pathway (Bessa, 2005).

In the eye disc, the cells specific response to wg and dpp enabled by tsh is superimposed onto the expression of eye-selector genes. Such combination of factors in turn would specify the eye primordium. The fact that Tsh and Ey have the potential to interact directly makes it possible for Ey to tether Tsh-containing transcriptional complexes to eye-specific targets genes (Bessa, 2005).

It is also observed that ato expression is induced in some of the tsh-overexpressing eye-disc cells. Therefore, tsh has the potential not only to sensitize eye disc cells to wg and dpp signals, but also to make them prone to neural differentiation. dpp and wg have been shown to regulate the spatial activation of ato to position several adult sensory organs, including the eye, within the corresponding imaginal discs. This mechanism for positioning ato would define a sensory organ prototype upon which selector genes, such as ey, would specify the final sensory type. Interestingly, the ectopic ato expression induced by tsh is not disc specific and, if tsh induction is transient, results in ectopic neurons. This ato induction might be mediated by tsh enabling cells to respond to dpp and wg (Bessa, 2005).

These results underlie the importance of the precise and dynamic spatiotemporal pattern of expression of tsh: although tsh expression must be confined to the ME layer of the eye disc, in order for eye development to proceed, tsh has to be first expressed in undifferentiated cells to be later turned off to allow retinal differentiation. The earlier paradox of tsh acting both as eye repressor and inductor, depending on the Gal4 promoters used, can now be explained as follows: Gal4 promoters that are not repressible by the gene expression changes induced upon tsh overexpression, such as ey-GAL4, will lead to sustained expression of tsh and, therefore, to a blockage of eye development. Other drivers that are turned off after tsh expression (i.e., MS1096, MD705) will mimic the situation found in the ME (that is, on/off), and in these cases, eye development will proceed. It is noted that in experiments where ey is ectopically expressed, eyes tend to develop in the proximal parts of appendages which derive from tsh-expressing domains in their respective imaginal discs. This correlation reinforces the idea of tsh as a potential eye-competence factor (Bessa, 2005).

At least three roles for tsh during eye development have been uncovered: promoting proliferation, acting as an eye repressor and acting as an eye inducer. The first two roles (proliferation and eye repression) are linked to the function of the transcription factor Hth. Thus, Tsh and Hth (together with Ey) maintain the eye disc cells in a proliferative, undifferentiated state, which is incompatible with eye differentiation. This state is kept as long as cells express hth, which is positively regulated by wg and repressed by dpp. Since tsh keeps hth on, sustaining tsh expression artificially in the disc blocks further eye differentiation. Once hth is repressed by Dpp signaling close to the MF, cells enter a preproneural state, that still maintains tsh expression, in which dpp activates the expression of retinal genes such as eya. The results suggest that tsh is required for the eye-specific interpretation of Wg and Dpp signals, and therefore for both the maintenance of proliferation and the specification of the retina. This model thus predicts that removal of the earliest tsh function (which corresponds to the most anterior regions in older discs) should result in eye loss due to either lack of proliferation or to the incorrect specification of the primordium; removal of later tsh function (which corresponds to more posterior regions of older discs) should cause a premature derepression of the eye differentiation program and excess of eye. In fact, both phenotypes have been described in tsh loss-of-function clones: eye loss and eye overgrowth. The current experiments, in which tsh function is reduced uniformly from early stages of eye development, agrees with an early role of tsh in eye specification and/or proliferation. This model of tsh function is further complicated by the fact that the dorsoventral genes also impinge on tsh function. Still, some tsh-clones show no phenotype. This might be explained by perdurance of the Tsh product, local differences in the requirement of tsh within the eye disc or the existence of compensatory functions (Bessa, 2005).

toy and ey lay atop the eye specification genetic network in Drosophila. However, neither Toy nor Ey is able to activate the expression of tsh in the PE, and tsh expression in maintained in ey mutant discs. The reverse is also true; tsh upregulates ey expression in the eye disc, but is unable to activate its expression de novo in any other disc. This indicates that tsh expression is regulated independently of the Pax6 genes in the eye disc. This situation is analogous to that of Optix, a Six3 homolog, which is expressed in the eye disc independently of ey with a pattern reminiscent of that of tsh. Nevertheless, Optix does not seem to regulate tsh; ectopic expression of Optix in the eye disc does not trigger tsh expression. Taking into account all these results, it is proposed that tsh functions in parallel to ey (and probably to toy) as an eye competence factor (Bessa, 2005).

Hedgehog and Decapentaplegic pathway function in the developing Drosophila compound eye: smoothened, thickveins and the genetic control of cell cycle and cell fate

The Hedgehog and Decapentaplegic pathways have several well-characterized functions in the developing Drosophila compound eye, including initiation and progression of the morphogenetic furrow. Other functions involve control of cell cycle and cell survival as well as cell type specification. This study used the mosaic clone analysis of null mutations of the smoothened and thickveins genes (which encode the receptors for these two signals) both alone and in combination, to study cell cycle and cell fate in the developing eye. It is concluded that both pathways have several, but differing roles in furrow induction and cell fate and survival, but that neither directly affects cell type specification (Vrailas, 2006).

Interestingly, though Hedgehog signaling is required for Decapentaplegic expression, the two pathways are not completely redundant. The data demonstrate that for some aspects of eye development, the two pathways have separable and independent functions, such as Hedgehog signaling regulation of rough expression and S phase of the second mitotic wave. However, both pathways have redundant roles in the apical constriction of the actin cytoskeleton and proper expression of elements of the Egfr/Ras and Notch/Delta signaling pathways as well as in cell fate specification, though neither pathway is required for differentiation. Finally, the Decapentaplegic pathway is epistatic to the Hedgehog pathway for G1 arrest in the furrow and G1, G2 and M phases of the second mitotic wave. These various ways in which the Hedgehog and Decapentaplegic pathways work together (or not) demonstrate the complexity of pathway integration for proper eye development (Vrailas, 2006).

A strong effect of loss of Hedgehog signaling was seen on the morphology of cells in the furrow, and it is suggested consequentially, in the distribution of the Egfr and Notch receptors. This disruption of the localization of elements of other signaling pathways, which is enhanced by the additional loss of thickveins, may explain some of the phenotypes observed. For example, cells at the edges of smoothened and double mutant clones near wild type tissue are still able to enter S phase. The Notch/Delta pathway has been shown to regulate the G1/S transition of the second mitotic wave with loss of pathway activity leading to a loss of S phase. Therefore, it may be that Notch/Delta signaling between cells in the wild type tissue and in the clone, allows for the S phases seen at the edges of the clones, while in the center of clones, where the Notch/Delta pathway is disrupted, S phase is lost. Cell fate specification can still occur at the edges of smoothened thickveins double mutant clones. It may be that the furrow does not really pass through the double mutant clones, but some signal from outside the clone can still induce photoreceptor cell fate, at least close to the clone margins. This is likely to be Spitz/Egfr signaling, which is present but disrupted in smoothened clones, since this signal can induce photoreceptor fate ectopically even anterior to the furrow and without the formation of R8/founder cells (Vrailas, 2006).

This study reports the roles of Hedgehog and Decapentaplegic signaling in eye development, however, these pathways are also instrumental for patterning and proliferation in the developing wing. Studies in the wing have shown that as in the eye, decapentaplegic expression is downstream of hedgehog, suggesting that these pathways may also rely on each other for proper wing development. Though smoothened and thickveins have no role in ommatidial cell fate, Hedgehog signaling is required for specification of intervein and vein territories in the central region of the wing, and Decapentaplegic signaling has been shown to be required for vein cell fate in the developing pupal wing (Vrailas, 2006).

As in the eye, Hedgehog and Decapentaplegic signaling have been implicated in cell cycle regulation in the developing wing. Studies in the wing found that overexpression of the Hedgehog signal induces proliferation through upregulation of Cyclin D and Cyclin E, as well as specifically promotes S phase in the wing margin. FACS analysis of wing discs revealed that thickveins loss of function clones (tkv7) have a reduced number of cells in S phase and an increase in the number of cells in G1 phase. Additionally, inhibition of the Hedgehog signal results in decreased growth and cell proliferation rates, and loss of Decapentaplegic pathway signaling results in small clones, suggesting that these pathways are important in cell survival and/or proliferation in the wing (Vrailas, 2006).

It appears that both tissues use Hedgehog signaling to promote S phase and possibly cell survival, since inhibiting Hedgehog signaling results in cell death in the eye and decreased growth in the wing. Additionally, the two tissues may use Hedgehog signaling to regulate the G1 phase, though this regulation may have subtle differences. In addition, Decapentaplegic signaling also appears to be necessary for proliferation in the developing eye and wing, though these tissues may use this signal to regulate the cell cycle differently. This is not surprising, since the developing third instar eye and wing discs may have fundamental differences in cell cycle regulation; the eye has a coordinated second mitotic wave and the wing does not. For example, the eye may utilize some factors that are not present in the wing disc to prevent the build up of too much Cyclin E. Therefore, Cyclin E levels are decreased in the eye but not in the wing. Additionally, thickveins appears to be responsible for G1 arrest in the furrow, while in the wing, G1 arrest in the zone of nonproliferating cells is mediated by Wingless signaling. However, it may be that the eye and wing regulate the cell cycle using Hedgehog and Decapentaplegic signaling in much the same way, but the techniques used to examine this phenomenon in the different tissues do not allow for a direct comparison of results. For example, it may be that FACS analysis is a more sensitive technique than immunohistochemistry, and thus subtle changes in the cell cycle that were observed in the wing were not observed in the eye. Alternatively, the FACS analysis was performed on wing discs that contained thickveins clones in a Minute background in order to achieve a larger sample of thickveins mutant cells. However, dying cells, such as those homozygous for Minute mutations, have been shown to have non-autonomous effects on the biology of the surrounding cells in the wing. Indeed, one study has reported that Minute mutations can non-autonomously affect pattering of photoreceptors in the developing eye. It may be that the Minute background partially masked the thickveins cell cycle phenotypes and the eye and wing may not be as different as it initially appears (Vrailas, 2006).

The data also shows that the Hedgehog and Decapentaplegic pathways are only partially redundant in the eye, which has also been shown in the wing. Hedgehog signaling alone is required for specification of veins 3 and 4 and the sensory organ precursors (SOPs) near the anterior/posterior boundary of the developing wing, whereas Decapentaplegic signaling mediated by Hedgehog promotes some SOP formation in the notum and some other regions of the wing (Vrailas, 2006).

In some instances, the data contrasts with previous reports from others. In one case, in which different alleles of smoothened (smo3 versus smoD16) were examined, phenotypic variation may be a result of allele specific effects. However, in another case, the same allele was used by two groups, and it may be that some other aspect of the genetic background of the stocks differed that influenced the results observed. The effects of removing a receptor (Smoothened) may also differ in some cases from those of removing a downstream element (Ci). It was also observed that clones the remove thickveins or smoothened and thickveins together often appear to be re-specified as other structures, resembling appendage discs. This may be due to other functions of the Decapentaplegic pathway on the disc margins and in defining the limits of the eye field. The interpretations of others may have been confounded by such re-specification in some cases. Indeed, in the developing wing, cells lacking Decapentaplegic pathway function actually leave the epithelium. Some care was taken to analyze only those small clones near the center of the eye field that do not have these characteristics. Indeed, the fact that photoreceptor specific markers were observed in some cells that lack both smoothened and thickveins demonstrates that even the double mutant clones do not always re-specify (Vrailas, 2006 and references therein).

In summary, it is concluded that the Hedgehog pathway has important roles in inducing furrow initiation and progression. The Hedgehog and Decapentaplegic pathways have redundant roles in actin constriction in the morphogenetic furrow, expression of Egfr, Notch and Delta, and differentiation with neither pathway essential for cell type specification. Likewise, no role was found for either Hedgehog or Decapentaplegic signaling in ommatidial rotation or chirality. It is also suggested that the Hedgehog pathway alone is required for rough expression and the G1/S transition in the second mitotic wave and provides a protective function against apoptosis. In contrast, the Decapentaplegic pathway appears critical for furrow initiation at the disc margins (but not progression in the center). In addition, the Decapentaplegic pathway is epistatic to Hedgehog signaling for maintenance of G1 arrest in the furrow and regulation of G1 phase and the G2/M transition in the second mitotic wave (Vrailas, 2006).

Polychaetoid appears to act in concert with the Dpp pathway to regulate select core components of the AJ during development

Correct cellular patterning is central to tissue morphogenesis, but the role of epithelial junctions in this process is not well-understood. The Drosophila pupal eye provides a sensitive and accessible model for testing the role of junction-associated proteins in cells that undergo dynamic and coordinated movements during development. Mutations in polychaetoid (pyd), the Drosophila homologue of Zonula Occludens-1, are characterized by two phenotypes visible in the adult fly: increased sensory bristle number and the formation of a rough eye produced by poorly arranged ommatidia. It was found that Pyd is localized to the adherens junction in cells of the developing pupal retina. Reducing Pyd function in the pupal eye results in mis-patterning of the interommatidial cells and a failure to consistently switch cone cell contacts from an anterior-posterior to an equatorial-polar orientation. Levels of Roughest, DE-Cadherin and several other adherens junction-associated proteins are increased at the membrane when Pyd protein is reduced. Further, both over-expression and mutations in several junction-associated proteins greatly enhances the patterning defects caused by reduction of Pyd. These results suggest that Pyd modulates adherens junction strength and Roughest-mediated preferential cell adhesion (Seppa, 2008).

The data demonstrate that Pyd is an AJ-associated protein that is required for patterning of the pupal lattice cells. Live imaging of the developing eye indicates that Pyd is necessary for the directed movements of interommatidial precursor cells (IPCs) that allow cell sorting into defined niches. Membrane contacts are dynamically exchanged in the pupal eye: each shift in the position of a cell requires the removal of previous contacts and the establishment of new ones. Pyd regulates patterning at least in part through modulating levels of the AJ-associated proteins DE-Cadherin, β-Catenin, and α-Catenin. Other studies have suggested that cell adhesion is necessary both to facilitate and restrict cell movement within the eye epithelium; the interplay between these two processes requires tight regulation of the levels of both cell adhesion molecules and junctional proteins. The data indicate that removal of Pyd from the AJ compromises this tightly-regulated system and biases the cells toward poorly-directed movements, perhaps because of dysregulation of the timing or function of the mechanisms that control the stability of AJ proteins. This failure in precise regulation of adhesion was also highlighted in the inability of cone cells to exchange their membrane contacts: the apical interfaces of pyd-RNAi expressing cone cells were locked in place. Ectopic DE-Cadherin further increased the percentage of ommatidia affected, again emphasizing the link between pyd activity and the AJ (Seppa, 2008).

The localization of Pyd to the AJ in the pupal eye is dependent on both DE-Cadherin and α-Catenin. However, it was found that ectopic expression of either junctional protein is not sufficient to alter the localization of Pyd. Taken together, these data indicate that DE-Cadherin and α-Catenin are necessary to build or maintain the AJ and to localize Pyd but that, in excess, they are not sufficient to attract ectopic Pyd. This suggests that either Pyd protein levels are not easily altered or that Pyd may be binding to proteins other than the core AJ constituents. Recent work demonstrated that E-Cadherin was necessary for the initial steps of AJ formation while α-Catenin was essential for both the establishment and maintenance of the junction; only when α-Catenin was reduced was ZO-1 lost from established junctions. The results suggest that in dynamically restructured tissues such as the eye, both E-Cadherin and α-Catenin are necessary for the localization of AJ-associated proteins (Seppa, 2008).

The immunoglobulin superfamily member Roughest is necessary for appropriate sorting of IPCs during pupal eye development. Reducing Pyd increased Roughest protein levels specifically at the AJ. Roughest is the Drosophila orthologue of Neph1, a cell adhesion molecule necessary for the structure and function of the glomerular slit diaphragm in the mammalian kidney. The slit diaphragm is the main size-selective barrier in the filtration apparatus of the kidney and retains many characteristics of both the tight and AJ complexes from which it was derived. The Hibris orthologue Nephrin also forms part of the physical structure of the slit diaphragm and both cell adhesion molecules have been reported to bind to each other as well as to ZO-1. Perhaps ZO-1, as with Pyd, has a role in regulating the localization or levels of cell adhesion molecules such as Neph1 and Nephrin (Seppa, 2008).

The Dpp pathway has emerged as a major contributor to patterning of the Drosophila pupal eye. Its role requires functional connections to both DE-Cadherin and Roughest. For example, mutations in shotgun (the locus that encodes DE-Cadherin) suppresses the roughest eye phenotype but enhances Dpp pathway-dependent phenotypes in the eye and wing. Together, these data suggest a model in which (1) Roughest acts to promote the stability of membrane contacts to drive directed cell movements and (2) the Dpp pathway and Pyd act to destabilize the adherens junction complex and local cell contacts to allow for proper IPC sorting. Consistent with this view, it was observed that reducing pyd enhances the effects of reduced Dpp pathway activity in the eye and wing. Thus, Pyd appears to act in concert with the Dpp pathway to regulate select core components of the AJ during development (Seppa, 2008).

This study has shown that Pyd is required specifically for patterning the interommatidial cells of the Drosophila pupal eye. Pyd appears to regulate both cell shape and cell positioning by controlling the levels of AJ proteins such as DE-Cadherin and adhesion proteins such as Roughest. Thus, Pyd provides a link between adhesion and junction formation; a further understanding of its role in the pupal eye will shed light on how these processes are coordinated to generate precise cellular movements during epithelial patterning (Seppa, 2008).

Transformation of eye to antenna by misexpression of a single gene

In Drosophila, the eye and antenna originate from a single epithelium termed the eye-antennal imaginal disc. Illumination of the mechanisms that subdivide this epithelium into eye and antenna would enhance understanding of the mechanisms that restrict stem cell fate. This study shows that Dorsal interacting protein 3 (Dip3), a transcription factor required for eye development, alters fate determination when misexpressed in the early eye-antennal disc, and this observation has been taken advantage of to gain new insight into the mechanisms controlling the eye-antennal switch. Dip3 misexpression yields extra antennae by two distinct mechanisms: the splitting of the antennal field into multiple antennal domains (antennal duplication), and the transformation of the eye disc to an antennal fate. Antennal duplication requires Dip3-induced under proliferation of the eye disc and concurrent over proliferation of the antennal disc. While previous studies have shown that overgrowth of the antennal disc can lead to antennal duplication, these results show that overgrowth is not sufficient for antennal duplication, which may require additional signals perhaps from the eye disc. Eye-to-antennal transformation appears to result from the combination of antennal selector gene activation, eye determination gene repression, and cell cycle perturbation in the eye disc. Both antennal duplication and eye-to-antennal transformation are suppressed by the expression of genes that drive the cell cycle providing support for tight coupling of cell fate determination and cell cycle control. The finding that this transformation occurs only in the eye disc, and not in other imaginal discs, suggests a close developmental and therefore evolutionary relationship between eyes and antennae (Duong, 2008).

Dip3 is able to bind DNA in a sequence specific manner and activate transcription directly. Dip3 possesses an N-terminal MADF domain and a C-terminal BESS domain, an architecture that is conserved in at least 14 Drosophila proteins, including Adf-1 and Stonewall. The MADF domain directs sequence specific DNA binding to a site consisting of multiple trinucleotide repeats, while the BESS domain directs a variety of protein-protein interactions, including interactions with itself, with Dorsal, and with a TBP-associated factor (Bhaskar, 2002).

Antagonism between the N and EGFR signaling pathways influences developmental fate in the eye-antennal disc leading to a loss of eye tissue and the appearance of extra antennae. Although this phenotype was originally suspected to represent eye-to-antennal transformation, subsequent analysis suggests that it most likely represents antennal duplication. Specifically, the absence of the N signal leads to a failure in eye disc proliferation resulting in compensatory over-proliferation of the antennal disc and its subdivision into multiple antennae. Consistent with the idea that the extra antennae result from under-proliferation of the eye field, it was found that the phenotype was largely suppressed by over-expression of CycE to drive the cell cycle (Duong, 2008).

In this study, it was found that inhibition of eye disc growth leads to antennal duplication. But in addition, it was shown that the same treatment that leads to antennal duplication can also direct the transformation of eyes to antennae. These two phenotypes are anatomically distinct. This anatomical distinction is evident in adults: antennae resulting from antennal duplication are found anterior to the antennal foramen, while the antennae resulting from eye-to-antenna transformation are found posterior to the antennal foramen. It is also apparent in larval eye-antennal imaginal discs: antennal duplication discs exhibit multiple circular dac expression domains within a single sac of epithelium (the antennal disc), while eye-to-antennal transformation discs exhibit two or more circular dac expression domains spread over both the eye and antennal discs. The two types of discs display distinct molecular signatures as well: the antennal duplication discs exhibit duplicated Dll expression domains, while the eye discs undergoing transformation to antennae lack Dll expression (Duong, 2008).

Perhaps the most persuasive evidence that Dip3 can direct eye-to-antennal transformation is provided by the observation of eyes that are only partially transformed to antennae since is very difficult to reconcile these partial transformations with the idea of antennal duplication. In some cases, proximal antennal segments tipped with eye tissue are observed. In accord with this phenotype, some third instar larval eye discs display a central domain of Elav-positive differentiating photoreceptors surrounded by a circular dac domain (Duong, 2008).

These arguments support the idea that antennal duplication and eye-to-antennal transformation are mechanistically distinct phenomena, and the remainder of the discussion assumes this to be the case. However, the possibility that these two phenotypes are two manifestations of a single mechanism cannot be excluded. For example, the discs exhibiting duplicated Dll domains may represent complete transformations, while the discs lacking duplicated Dll domains, but containing Elav may represent partial transformations (Duong, 2008).

The data show that discs undergoing antennal duplication as a result of Dip3 expression are comprised of a severely diminished eye region and an enlarged antennal region. As shown by BrdU labeling experiments, these antennal duplication discs most likely result from suppression by Dip3 of cell proliferation in the eye field leading to overproliferation of the antennal disc. This conclusion is supported by the ability of factors that drive cell proliferation (e.g., Cyclin E) to alleviate the Dip3 misexpression defect (Duong, 2008).

Many experimental manipulations that reduce the size of the eye disc (e.g., surgical excision, induction of cell death, or suppression of cell proliferation) lead to enlargement and duplication of the antennal primordium. How might reduction of the eye field lead to antennal field over-growth? One possibility is that the eye field produces a growth inhibitory signal. Alternatively, the eye field and the antennal field may compete with each other for limited nutrients or growth factors. In support of this latter possibility, recent studies of the role of dMyc in wing development have demonstrated growth competition between groups of imaginal disc cells (Duong, 2008).

While the results imply that antennal disc overgrowth is required for antennal duplication, overgrowth is thought not to be sufficient for duplication. This conclusion derives from experiments in which an antennal disc specific driver is used to direct over-expression of CycE or Nact. This resulted in antennal overgrowth without concurrent reduction in the eye disc. In this case, antennal duplication was not observed. Thus, in addition to antennal overgrowth, antennal duplication also appears to require reduction or elimination of the eye disc. Regulatory signals from the eye disc may act to prevent antennal duplication (Duong, 2008).

The eye and antenna discs differ in several respects: (1) Specific expression of the organ-specification genes. The eye disc expresses the retinal determination gene network (RDGN) genes, including eyeless (ey), twin of eyeless (toy), eyes absent (eya), sine oculis (so), and dachshund (dac), while the antennal disc expresses Dll and hth. hth is also expressed in the eye disc but in a distinct pattern from that seen in the antennal disc. In the second instar eye disc, hth is expressed throughout the eye disc, and collaborates with ey and teashirt (tsh) to promote cell proliferation. The hth expression domain later retracts to only the anterior-most region of the eye disc. This pattern is different from the circular expression pattern observed in the antennal disc. (2) In the antennal disc, dpp is expressed in a dorsal anterior wedge and wg is expressed in a ventral anterior wedge. The intersection of Dpp and Wg signaling is required to specify the proximodistal axis in the leg and antenna. In the early eye disc, Wg and Dpp signaling may overlap. But as the disc grows in size, the wg and dpp expression domain are separated, so that there is probably no intersection between high levels of Wg and Dpp signaling. (3) Whereas the partial overlap of Dll and hth expression domains in the antennal disc is important for proximodistal axis specification, there is no Dll expression in the eye disc. Dll expression in the center of the antennal and leg discs is induced by the combination of high levels of Dpp and Wg signaling. Because there is no overlap of Dpp and Wg signaling in the eye disc, Dll is not induced (Duong, 2008).

Therefore, efficient transformation of the eye disc into an antennal disc requires at least three things: (1) repression of the eye fate pathway; (2) activation the antennal fate pathway; and (3) the intersection of Dpp and Wg signaling, mimicking the situation in the antenna and leg disc that induces proximodistal axis formation. Any one of these three conditions by itself is not sufficient. (1) Loss of the RDGN genes leads only to the loss of the eye. However, if apoptosis is blocked, or cell proliferation is induced, in the ey2 mutant (ey>p35 or ey>Nact in ey2), then Dll can be induced in the eye disc and extra antenna are formed. The induction of Dll is not ubiquitous in the eye disc, suggesting that the loss of ey does not autonomously lead to the expression of Dll and the transformation to the antennal fate. (2) Simply expressing the antennal determining genes Dll or hth in the eye disc does not change the eye fate into antennal fate. It was found that uniform expression of Dll in the eye disc (ey>Dll) resulted in mild eye reduction, whereas ey>hth completely abolished eye development. E132>Dll caused the formation of small antenna in the eye in about 46% of flies, whereas ptc>Dll and C68a>Dll induced extra antenna but not within the eye field. Therefore, although Dll and hth are important determinants for antennal identity, it is their specific spatial expression patterns that determine antennal development. (3) Creating the intersection of Wg and Dpp signaling does not change the eye into antenna. Such manipulation in the leg disc turned on vg and transdetermined the leg disc into wing disc. Therefore, the specific genes induced by Dpp and Wg signaling may depend on disc-specific factors. In the eye disc, turning on Wg signaling in the dpp expressing morphogenetic furrow only blocked furrow progression (Duong, 2008).

In this study, it was found that the ectopic expression of a single gene, Dip3, can cause eye-to-antenna transformation. Dip3 apparently satisfied all three requirements. (1) Overexpression of Dip3 repressed (non-cell-autonomously) ey and dac. The repression of ey may be due to the induction of ct. The ability of Dip3 to simultaneously repress multiple retinal determination genes is completely consistent with the many known cross-regulatory interactions between these genes. (2) ey>Dip3 turned on ct and hth. (3) By blocking cell proliferation, ey>dip3 reduced the eye field size and allowed the intersection of Dpp and Wg signaling. Furthermore, ey>Dip3 induced en, which probably created an ectopic A/P border and induced ectopic dpp/wg expression (Duong, 2008).

Interference with cell cycle progression appears to be a common link between the two phenotypes described in this study. In the case of antennal duplication, interference with eye disc growth leads to antennal disc overgrowth, which is a prerequisite for duplication. In the case of eye-to-antenna transformation, eye disc undergrowth allows the required intersection between Dpp and Wg signaling (Duong, 2008).

The observation that Dip3 misexpression can transform the eye field, but not other tissues, to an antennal fate suggests a close evolutionary relationship between the eye and the antenna. Previous studies have emphasized the homology between antennae and legs. The findings presented here that misexpression of a single transcription factor, namely Dip3, can transform eyes to antennae provides support for the notion that the eye and antenna may also, in some sense, be homologous to one another. Previous evidence in support of this idea comes from the observation that similar spatial arrangements of Wg and Dpp signaling along with a temporal cue provided by the ecdysone signal are required for the formation of the eye and the mechanosensory auditory organ. Small mechanosensory sensilla, such as Johnston's organ and the chordotonal organs (stretch receptors) are thought to represent the earliest evolving sense organs. Perhaps the eye resulted from a duplication and specialization of such a sensillum (Duong, 2008).

Dpp and Tracheal Development

In the developing tracheal system of Drosophila, six major branches arise by guided cell migration from a sac-like structure. The chemoattractant Branchless/FGF (Bnl) appears to guide cell migration and is essential for the formation of all tracheal branches, while Decapentaplegic signaling is strictly required for the formation of a subset of branches, the dorsal and ventral branches. Using in vivo confocal video microscopy, it has been found that the two signaling systems affect different cellular functions required for branching morphogenesis. Bnl/FGF signaling affects the formation of dynamic filopodia, possibly controlling cytoskeletal activity and motility as such, and Dpp controls cellular functions allowing branch morphogenesis and outgrowth (Ribeiro, 2002).

To investigate possible dynamic cell shape changes accompanying tracheal cell movement in vivo and link them to the different signaling systems, three-dimensional reconstructions were used of confocal images of living embryos expressing different GFP-tagged proteins in the developing tracheal system. Expression of GFP-actin, driven in tracheal cells by the btl-Gal4 driver line, revealed fine cellular protrusions from cells at the tip of growing branches after initiation of germ band retraction when migration starts. Such cell extensions are most prominently observed in the developing dorsal and ganglionic branches as well as in the dorsal trunk anterior and posterior. During the early stages of branch outgrowth, these cellular extensions were generally short and relatively few in number (Ribeiro, 2002).

In order to visualize possible cell shape changes during later migratory phases, a GFP protein fused to the myristilation site of the Src protein was expressed under the indirect control of the btl enhancer. This GFP fusion protein labels cellular membranes and thus traces the outline of tracheal cells. Three-dimensional reconstruction of dorsal branches using a stack of optical sections through a living embryo expressing this construct revealed that the two leading cells form numerous membranous extensions in all directions; extensions from more proximal cells of the dorsal branch or from cells of the dorsal trunk were only seen very rarely. To ascertain that these membranous extensions contain actin, embryos of the same developmental stage expressing the GFP-actin construct were also examined. Clearly, a similar network of cell extensions was also discernable with actin-coupled GFP. The diameter of these extensions was in the range of 0.3 to 0.4 µm (Ribeiro, 2002).

To investigate the dynamics of the formation of these cellular extensions, a time-lapse confocal analysis was performed of actin cytoskeletal activity in tracheal cells during the migration process, with special emphasis on dorsal and ganglionic branches. In both cases, actin-containing extensions were seen most prominently in the cells at the tip of the branches. Each of the two leading cells in the dorsal branches formed numerous dynamic cellular outgrowths. In the ganglionic branches, cell extensions were most prominently seen in the single leading cell. The formation of cell extensions was extremely dynamic and their topology changed dramatically with time. Some extensions were found to be short lived; others were more stable and rather long (up to 20 µm). It is concluded from these data that tracheal cell migration is accompanied by the formation of thin, dynamic actin-containing cell extensions, referred to here as filopodia (Ribeiro, 2002).

The formation of tracheal branches via directed cell migration requires input from other signaling systems in addition to Bnl/FGF. Activation of the Dpp signal transduction cascade is essential in dorsal and ventral tracheal cells prior to migration for the subsequent formation of dorsal and ventral (ganglionic and lateral trunk anterior and posterior) branches. In the absence of the Dpp receptors Thick veins (Tkv) or Punt (Put), dorsal branches completely fail to develop and ventral branches are strongly affected. Dpp induces the expression of the genes kni and knrl in the ventral and dorsal cells of the placode; in the absence of these two nuclear proteins, dorsal branches are absent and ventral branches are strongly abnormal (Ribeiro, 2002).

Knowing that Bnl/FGF acts as a chemoattractant for tracheal cells, and having shown above that Bnl/FGF signaling induces filopodial activity, one must wonder why cells need input from the Dpp signaling cascade for a directed movement to the Bnl/FGF source. Is the Dpp response a prerequisite for the subsequent induction of filopodia by Bnl/FGF? Or do dorsal branch cells respond to Bnl/FGF with the formation of filopodia even in the absence of Dpp signaling input, yet fail to migrate properly? In order to find out how these different signaling systems interact in vivo, the cytoskeletal activity of tracheal cells was examined in the absence of Dpp signaling, with particular emphasis on dorsal branches. However, both tkv and put mutants lack dorsal expression of bnl; therefore, they not only lack the Dpp signaling input but also the Bnl/FGF signaling input. In line with the absence of dorsal bnl expression, cellular extensions were not observed in dorsal tracheal cells in put mutants when analyzed in vivo using the GFP-actin fusion protein (Ribeiro, 2002).

To circumvent the problem of the absence of dorsal bnl expression in mutants defective in Dpp signaling, use was made of the inhibitory SMAD protein encoded by the Drosophila Daughters against dpp (Dad) gene. Specific inhibition of Dpp signaling in tracheal cells via trachea-specific ectopic expression of Dad led to the absence of dorsal branches, despite the presence of bnl expression on the dorsal side of the embryo. Consistent with the absence of dorsal branches upon ectopic expression of Dad, kni expression was not detectable in dorsal tracheal cells. Loss of dorsal branches was also readily visible in the later larval stages; no dorsal branches were observed in third instar larvae upon the expression of Dad in the tracheal system during embryogenesis. In embryos and in larvae expressing Dad, stump-like dorsal outgrowths were occasionally observed at positions where dorsal branches form in wild-type animals. It is argued that these stumps are misrouted dorsal trunk outgrowths; such outgrowths are never observed in tkv or put mutants, presumably due to the lack of bnl expression dorsal to the invaginating placode. It is concluded from these experiments that ectopic expression of Dad mimics the tkv and put mutant phenotypes with regard to the lack of dorsal branch formation, and that dorsal branches fail to form through guided cell migration in this particular Dpp loss-of-function situation despite the presence of dorsal bnl expression (Ribeiro, 2002).

To investigate the possible cell shape changes or cytoskeletal rearrangements in dorsal tracheal cells in the absence of Dpp signaling in vivo, confocal imaging was performed of living embryos expressing both a Dad and a GFP-tagged actin transgene in the developing trachea. Confirming the observations made in fixed embryos and in third instar larvae, the phenotype observed in late embryonic stages (stages 15 and 16) in vivo is the complete absence of dorsal branches. However, analysis of a time-lapse study of three-dimensional reconstructions, in which tracheal GFP-actin dynamics were recorded in an interval of 5 min for 135 min, revealed a strikingly different picture. Unlike put mutants, embryos in which Dpp signaling is inhibited specifically in tracheal cells by ectopic expression of Dad clearly show dorsal outgrowths and filopodial activities in positions where dorsal branches normally form. These outgrowths look bud-like and showed dynamic filopodial extensions, but never refine to single-cell diameter, tubular dorsal branches. Although tracheal cells migrated dorsally, they never migrated over a large distance, and in most cases all the cells forming these buds eventually reintegrated into the main dorsal trunk, leading to a general absence of dorsal branches (Ribeiro, 2002).

These results demonstrate that in the absence of Dpp signaling, tracheal cells close to the dorsal bnl-expressing ectodermal cells are able to form actin-containing filopodial extensions and initiate dorsal migration. However, the lack of Dpp signaling, which results in the lack of expression of the kni/knrl target genes, leads to failure to form a dorsal branch, and the short, bud-like dorsal outgrowths eventually reintegrate into the main dorsal trunk. Consistent with this interpretation, cells forming the initial dorsal outgrowth in Dad-expressing embryos in rare cases generated a dorsal trunk-sized lumen. These dorsally directed stumps of dorsal trunk were also visible in third instar larvae. Such dorsal trunk-like buds are also seen in mutants that lack Dpp-induced kni/knrl in the tracheal system, indicating that dorsal migration also takes place in these mutants. These buds are never observed in put mutants, presumably due to the lack of dorsal expression of the chemoattractant Bnl/FGF (Ribeiro, 2002).

These results demonstrate that while Bnl/FGF signaling is necessary and sufficient for the induction of filopodial activity in tracheal cells and for cell migration in the strictest sense (cells do start to migrate dorsally when Dpp signaling is inhibited by Dad), Bnl/FGF is apparently not sufficient to allow productive dorsal branch outgrowth. For dorsal branches to grow out and form, Dpp signaling input is strictly required, in addition to filopodial activity induced by Bnl/FGF. Thus, Dpp signaling does not appear to collaborate with Bnl/FGF in filopodia production and motility, but instead to target cellular functions distinct from those targeted by Bnl/FGF signaling. Thus, despite the essential and crucial role of Bnl/FGF, chemoattraction is not sufficient for successful tracheal branching and, despite the requirement of Dpp for dorsal branch formation, migration per se is not affected (Ribeiro, 2002).

A number of potential, possibly overlapping, roles for Dpp signaling in dorsal branch formation are envisaged. (1) Dpp might induce branch-specific cell rearrangements allowing the formation of an extended unicellular tube via cell intercalation. In the absence of this information, branch elongation cannot take place and the 5-7 cells that would line up to form the dorsal branch under normal conditions start to migrate dorsally as a cell group, adopt dorsal trunk identity, and later reintegrate into the resident dorsal trunk. (2) Dpp signaling might influence adhesion among tracheal cells, generating groups of cells with higher affinity for each other. It is possible that cell movement during branching morphogenesis is the result of a balance between the forces generated by Bnl/FGF-induced forward cell migration and the forces generated by adhesive properties among neighboring tracheal cells. Reintegration of the dorsal bud formed in the absence of Dpp signaling into the dorsal trunk could be the result of the higher affinity of the respecified dorsal trunk cells for their own dorsal trunk 'affinity group', which might be stronger than the force generated by the Bnl/FGF chemoattractant. (3) Dpp signaling could alter the adhesion between tracheal cells and migratory substrates, for example by changing the selective adhesion of extending filopodia to distinct target regions. Such a model would also be consistent with the demonstrated capability of Dpp to redirect cells toward dorsal migratory behavior (Ribeiro, 2002).

Evidence for a fruit fly hemangioblast and similarities between lymph-gland hematopoiesis in fruit fly and mammal aorta-gonadal-mesonephros mesoderm

The Drosophila lymph glands is a hematopoietic organ and, together with prospective vascular cells (cardioblasts) and excretory cells (pericardial nephrocytes), arises from the cardiogenic mesoderm. Clonal analysis provided evidence for a hemangioblast that can give rise to two daughter cells: one that differentiates into heart or aorta and another that differentiates into blood. In addition, the GATA factor gene pannier (pnr) and the homeobox gene tinman (tin), which are controlled by the convergence of Decapentaplegic (Dpp), fibroblast growth factor (FGF), Wingless (Wg) and Notch signaling, are required for the development of all cardiogenic mesoderm, including the lymph gland. An essential genetic switch differentiates between the blood or nephrocyte and vascular lineages involves the Notch pathway. Further specification occurs through specific expression of the GATA factor Serpent (Srp) in the lymph-gland primordium. These findings suggest that there is a close parallel between the molecular mechanisms functioning in the Drosophila cardiogenic mesoderm and those functioning in the mammalian aorta-gonadal-mesonephros mesoderm (Mandal, 2004).

Blood and vascular cells in the vertebrate embryo are thought to derive from oligopotent progenitor cells, called hemangioblasts, that arise in the yolk sac and in the aorta-gonadal-mesonephros (AGM) mesenchyme. A close relationship between blood and vascular progenitors is well established, but in vivo evidence that a single cell can divide to produce a blood cell and an endothelial cell is lacking in vertebrate systems. Similarly, the molecular mechanism that distinguishes between the two lineages is not well understood. To address these issues in a simple, genetically amenable system, the genetic control of hematopoiesis was analyzed in Drosophila. The results show that there are close lineage relationships between hematopoietic and vascular cells, similar to those present in the AGM of mammalian systems. Evidence is provided for conserved cassettes of transcription factors and signaling cascades that limit the pool of hemangioblastic cells and promote the blood versus vascular fate (Mandal, 2004).

In the mature Drosophila embryo, the lymph gland is formed by a paired cluster of ~20 cells flanking the aorta. The aorta and heart represent a contractile tube lined by a layer of myoepithelial vascular cells called cardioblasts. The cells flanking the aorta and heart posterior to the lymph gland are the pericardial cells, which function as excretory cells (nephrocytes). Lymph gland progenitors express the prohemocyte marker Srp and ultrastructurally resemble prohemocytes that develop at an earlier stage from the head mesoderm. Monitoring expression of the zinc-finger protein Odd-skipped (Odd) shows that the lymph gland originates from the dorsal thoracic mesoderm. Odd is expressed in segmental clusters in the dorsal mesoderm of segments T1-A6. The three thoracic Odd-positive clusters coalesce to form the lymph gland, whereas the abdominal clusters formed the pericardial nephrocytes (Mandal, 2004).

Lymph-gland progenitors, cardioblasts and pericardial cells are closely related by lineage. Labeled 'flipout' (FLP/FRT) clones were induced in embryos aged 3-4 h such that the clones contained only 2-4 cells. Of the two-cell clones, ~50% contained cardioblast and lymph-gland cells; the other clones comprised either cardioblasts or lymph-gland cells alone. Mixed clones were recovered at the late third larval stage. The finding of mixed clones indicates that the cardiogenic mesoderm of D. melanogaster contains oligopotent progenitors that, up to the final division, can give rise both to Srp-positive blood-cell progenitors that form the lymph gland and to vascular cells (Mandal, 2004).

The cardiogenic mesoderm forms part of the dorsal mesoderm, which requires the homeobox protein Tin and the GATA factor Pnr. In embryos with mutations in tin or pnr, the lymph gland was absent. Maintenance of Tin expression in the dorsal mesoderm requires the activity of at least two signaling pathways regulated by Dpp (the Drosophila homolog of transforming growth factor-ß) and Heartless (Htl; one of the D. melanogaster homologs of the FGF receptor); the dependence of cardioblast and pericardial nephrocyte development on these signaling pathways has been documented. Lymph-gland progenitors did not develop in loss-of-function dpp and htl mutants (Mandal, 2004).

Between 6 h and 8 h of development, the dorsal mesoderm splits into the cardiogenic mesoderm and the visceral mesoderm. The cardiogenic mesoderm is regulated positively by Wg and negatively by Notch. Lack of Wg signaling results in the absence of all cardiogenic lineages including lymph gland. Notch signaling has the opposite effect and restricts cardiogenic mesodermal fate. Notch is active in the dorsal mesoderm from 6 h to 10 h of development. Eliminating Notch during the first half of this interval by raising embryos homozygous with respect to the temperature-sensitive allele Nts1 at the restrictive temperature resulted in substantially more cardioblasts, pericardial cells and lymph-gland progenitors (Mandal, 2004).

Lymph-gland progenitors, cardioblasts and pericardial nephrocytes are specified in the cardiogenic mesoderm around the phase of germband retraction 8-10 h after fertilization. At this stage, Tin, which was initially expressed in the whole cardiogenic mesoderm, becomes restricted to a narrow medial compartment containing the cardioblasts. Pnr follows the same restriction. Cells located at a more lateral level in the cardiogenic mesoderm give rise to lymph-gland progenitors (in the thoracic domain) and pericardial nephrocytes (in the abdominal domain) and activate the gene odd. Slightly later, Srp is expressed in lymph-gland progenitors. As reported for the early hemocytes derived from the embryonic head, srp is centrally involved in lymph-gland specification. In srp-null embryos, Odd-expressing cells still formed a lymph gland-shaped cluster flanking the aorta, but these cells also express the pericardial marker pericardin (Prc), suggesting that they lose some aspects of hemocyte precursor identity or gain properties of nephrocytes. As a countercorrelate, ectopic expression of Srp in the whole cardiogenic mesoderm directed by mef2-Gal4 induces pericardial cells to adopt lymph-gland fate (Mandal, 2004).

Downregulation of tin and pnr in cells in the lateral domain of the cardiogenic mesoderm is essential for lymph-gland specification. Ectopic expression of tin or pnr by twist-Gal4 (or mef2-Gal4) causes a marked reduction in the number of lymph-gland and pericardial cells. The antagonistic effect of tin on lymph-gland progenitors resembles its earlier role in the head mesoderm that gives rise to the larval blood cells; here too, ectopic expression of tin causes a reduction in the number of hemocytes (Mandal, 2004).

Inhibiting tin and upregulating odd and srp requires input from the Notch signaling pathway. A function of Notch at 6-8 h in specification of the cardiogenic mesoderm is described. Reducing Notch function between 8 h and 10 h causes an increase in the number of cardioblasts and a concomitant loss of pericardial and lymph-gland cells. Overexpressing an activated Notch construct causes a marked increase in lymph-gland size. This late requirement for Notch signaling is separable from the earlier role of Notch in restricting the overall size of the cardiogenic mesoderm. Thus, the sum total of cardioblasts and pericardial or lymph-gland cells in Nts1 embryos shifts between 8 h and 10 h and does not differ substantially from that in wild type, whereas a combined effect on cell number and cell fate is seen in embryos with a Notch deletion. In these embryos, the cardiogenic mesoderm is hyperplasic and develops as cardioblasts at the expense of lymph-gland progenitors and pericardial nephrocytes. The dual role of Notch in restricting the numbers of a pluripotent progenitor pool and in distinguishing between the progeny of these progenitors is reminiscent of the function of Notch in sense-organ development (Mandal, 2004).

Lymph-gland formation is restricted to the thoracic region by positional cues that are provided by expression of the homeobox proteins of the Antennapedia and Bithorax complex. Specifically, Ultrabithorax (Ubx), which is expressed in segments A2-A5 of the cardiogenic mesoderm, inhibits lymph-gland formation. Loss of Ubx results in the expansion of the lymph-gland fate into the abdominal segments. Conversely, overexpression of Ubx driven by mef2-Gal4 causes the transformation of lymph-gland progenitors into pericardial nephrocytes (Mandal, 2004).

These findings are suggestive of a model of lymph-gland development in Drosophila that is similar to mammalian hematopoiesis. Lymph-gland progenitors develop as part of the cardiogenic mesoderm that also gives rise to the vascular cells (aorta and heart) and to excretory cells. Similarly, progenitor cells of the blood, aorta and excretory system are closely related both molecularly and developmentally in mammals, where they form part of the AGM. Specification of the cardiogenic mesoderm requires the input of FGF and Wg signaling, as in vertebrate hematopoiesis, where the AGM region is induced in response to several converging signaling pathways including FGF, BMP and Wnt (Mandal, 2004).

The cardiogenic mesoderm in Drosophila evolves from the dorsal mesoderm and requires input from the Htl, Dpp, Wg and Notch (N) signaling pathways. The cardiogenic mesoderm then differentiates into lymph gland, vascular cells (cardioblasts) and excretory cells (pericardial nephrocytes). A subpopulation of cardioblasts and lymph-gland cells is derived from one progenitor (hemangioblast; HB). Essential for the differentiation of the cardiogenic mesoderm is the Notch-Delta (Dl)-dependent restriction of Tin and Pnr to cardioblasts and the expression of Srp in the lymph gland. In vertebrates, similar cell types are derived from a mesodermal domain called the AGM, which also requires the input of FGF, BMP and Wnt signaling. A subset of AGM-derived cells has been proposed to constitute hemangioblasts, which produce blood progenitors and endothelial cells (Mandal, 2004).

These findings show that in Drosophila, the cardiovascular and blood-cell lineages are differentiated by an antagonistic relationship between Tin or Pnr expression in the cardioblasts and Srp expression in the lymph-gland progenitors. In vertebrates, GATA factors also have a pivotal role in specifying different lineages among blood-cell progenitors, although not much is known about what differentiates between blood progenitors as a group and endothelial progenitors. The results indicate that this step is driven by input from the Notch signaling pathway. In the thoracic cardiogenic mesoderm, Notch antagonizes tin and pnr expression and aortic cardioblast formation, and promotes srp expression and the development of lymph-gland progenitors. In vertebrates, Notch signaling is also involved in both blood and vascular development. The role of Notch during AGM morphogenesis remains to be investigated (Mandal, 2004).

Cardioblasts and lymph-gland cells can arise from the division of a single cardiogenic mesodermal cell, which should be called a hemangioblast. A previous study induced clones in the cardiogenic mesoderm but used only Tin as a marker. This study also yielded mixed two-cell clones comprising a cardioblast and a nonlabeled cell, which, in light of the current findings, must be interpreted as a lymph-gland cell. Hemangioblasts have been proposed in vertebrates, although the definitive experiment in which a precursor is marked and its lineage is tracked has not been done. Blast colony-forming cells that give rise to both lineages in vitro and common markers that belong to both cell types in vivo have been identified, but direct evidence for the existence of a common precursor has not yet been found. This study, using genetic analysis of two-cell clones, establishes the existence of such a population in Drosophila. On the basis of these results, and given the conservation of the signaling and transcriptional components described here, the prediction is that many cells of the AGM in vertebrates may give rise to only blood or only vascular cells, but a number of intermixed hemangioblasts may give rise to mixed lineages. Future genetic screens aimed at finding components in early lymph-gland development will probably identify additional pathways and strategies important for vertebrate hematopoiesis (Mandal, 2004).

Defective decapentaplegic signaling results in heart overgrowth and reduced cardiac output in Drosophila

During germ-band extension, Dpp signals from the dorsal ectoderm to maintain Tinman (Tin) expression in the underlying mesoderm. This signal specifies the cardiac field, and homologous genes (BMP2/4 and Nkx2.5) perform this function in mammals. A second Dpp signal from the dorsal ectoderm restricts the number of pericardial cells expressing the transcription factor Zfh1. Via Zfh1, the second Dpp signal restricts the number of Odd-skipped-expressing and the number of Tin-expressing pericardial cells. Dpp also represses Tin expression independently of Zfh1, implicating a feed-forward mechanism in the regulation of Tin pericardial cell number. In the adjacent dorsal muscles, Dpp has the opposite effect. Dpp maintains Krüppel and Even-skipped expression required for muscle development. The data show that Dpp refines the cardiac field by limiting the number of pericardial cells. This maintains the boundary between pericardial and dorsal muscle cells and defines the size of the heart. In the absence of the second Dpp signal, pericardial cells overgrow and this significantly reduces larval cardiac output. This study suggests the existence of a second round of BMP signaling in mammalian heart development and that perhaps defects in this signal play a role in congenital heart defects (Johnson, 2007).

A previous study suggested that a second round of Dpp dorsal ectoderm-to-mesoderm signaling, stimulated by enhancers located in the dpp disk region, initiates during germ-band retraction (stage 12; Johnson, 2003). This is referred to as the second round of signaling because a distinct set of enhancers located in the dpp Haplo-insufficiency (Hin) region activates Dpp dorsal ectoderm-to-mesoderm signaling during germ-band extension (stage 8). Further, the data revealed that dpp dorsal ectoderm expression driven by the Hin region enhancers persists long after germ-band retraction. These studies showed that Hin-region-driven dpp expression is sufficient for Dpp ectodermal functions such as dorsal closure and dorsal branch migration (Johnson, 2007).

Given these data, it appears that the dppd6 inversion prevents the augmentation of dpp expression in the dorsal ectoderm during germ-band retraction that is normally provided by disk region enhancers. The presence of numerous mesodermal phenotypes in dppd6 mutants (Johnson, 2003) suggests that the augmentation of dpp expression is necessary to boost Dpp dorsal ectoderm signals so that they can reach the underlying mesoderm. Perhaps there are barriers of distance or extracellular matrix density between these germ layers that must be overcome (Johnson, 2007).

The data are wholly consistent with the hypothesis that the dppd6 inversion prevents the augmentation of dpp expression provided by disk region enhancers during germ-band retraction. The data further suggest that the augmentation of dpp expression is necessary to boost Dpp dorsal ectoderm signals such that they can reach the underlying mesoderm. Finally, this study shown that during germ-band retraction Dpp signals maintain the boundary between pericardial cells and dorsal muscle cells via two distinct mechanisms: the regulation of gene expression and the restriction of cell proliferation. To regulate gene expression, Dpp signals directly to pericardial cells and restricts Odd and Tin expression in a zfh1-dependent manner. Dpp also limits Tin expression, independently of zfh1, by repressing the expression of mid, a stimulator of proliferation (Johnson, 2007).

With respect to zfh1-dependent regulation, the data support the hypothesis that Dpp restricts Zfh1 expression to regulate the number of pericardial cells derived solely from symmetrically dividing lineages. Lineage analyses have identified both symmetric and asymmetric cell divisions of myogenic and pericardial precursor cells. Pericardial cells are derived from four separate lineages that arise from four distinct precursor cells. Asymmetric precursor cell divisions initiating between stages 8 and 10 give rise to the Odd-positive/Seven up (Svp)-positive pericardial cells and the Eve-positive/Tin-positive pericardial cells (EPCs). In contrast, symmetric division, initiating at the same stage, establishes the Odd-positive/Svp-negative pericardial cells (OPCs) and the Tin-positive/Eve-negative pericardial cells (TPCs). dpp mutations do not affect the number of EPCs or the number of Odd-positive/Svp-positive cells. However, embryos bearing dpp mutations show an increase in the number of OPCs and TPCs. Therefore, the ectopic pericardial cells seen in dpp mutants derive from symmetrically dividing lineages (Johnson, 2007).

Previous reports have shown that regulation of asymmetric cell division is a key mechanism in establishing boundaries among the various cell types in the dorsal mesoderm. For instance, in the absence of Numb, a Notch pathway antagonist, asymmetric progenitor cell division is abrogated and the number of Odd-positive/Svp-positive cells and EPCs increases at the expense of the Svp-expressing cardial cells and Eve-expressing dorsal muscle cells, respectively. This study extends these observations by showing that pericardial cell types derived from symmetrically dividing lineages are also under strict regulatory control (Johnson, 2007).

With respect to zfh1-dependent regulation of pericardial cell number, Dpp restricts cell proliferation and, in turn, Tin expression by limiting mid expression. In wild-type embryos, cell division in the dorsal mesoderm is largely complete by the early stages of germ-band retraction (stage 11), whereas in dppd6 embryos cell proliferation in the dorsal mesoderm continues through stage 13. Interestingly, the number of cells expressing Zfh1 increases from stage 12 to stage 13 in wild-type embryos in the absence of cell division, demonstrating that patterning events subsequent to cell division regulate cell fate choices in the dorsal mesoderm. This hypothesis is supported by the fact that tracing pericardial cell lineages requires inducing mitotic clones by stage 8. Therefore, the ectopically dividing mesoderm cells observed in dppd6 embryos are derived from cells with the potential to become Tin-expressing cells (Johnson, 2007).

During stage 12, tin expression is reactivated in a subset of cardiac cells in a mid-dependent fashion, suggesting that tin expression in precursor cells alone is not sufficient for specifying the ultimate fate of their daughter cells. Moreover, misexpression of mid results in both ectopic cell division and expanded tin expression. Lineage studies support the necessity of reactivating Tin by showing that a single precursor cell gives rise to two Tin-positive/Eve-negative pericardial cells and two siblings that do not express Tin. Thus tin is not reactivated in all subpopulations of pericardial cells. The data suggest that, during stage 12, Dpp prevents tin reactivation in cells occupying lateral regions of the dorsal mesoderm by limiting mid expression (Johnson, 2007).

Development of the dorsal musculature initiates when founder cells are specified in the mesoderm. These founder cells then fuse with neighboring cells to form syncitial myofibers. In the absence of Dpp, the pericardial cell domain expands into the dorsal muscle domain and reduces expression from the dorsal muscle genes Kr and Eve. Since the separation between pericardial and dorsal muscle cells is lost in dpp mutant embryos, it is concluded that Dpp maintains the pericardial-dorsal muscle cell boundary after it is established. Moreover, reducing pericardial cell number increases Kr expression after germ-band retraction, suggesting that cross-repressive interactions between pericardial and dorsal muscle cells contribute to patterning of the dorsal mesoderm. The presence of ectopic pericardial cells in the dorsal mesoderm reduces the number of myofibers comprising the dorsal muscles even though the dorsal muscle founder cells are, for the most part, correctly specified. pMad does not accumulate in Kr-expressing founder cells yet Kr expression is significantly reduced in dpp mutant embryos. Therefore, changes in Kr and Eve expression observed in embryos with altered dpp or zfh1 activity reflect alterations in the number of myoblast fusion events in the dorsal mesoderm (Johnson, 2007).

These data extend a previous study showing that misexpressing Zfh1 reduces dMef2 expression in somatic muscles. This study demonstrates that misexpression of Zfh1 induces ectopic pericardial cells and that the presence of pericardial cells in the dorsal muscle domain reduces myoblast fusion. Therefore, reduced dMef2 expression in embryos misexpressing Zfh1 is likely the result of reduced myoblast fusion and not of direct repression of dMef2 expression by Zfh1. Further, analysis of lmd mutants that have reduced numbers of myoblasts revealed that they also contain an excessive number of pericardial cells. Together, these results suggest that maintaining the pericardial-dorsal muscle cell boundary requires Dpp-mediated cross-repressive interactions between these cell types. Thus, in the absence of Dpp, the transformation of dorsal muscle cells into pericardial cells reduces the number of myoblasts available for fusion (Johnson, 2007).

Experiments in the larvae of Drosophila and other insects suggested that pericardial cells act as nephrocytes that filter the hemolymph. These studies also showed that pericardial cells secrete proteins into the hemolymph, suggesting that one pericardial cell function may be to provide short- or long-range signals. Consistent with this, reducing pericardial cell number reduces heart rate and increases the cardiac failure rate, suggesting that pericardial cells influence the development of cardiac cells (Johnson, 2007).

This study shows that pericardial cell hyperplasia reduces the luminal distance of the heart during systole as well as diastole, resulting in an overall decrease in average pulse distance of each contraction. However, pericardial overgrowth does not alter heart rate, indicating that cardiac cells develop appropriately in the presence of ectopic pericardial cells. Luminal measurements suggest a role for pericardial cells in the mechanics of heart function. One hypothesis for this is based on the fact that pericardial cell hyperplasia results in excess levels of extracellular matrix protein Pericardin (Prc) in the extracellular matrix (ECM) surrounding the heart. Prc is a collagen IV-like ECM protein secreted at high levels from pericardial cells. In dpp mutants, excess Prc is seen predominantly in the posterior of the heart where the pulse-distance reduction was observed. It is proposed that Prc secreted by pericardial cells limits the width of the dorsal vessel at diastole and thus modulates the pulse distance of each heart contraction. Pericardial cell overgrowth would increase Prc deposition, thereby reducing the size of the diastolic heart and the pulse distance. Consistent with this hypothesis, excessive expression of ECM proteins, including collagen IV, was correlated with heart failure in patients presenting with end-stage cardiomyopathy (Johnson, 2007).

It is well documented that many of the early events driving Drosophila embryonic heart development have been conserved in vertebrates. The data provide the first basis upon which to determine if Dpp regulation of Zfh1 or Tin late in heart development is also conserved (Johnson, 2007).

Two orthologs of zfh1, Sip1 and Kheper, have been identified in vertebrates. Zebrafish embryos injected with the Dpp homolog BMP4 show reduced Kheper expression while Xenopus embryos injected with the BMP antagonist Chordin display elevated Sip1 expression. These results suggest the possibility that Dpp repression of zfh1 expression may be conserved in vertebrates. In addition, mammalian Sip1 plays an essential role in heart development. In mice, Sip1 is expressed in neural crest cells (NCCs), paraxial mesoderm, and neuroectoderm. The subset of NCCs that express Sip1 give rise to the septum and large arteries of the heart. Sip1 knockout mice fail to form these NCCs and these mice die midway through gestation with numerous heart defects. Mice lacking the BMP receptors BMPRIA or ALK2 specifically in NCCs also display numerous cardiac phenotypes. In conditional knockout of ALK2 in NCCs, abnormalities are seen in the heart's outflow tract, and conditional knockout of BMPRIA in NCCs results in heart failure and early embryonic lethality similar to Sip1 knockout mice. Thus BMP signals are required for proper specification of NCCs, and loss of BMP signaling in NCCs phenocopies Sip1 knockout mice to an extent. It is tempting to speculate that, as in Drosophila, BMP signals regulate the Zfh1 ortholog Sip1 to correctly specify NCCs and, in turn, to properly pattern the mammalian heart (Johnson, 2007).

With regard to the conservation of late-stage Dpp regulation of Tin, a recent article describing a study of mice with a conditional knockout of Nkx2.5 where expression is missing only during late stages of heart development (post E14.5) is highly relevant. Utilizing rescue of Nkx2.5 mutant embryos with BMP-signaling-pathway components, the study identified a direct connection among BMP4 signaling, Nkx2.5 activity, and heart cell proliferation. Since Nkx2.5 is the Tin homolog, BMP4 is the Dpp homolog, and the mutant phenotype (heart cell hyperplasia) is the same in both species, this suggests that this aspect of Dpp signaling is conserved in mammals. Together with this study, these results suggest that defects in late-stage BMP signaling may play a role in congenital heart defects (Johnson, 2007).

Tip cells act as dynamic cellular anchors in the morphogenesis of looped renal tubules in Drosophil

Tissue morphogenesis involves both the sculpting of tissue shape and the positioning of tissues relative to one another in the body. Using the renal tubules of Drosophila, this study shows that a specific distal tubule cell regulates both tissue architecture and position in the body cavity. Focusing on the anterior tubules, it was demonstrated that tip cells make transient contacts with alary muscles at abdominal segment boundaries, moving progressively forward as convergent extension movements lengthen the tubule (see Tip-Cell-Dependent Anchorage of Anterior Tubules to Alary Muscles). Tip cell anchorage antagonizes forward-directed, TGF-beta-guided tubule elongation, thereby ensuring the looped morphology characteristic of renal tubules from worms to humans. Distinctive tip cell exploratory behavior, adhesion, and basement membrane clearing underlie target recognition and dynamic interactions. Defects in these features obliterate tip cell anchorage, producing misshapen and misplaced tubules with impaired physiological function (Weavers, 2013; see Graphical Abstract).

As the embryonic renal tubules assume their mature shape they interact with other tissues, responding to Dpp guidance cues as they take up their characteristic positions in the body cavity (Bunt, 2010). This study shows that, in addition, a single cell at the distal end of each renal tubule makes specific transitory, and finally long-term, contacts with target tissues. These cells express a distinctive pattern of genes and show characteristic exploratory activity, which is crucial for the stereotypical looped shape and position of the tubules in the body cavity. In turn, these features have profound consequences for the efficacy of fluid homeostasis in the whole animal (Weavers, 2013).

It is suggested that the elongation and forward extension of the tubules result from the combined effects of cell rearrangements that lengthen the tubule and the response of kink region cells to regional Dpp guidance cues. The evidence indicates that the tip cells act as anchors, through their interactions with alary muscles, so that tubules are tethered at both ends, the proximal end being attached through ureters to the hindgut. These attachments perform two functions: they stabilize the looped architecture, maintaining the kink close to the tubule midpoint, and they limit forward and ventral movement to ensure the stereotypical tubule arrangement in the body cavity (Weavers, 2013).

If tip cell contact with the alary muscles is lost, the kink 'unravels,' shifting distalward, and the tubule as a whole extends too far into the anterior, with the distal region lying more ventrally close to the Dpp-expressing gastric caeca. Confirming the existence of a forward tractive force responsible for tip cell detachment are the distortion of transient alary muscle targets before the tip cell detaches and the characteristic 'recoil' seen when the tip cell is ablated. Evidence that this results from the response to guidance cues is the failure of tip cells to detach from their first alary muscle contact (A5/A6) in the absence of the midgut Dpp guidance cue and the more anterior location of the kink region (close to the gastric caeca) in tubules where the tip cell stalk is greatly extended, for example when the activity of RhoA is repressed. The critical nature of the balance between these forward and restraining influences is also revealed when adhesion between the tip cells and alary muscles is increased by manipulating tip cell number or adhesive strength. In each case, the tip cells remain attached to alary muscles posterior to their normal final contacts, and this results in more posterior positioning of the whole tubule. Together, these results strongly suggest that tip cells detach because the forward movement of tubules overcomes the adhesive strength of their early transient contacts (Weavers, 2013).

The final tip cell/alary muscle target is highly reproducible, suggesting recognition through segmental identity, the A3/A4 target being the first encountered by the tip cell that expresses Ubx. However, altering Ubx expression in alary muscles has no effect on the final tip cell contact. Instead, it appears that tip cells adhere to each alary muscle they contact, and the final target depends on the balance between forward tubule movement and the strength of tip cell/target adhesion. Consistent with this view, when all the normal muscle targets are ablated, tip cells can make stable contacts with the A2/A3 alary muscle (Weavers, 2013).

From the time that they are specified, tip cells show distinctive patterns of gene expression, morphology, and behavior critical to their ability to make alary muscle contacts; they form dynamic filopodia, which explore the alary muscle surface, remain denuded of the BM that envelops the rest of the tubule, and express cell-adhesion proteins, including Neuromusculin (Nrm) and integrins. Protrusive activity depends on the rapid turnover of actin, mediated by regulators such as Rac GTPase and the actin-capping protein Enabled, which are active in tip cells (Weavers, 2013).

BM deposition basally around the tip cells severely inhibits protrusive behavior, and tip cells therefore employ multiple mechanisms to ensure that they remain denuded. These include the absence of expression of factors that promote BM deposition and stabilization (hemocyte attractants, BM components, or receptors), the removal by transcytosis of any BM that is deposited, and expression of MMP1, which is able to cleave matrix components. MMP1 is expressed late during tubule elongation and the protein is localized apically in tip cells, suggesting that its function might be to degrade transcytosed BM proteins (Weavers, 2013).

Protrusive exploratory behavior results in adhesive contacts made possible through tip cell expression of Nrm and integrins. Nrm is a homophilic cell-adhesion molecule of the Ig-domain superfamily. The binding partner for tip cell Nrm is unclear, as alary muscles do not express it. However, driving nrm expression in alary muscles induces strong adhesion, resulting in tip cells remaining bound to their first target in A5/A6. It is possible that Nrm in tip cells normally makes heterophilic associations with Ig domain-containing proteins such as Dumbfounded (Kirre), which is expressed in alary muscles and is sufficient, when overexpressed, to induce more posterior target adhesio (Weavers, 2013).

Tip cells express integrins, and complexes accumulate as each target contact is made, but initially they do not lead to long-term adhesion. It is suggested that the strength of adhesion increases with successive contacts, either through increased expression of integrins and their associated factors or by regulated adhesive complex turnover, as shown in other tissues. Once the final tip cell contact is made, BM accumulates around the tip cellalary muscle surface, increasing the concentration of integrin ligands at the junction. The accompanying decline in the protrusive activity of the tip cell could also result from integrin-mediated adhesion, which is known to reduce levels of the actin-capping protein Enabled. This sequence of events parallels the mechanism by which elongated myotubes and tendon cells establish their myotendinous junctions (Weavers, 2013).

Once the anterior tubule tip cells make their final alary muscle contact, they remain attached throughout development into adult life. Such interaction of excretory tubule tips with muscles is a common feature of renal systems in insects, either with alary muscles or with fine striated muscles that spiral along the tubule. Muscle contacts increase tubule movement, maximizing the effectiveness of excretion, by increasing hemolymph sampling and enhancing tubule flow. Similar contacts are found outside the arthropods; the flame cells that cap planarian protonephridial tubes develop prominent filopodia and interact with nearby muscle fibers, providing anchorage, thought to be important during branching morphogenesis in this system (Weavers, 2013).

Tip cells or groups of cells at the distal tips of outgrowing epithelial tubes act as organizers in tubular systems, from the migrating Dictyostelium slug to the branching epithelial scaffolds of human organs. As in fly renal tubules, these distinctive cells regulate cell division and guided tubule extension, and in mammalian systems they control branching morphogenesis (Weavers, 2013).

However, in distinct contrast to the role of tip cells in the morphogenesis of these systems, the tip cells of the anterior renal tubules play no role in leading outgrowth. Instead, they act to counteract outgrowth, and importantly this leads to the development of a looped tubular structure both by tethering the distal tips of tubules close to their proximal junction with the ureter and by maintaining the tightness of the tubule kink region. Looped tubular structures are relatively uncommon; a tubule tree as in the lung, pancreas, or liver or an anastomosing network as in the vascular system is more frequently seen. However, a striking example of looped tubules is found in the mammalian kidney, where the distal and proximal convoluted tubules together with the loop of Henle connect the tubule tip (at the glomerulus) to the collecting duct (close to the site of urine outflow). Looping of both the nephron and its vascular supply creates a countercurrent system that maximizes the efficiency of ion and fluid homeostasis. Such exchange systems also occur in insects with specialized diets or those living in dry conditions. Countercurrent exchange has not been demonstrated in Drosophila melanogaster tubules, where it is more likely that the looped tubule structure is important for effective hemolymph sampling (Weavers, 2013).

In the development of the mammalian nephron, as in fly renal tubules, both the site of connection to the ureter and the tubule tip, the renal corpuscle, are established early in organ development so that tubule extension, by both cell proliferation and rearrangements, occurs between these fixed points. It will be interesting to discover whether similar tissue interactions stabilize the position of the developing glomerulus, and so play a prominent role in maintaining the looped structure as kidney tubules extend, resulting in the final intricate and regular array of nephrons apparent in the mature mammalian kidney (Weavers, 2013).

Dally proteoglycan mediates the autonomous and nonautonomous effects on tissue growth caused by activation of the PI3K and TOR pathways

How cells acquiring mutations in tumor suppressor genes outcompete neighboring wild-type cells is poorly understood. The PTEN and TOR pathways are frequently activated in human cancer, and this activation is often causative of tumorigenesis. This study used the Gal4-UAS system in Drosophila imaginal primordia, highly proliferative and growing tissues, to analyze the impact of restricted activation of these pathways on neighboring wild-type cell populations. Activation of these pathways leads to an autonomous induction of tissue overgrowth and to a remarkable nonautonomous reduction in growth and proliferation rates of adjacent cell populations. This nonautonomous response occurs independently of where these pathways are activated, is functional all throughout development, takes place across compartments, and is distinct from cell competition. The observed autonomous and nonautonomous effects on tissue growth rely on the up-regulation of the proteoglycan Dally, a major element involved in modulating the spreading, stability, and activity of the growth promoting Decapentaplegic Dpp signaling molecule. The findings indicate that a reduction in the amount of available growth factors contributes to the outcompetition of wild-type cells by overgrowing cell populations. During normal development, the PI3K/PTEN and TSC/TOR pathways play a major role in sensing nutrient availability and modulating the final size of any developing organ. This study presents evidence that Dally also contributes to integrating nutrient sensing and organ scaling, the fitting of pattern to size (Ferreira, 2015).

Evidence is presented that targeted deregulation of the PI3K/PTEN, TSC/TOR, or hippo/Yorkie pathways, known to promote tissue overgrowth by increasing the number and/or size of cells, induces a nonautonomous reduction in tissue size of adjacent cell populations. This nonautonomous effect is a consequence of a reduction in both cell size and proliferation rates (cell number), and it is not a consequence of programmed cell death or the withdrawal of nutrients from neighboring tissues, as reducing the levels of proapoptotic genes or subjecting larvae to different amino-acid diets does not have any impact on the size reduction of neighboring cell populations. The glypican Dally, which plays a major role in regulating the spread of Dpp in Drosophila tissues, is up-regulated upon deregulation of these tumor suppressor pathways, and the increase in Dally expression levels contributes to the autonomous effects on tissue size and to the nonautonomous reduction in cell number. Whereas the autonomous effects on tissue size caused by deregulation of these tumor suppressor pathways are most probably due, as least in part, to the capacity of Dally to facilitate Dpp spreading throughout the tissue, it is proposed that the nonautonomous effects on cell number are a consequence of withdrawal of Dpp from neighboring tissues. This proposal is based on a number of observations. First, the width of the Dpp activity gradient as well as the total amount of Dpp activity was reduced in adjacent cell populations upon targeted depletion of tumor suppressor pathways. Second, the nonautonomous effects on tissue size were fully rescued by Dally depletion, which has a rather specific role on the spread of Dpp when overexpressed. Third, the nonautonomous effects on tissue size, growth and proliferation rates, and/or Dpp availability and signaling can be phenocopied by overexpression of Dally or the Dpp receptor Tkv (Ferreira, 2015).

Different strengths of the autonomous and nonautonomous effects were observed upon deregulation of these tumor suppressor pathways or overexpression of Dally in either the A or P compartments. Despite the mild autonomous induction of tissue growth caused by the ci-gal4 driver in A cells, it caused a relatively strong nonautonomous reduction of the neighboring compartment. On the contrary, the en-gal4 driver caused a strong autonomous induction of tissue growth in P cells but a relatively weak nonautonomous reduction of the neighboring compartment. The differential autonomous response might simply reflect different strengths of these Gal4 drivers. By contrast, the strongest nonautonomous effects caused by the ci-gal4 driver (when compared to the en-gal4 driver) might be because Dpp expression is restricted to the A compartment and increased levels of Dally in Dpp expressing cells are more efficient at titrating out the levels of this growth factor from the neighboring compartment. It was noticed that the nonautonomous effects on cell size observed upon deregulation of the PI3K/PTEN, TSC/TOR, or hippo/Yorkie pathways are Dally independent, as overexpression of Dally did not cause a nonautonomous reduction in cell size. Moreover, depletion of Dally did not rescue the nonautonomous reduction in cell size caused by activation of these pathways. These results are consistent with the fact that changes in Dpp signaling do not cause any effect on cell size and indicate that Dally and Dpp are regulating cell number but not cell size. Somatic mutations in tumor suppressor genes such as PTEN or TSC are frequently accumulated in early events of tumor development, and these mutations are thought to contribute to the selection of tumorigenic cells. Competition for available growth factors, by modulating the levels of glypicans, such as Dally, might contribute to the outcompetition of wild-type cells and to the selection of malignant mutation-carrying cells in human cancer (Ferreira, 2015).

The PI3K/PTEN and TSC/TOR signaling pathways play a role not only in disease but also during normal development. These two pathways modulate the final size of the developing organism according to nutrient availability. The current results also identify, in this context, Dally as a molecular bridge between nutrient sensing and wing scaling in Drosophila. In a condition of high nutrient availability, which leads to the activation of the nutrient-sensing PI3K/PTEN and TSC/TOR pathways, increased levels of Dally facilitate the spread of Dpp throughout the growing tissue and contribute to the generation of larger but well-proportioned and scaled adult structures. Depletion of Dally expression levels rescues the tissue growth caused by high levels of nutrients or activation of the nutrient-sensing pathways and gives rise to smaller and, again, well-proportioned and scaled adult structures. Of remarkable interest is the capacity of Dally to induce tissue overgrowth when overexpressed or to mediate tissue growth upon deregulation of the PI3K/PTEN, TSC/TOR, or hippo/Yorkie pathways. Interestingly, deregulation of these pathways, and the resulting tissue overgrowth, leads to the expansion of the Dpp gradient without affecting the total levels of Dpp signaling (Ferreira, 2015).

These results imply that Dpp activity levels do not play an instructive role in promoting tissue growth but rather that it is the range of the Dpp gradient that regulates final tissue size. Consistent with this proposal, depletion of Dally levels in one compartment (which might lead to increased levels of available Dpp in the neighboring cell population) does not cause any visible nonautonomous effect in tissue size. These results are reminiscent of the capacity of Dpp to restrict its own spreading through the repression of Pentagone, a diffusible protein that interacts with Dally and contributes to the expansion of the Dpp gradient. The graded distribution of Dpp leads, via the interaction with its receptor complex, to the graded activation of Mad/Medea, which in turn represses the transcription of brinker (brk). This creates a gradient of Brk expression that is reciprocal to the Dpp gradient. Brk is a transcriptional repressor that acts negatively to establish, in a dose-dependent manner, the expression domain of Dpp target genes like spalt. Thus, Dpp regulates the expression of target genes by repressing brinker. Remarkably, the reduced size of the wing primordium observed in hypomorphic alleles of dpp is restored when combined with brk mutants. This experimental evidence indicates that Dpp controls wing growth entirely via repression of brk. The Dally-mediated increase in the width of the Dpp gradient observed upon deregulation of the PI3K/PTEN, TSC/TOR, or hippo/Yorkie pathways might contribute to restrict the expression domain of brk to the lateral sides of the wing primordium. Similarly, the nonautonomous decrease in the width of the Dpp gradient might cause an expansion of the brkdomain, which is known to repress growth. Interestingly, Dally-mediated spreading of other secreted growth factors might also contribute to the autonomous effects on tissue growth caused by deregulation of the PI3K/PTEN, TSC/TOR, or hippo/Yorkie pathways. This is revealed by the fact that Dally depletion rescues both the autonomous and the nonautonomous effects, whereas deregulation of these pathways are still able to induce some growth upon knocking down Dpp (Ferreira, 2015).

Compartments have been proposed to be units of growth control. In other words, the size of each compartment is controlled independently. The results on the lack of nonautonomous effects on tissue growth upon depletion of Dally or Sfl, the enzyme needed for the modification of HS chains within glypicans, indicate that this is the case. Targeted depletion of glypican expression or activity in the developing compartments gave rise to an autonomous reduction in tissue size without affecting the neighboring compartment. However, independent lines of evidence support the view that adjacent compartments buffer local variations in tissue growth caused by different means, including a nonautonomous reduction in tissue size upon depletion of the protein biosynthetic machinery or reduced epidermal growth factor receptor (EGFR) activity. The current results on the capacity of overgrowing compartments to withdraw Dpp from neighboring tissues upon targeted deregulation of the PI3K/PTEN, TSC/TOR, or hippo/Yorkie pathways and to cause a nonautonomous reduction in growth and proliferation rates reinforce the view that compartments are susceptible to modulate their growth rates upon different types of stress, including depletion of tumor suppressor genes. Interestingly, the halteres and wings of Drosophila are homologous thoracic appendages, and the activity of the Ultrabithorax (Ubx) Hox gene in the haltere discs contributes to defining its reduced size. Remarkably, it does so by reducing the expression levels of Dally, thus reinforcing the role of Dally in modulating tissue growth in epithelial organs (Ferreira, 2015).

Dpp and the Wing Disc

continued: decapentaplegic Developmental Biology part 2/3 | part 3/3

decapentaplegic: Biological Overview | Evolutionary Homologs | Transcriptional regulation | Targets of activity | Protein Interactions | Post-transcriptional Regulation | Effect of mutation | References

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.