Targets of Activity (part 2/2)

Hedgehog targets in the wing disc

The principal function of HH is to activate transcription of dpp at the boundary between anterior and posterior compartments of the developing wing, thereby establishing a source of DPP activity, the primary determinant of anterior-posterior patterning. In a remarkably similar fashion, the function and expression of the vertebrate sonic hedgehog (shh) gene is closely associated with the 'zone of polarizing activity' (ZPA) that controls anterior-posterior patterning of the vertebrate limb. Both of these functions suggest a role for Hedgehog family proteins as morphogens. Vertebrate shh can be inserted ectopically into Drosophila. The effects on Drosophila wing patterning have been explored by ectopically expressing varying levels of hh and shh, as well as of the HH target gene, dpp. Different levels of hh activity can induce graded changes in the patterning of the wing. Zebrafish shh acts in a similar though attenuated fashion. Varying levels of ectopic hh and shh activity can differentially activate transcription of the patched and dpp genes. Ectopic expression of dpp alone is sufficient to induce the pattern alterations caused by ectopic hh or shh activity (Ingham, 1995).

Genetic analysis has identified wg transcription as one of the targets of hh activity. It has been suggested that the spatial control of wg expression depends on the limited range of the HH signal and the differential competence of responding cells. Ubiquitous expression of hh causes the ectopic activation of wg in only a subset of the cells of each parasegment. Competence of cells to express wg is independent of their ability to receive the HH signal (Ingham, 1993).

Hedgehog is a secreted protein; it can cross embryo parasegment borders and the anterior-posterior compartment border of imaginal discs to neighboring cells that express neither engrailed nor hedgehog. In these cells, it is localized in discrete punctate structures sequestered within the polarized epithelium. Analysis of animals that have expressed hedgehog ectopically, or of a mutant that expresses hedgehog abnormally in the anterior compartment of the wing disc, indicates that Hedgehog is involved in regulating patched. In the embryo, Hedgehog regulation of patched apparently facilitates patched and wingless expression. In the discs, Hedgehog regulation of patched and other genes in the anterior compartment helps to establish the proximodistal axis. It is proposed that cell-cell communication mediated by Hedgehog links the special properties of compartment borders with specification of the proximodistal axis in imaginal development (Tabata, 1994).

Patched and Hedgehog have reciprocal effects on gene expression in the wing disc. High Patched levels, expressed in either normal or ectopic patterns, result in loss of wing vein patterning in both compartments, centering at the anterior/posterior border. In addition, patched inhibits the formation in the wing blade of the mechanosensory neurons, the campaniform sensilla. The patched wing vein phenotype is modulated by mutations in hedgehog and cubitus interruptus (ci). patched overexpression inhibits transcription of patched and decapentaplegic and post-transcriptionally decreases the amount of CI protein at the anterior/posterior boundary. In wing discs, which express ectopic hedgehog, CI levels are correspondingly elevated, suggesting that hedgehog relieves patched repression of CI accumulation. Protein kinase A also regulates CI; protein kinase A mutant clones in the anterior compartment have increased levels of CI protein. Thus patched influences wing disc patterning by decreasing CI protein levels and inactivating hedgehog target genes in the anterior compartment (Johnson, 1995).

Hedgehog, along with Engrailed, regulates invected. engrailed expression has been targeted to different regions of the wing disc. In the anterior compartment, ectopic en expression gives rise to the substitution of anterior structures for posterior ones, thus demonstrating Engrailed's role in specification of posterior patterns. The en-expressing cells in the anterior compartment induce high levels of HH and DPP, resulting in local duplications of anterior patterns. Additionally, HH is able to activate en and the engrailed-related gene invected in this compartment (Guillen, 1995).

The transcription factor encoded by spalt major gene, which is expressed in the wing disc in a broad wedge centered over the dpp stripe, is one target of Dpp signaling. The anterior edge of the salm expression domain abuts a narrow stripe of rhomboid (rho)-expressing cells corresponding to the L2 longitudinal vein primordium. rhomboid is transcribed in a pattern that matches the future site of vein formation. hedgehog mis-expression along the anterior wing margin induces a surrounding domain of salm expression, the anterior edge of which abuts a displaced rho L2 stripe. salm plays a key role in defining the position of the L2 vein, since loss of salm function in mosaic patches induces the formation of ectopic L2 branches, composed of salm- cells running along clone borders at positions where salm- cells confront salm+ cells (Sturtevant, 1997).

It is believed that wing veins L3 and L4 do not respond to DPP signaling but instead L3 is determined directly by a threshold response to Hedgehog secreted across the A-P compartment boundary. It has been observed that clones of mutant patched cells in the middle of the anterior compartment are surrounded by an ectopic L3 vein which comprises wild-type cells. Similarly, loss-of-function clones of Protein kinase A, which functions like Ptc to repress ptc expression, are encircled by ectopic veins consisting of wild-type cells. Thus, cells with low levels of ptc may induce adjacent ptc+ cells to assume L3 fates. Since secreted HH is thought to be responsible for inactivating PTC, the position of the L3 primordium might be determined by a threshold response to HH diffusing from the posterior compartment (Sturtevant, 1997 and references).

Hedgehog (Hh) and Decapentaplegic (Dpp) direct anteroposterior patterning in the developing Drosophila wing by functioning as short- and long-range morphogens, respectively. The activity of Dpp is graded and is directly regulated by a novel Hh-dependent mechanism. Dpp activity is monitored by visualizing the activated form of Mothers against dpp (Mad), a cytoplasmic transducer of Dpp signaling. Activated Mad levels are highest near the source of Dpp but are unexpectedly low in the cells that express dpp. Hh induces dpp in these cells; it also attenuates their response to Dpp by downregulating expression of the Dpp receptor thick veins (tkv). It has been suggested that regulation of tkv by Hh is a key part of the mechanism that controls the level and distribution of Dpp (Tanimoto, 2000).

To determine whether the low levels of phosphorylated Mad (p-MAD) in dpp-expressing cells reflect an autoregulatory influence on Dpp signaling that functions only in dpp-expressing cells, or whether p-MAD levels are controlled by another signal such as Hh, p-MAD was monitored in wing discs carrying clones that ectopically express hh or dpp. If Dpp has the potential to attenuate its own signal transduction, the level of p-MAD would be expected to decrease cell autonomously in cells that express dpp. However, the level of p-MAD is elevated around the clones with no obvious reduction within the clones. This observation is not consistent with an autoregulatory mechanism. In contrast, clones expressing hh in the A compartment cause both a nonautonomous reduction of p-MAD level as well as ectopic elevation of p-MAD around the clones, which is due to ectopic induction of dpp by Hh. No changes in p-MAD levels are observed when Hh is expressed in the P compartment. These observations suggest that Hh attenuates Dpp signal transduction by regulating the type I receptor-dependent phosphorylation of MAD. Thus, the activity of Dpp depends on both the concentration of Dpp and Hh (Tanimoto, 2000).

It was next asked whether Hh directly controls p-MAD levels. First, clones that ectopically express HhCD2, a membrane-tethered form of Hh, were compared to clones that express a diffusible form of Hh. P-MAD levels are reduced both in cells that expressed HhCD2 and in cells that are immediately adjacent. In contrast, the levels of p-MAD are reduced more broadly when cells express diffusible Hh. The distribution of p-MAD was examined when clones of cells mutant for either Protein kinase A (Pka) alone or both Pka and dpp were induced. Pka functions to antagonize Hh signaling: Pka- clones in the A compartment fail to repress Hh signaling in a cell-autonomous manner. Pka- clones in the A compartment have reduced levels of p-MAD. Moreover, p-MAD accumulates to high levels in cells surrounding Pka- clones, due, presumably, to the ectopic induction of Dpp by constitutive Hh signaling in the mutant cells. In clones mutant for both Pka and dpp, the reduction was evident, but it was not accompanied by elevated levels in the surrounding cells. Outside the double-mutant clones, p-MAD appears to respect the endogenous gradient of Dpp. Autonomous reduction in the level of Spalt is also seen in both types of mutant clones, confirming both that p-MAD accurately manifests the active state of Dpp signaling and that the reduction of p-MAD leads to the decrease of target gene expression. It is concluded that Hh signaling directly attenuates Dpp signal transduction regardless of the level of endogenous Dpp (Tanimoto, 2000).

The contribution of Cubitus interruptus (Ci) to Hh-dependent attenuation of Dpp signaling was examined. ci is expressed in the A compartment only. In clones that ectopically express Ci in the P compartment where Hh is abundant, the levels of p-MAD were reduced in a cell-autonomous manner. Levels of Sal are similarly reduced in these cells. Clones that ectopically express ci in A compartment regions that receive little or no Hh have no effect on p-MAD levels. It is concluded that Hh both induces dpp expression and downregulates Dpp signal transduction in the same cells (Tanimoto, 2000).

Because Ci is involved in regulating p-MAD levels, it is possible that Hh controls the expression of a component that functions upstream of Mad phosphorylation. The possibiliy that Hh regulates the transcription of the Dpp receptors was therefore investigated. Dpp preferentially signals through the Tkv receptor and also negatively regulates tkv expression. Therefore, the correlation between tkv expression and the activity of Dpp was examined. The level of tkv expression is higher in cells located peripherally and is lower in the central region. In addition, a sharp reduction in expression at the A/P border was observed, a pattern very similar to that of p-MAD. The level of tkv expression in the area between the peripheral region and the A/P border is referred to as 'basal' and the reduced level at the A/P border as 'hyperrepression'. Interestingly, the basal level in the P compartment is higher than it is in the A compartment. This may account for the steeper gradient of p-MAD in the P compartment, since high levels of Tkv limit the movement of Dpp; since the spread of Dpp would be less in the P compartment, its gradient of activity would be expected to be steeper (Tanimoto, 2000).

In order to investigate the role of Dpp in generating the complex pattern of tkv expression, tkv and p-MAD were monitored in discs that ubiquitously express dpp. These discs exhibit significant overgrowth. tkv-lacZ at the A/P border remains hyperrepressed, but the level of peripheral tkv is significantly reduced, apparently to the basal level. This suggests that the level of tkv in the peripheral regions is directly regulated by Dpp but that the level at the A/P compartment border is not. The basal level in the P compartment remains higher than in the A compartment. Distribution of p-MAD is consistent with the expression pattern of tkv: overexpression of dpp caused the ubiquitous elevation of p-MAD, except at the A/P border. In contrast, phosphorylation of Mad at the A/P border is unchanged in the presence of excess Dpp. This suggests that the level of tkv along the A/P border is limiting. When constitutively active Tkv (tkv*) is ubiquitously expressed, peripheral tkv expression is reduced to the basal level, comparable to Dpp overexpression. Unlike the discs in which Dpp is overexpressed, no reduction in p-MAD is observed at the A/P border in these discs. These results suggest that in middle regions of the A and P compartments in normal discs, Dpp signaling reduces tkv expression from the levels at the periphery to the basal level. However, Dpp signaling is not responsible for the hyperrepression at the A/P border or for the difference between the basal levels in the A and P compartments. Given that hyperrepression of p-MAD along the A/P border is not observed in the presence of TKV*, it is more likely that in this region, Hh modulates p-MAD levels by regulating tkv (Tanimoto, 2000).

To ask whether Hh controls tkv expression directly at the A/P border, clones of cells mutant for patched (ptc), expressed in all A cells, were examined. The Hh signal transduction is constitutively active in the absence of the ptc activity. p-MAD and tkv-lacZ were monitored in ptc- clones. The clones in the A compartment ectopically activate Hh signaling in a cell-autonomous manner; they also caused cell-autonomous repression of tkv. Outside of the clone, the level of tkv is reduced to the basal level, but this level is higher than within the clone. This behavior is likely due to ectopic expression of dpp. P-MAD levels are also reduced in the clone and are higher around the clone. This establishes that Hh signaling directly represses tkv and, as a consequence, represses phosphorylation of MAD. This conclusion is confirmed by examining discs carrying clones of cells in the P compartment that ectopically express ci. tkv is autonomously repressed in these clones, indicating that Ci-mediated Hh signaling directly represses tkv (Tanimoto, 2000).

In order to understand the significance of Hh-dependent repression of tkv, the level of tkv expression was changed. In the wing disc, when overexpressing tkv in the dorsal compartment under the control of apterous (ap) enhancer, the levels of the dorsal p-MAD and Spalt expression at the A/P border are elevated, and their distribution is significantly narrower than in the ventral compartment. This is probably due to the sequestration of Dpp protein by the elevated Tkv. In addition, the adult wing misexpressing tkv was examined in dpp-expressing cells where tkv is hyperrepressed. The overexpression caused small wings with abnormal patterning in the central region of the wing. These results are consistent with the proposal that Hh, but not Dpp, patterns the central wing region. It should be noted, however, that the level of tkv expression in the experimental clones is probably at least several fold higher than normal, since these experiments utilized the Gal4/UAS system (Tanimoto, 2000).

The activities of Dpp were examined when tkv expression differed from wild type by only 2-fold. Clones mutant for tkv were generated in a heterozygous background such that both mutant (0 copies of the wild-type tkv) and fully wild-type sister clones (2 copies of wild-type tkv) were produced. However, since clones of cells with no Tkv activity do not survive in the wing pouch because they need Tkv activity to grow, only their sister clones survive. Homozygous (+/+) clones at the A/P border increase p-MAD and Spalt levels autonomously. The same analysis was performed using a null sax allele: differences in the levels of p-MAD and Spalt between clones carrying two wild-type sax and cells lacking one copy of sax were negligible. Taken together, it is proposed that the precise regulation of the Tkv receptor level by Hh signal is necessary for Dpp morphogen to shape the correct activity gradient (Tanimoto, 2000).

The nuclear proteins Spalt and Spalt-related belong to a conserved family of transcriptional regulators characterized by the presence of double zinc-finger domains. In the wing, they are regulated by the secreted protein Decapentaplegic and participate in the positioning of the wing veins. Regulatory regions in the spalt/spalt-related gene complex have been identifed that direct expression in the wing disc. The regulatory sequences are organized in independent modules, each of them responsible for expression in particular domains of the wing imaginal disc. In the thorax, spalt and spalt-related are expressed in a restricted domain that includes most proneural clusters of the developing sensory organs in the notum, and are regulated by the signaling molecules Wingless, Decapentaplegic and Hedgehog. spalt/spalt-related are found to participate in the development of sensory organs in the thorax, mainly in the positioning of specific proneural clusters. Later, the expression of at least spalt is eliminated from the sensory organ precursor cells and this is a requisite for the differentiation of these cells. It is postulated that spalt and spalt-related belong to a category of transcriptional regulators that subdivide the thorax into expression domains (prepattern) required for the localized activation of proneural genes (de Celis, 1999).

The sal and salr genes are expressed in only part of the thorax in three domains that have been defined with reference to en, wg and ci: the thoracic posterior compartment marked by En, an adjacent stripe anterior to the anteroposterior compartment boundary corresponding to the stripe of maximal accumulation of Ci, and a zone between the stripe of wg expression and the hinge. A fourth domain in the central thorax, from where only microchaetae develop, does not express sal/salr. To explore the regulatory mechanisms that localize sal/salr expression with respect to the anteroposterior compartment boundary and wg, experiments were performed in which genes that function in developmental signaling were expressed ectopically using the Gal4 system. A series of experiments led to the conclusion that, in the thorax, dual hh signaling is required to induce sal/salr expression: signaling through dpp and signaling that is dpp independent. Thus, expression of hh in clones within the central thorax (presumably accompanied by induction of dpp) leads to ectopic expression of sal/salr; interestingly, this ectopic expression is observed both in hh-expressing cells and in adjacent cells. In contrast, ectopic expression of dpp does not result in activation of sal transcription in the thorax or in the hinge, but it does so in the wing blade. The transcription factor Cubitus interruptus (Ci), a key mediator of Hedgehog signaling, was also experimentally mis-expressed in clones of cells. Ci is only able to activate sal ectopically in the wing blade, a place where ectopic expression of Ci results in novel expression of dpp, but not in the central thorax or wing hinge. In any other tissue studied to date, Hh signaling depends on Ci; since hh positively regulates sal in the thorax, the failure of ectopic Ci to activate sal expression there may be ascribed to the presence of countermanding repressors. However, even though dpp is not sufficient to induce sal/salr in the thorax, it is required. Thus, mitotic clones of Pka (corresponding to constitutive activation of hh signaling show cell autonomous expression of sal; in contrast, Pka;dpp double mutant clones do not express sal, indicating that, close to the Dpp source, hh and dpp signaling must cooperate to activate sal expression in the thorax. In agreement with this, the expression of sal can be reduced in tkv mutant cells, which have reduced levels of a Dpp receptor. The requirement of dpp function for induction of sal differs in different parts of the thorax. In the central thorax, where sal is normally not expressed, tkv clones have no effect. In region 2, where dpp is normally expressed, tkv clones result in reduced expression of sal. In other regions of the thorax, expression of sal is unaffected by the reduction of tkv (de Celis, 1999).

The secreted Hedgehog (Hh) proteins have been implicated as mediators of positional information in vertebrates and invertebrates. A gradient of Hh activity contributes to antero-posterior (A/P) patterning of the fly wing. In addition to inducing localised expression of Decapentaplegic (Dpp), which in turn relays patterning cues at long range, Hh directly patterns the central region of the wing. Short-range, dose-dependent Hh activity is mediated by activation of the transcription factor Collier (Col). In the absence of col activity, longitudinal veins 3 and 4 (L3 and L4) are apposed and the central intervein is missing. Hh expression induces col expression in a narrow stripe of cells along the A/P boundary through a dual-input mechanism: inhibition of proteolysis of Cubitus-interruptus (Ci) and activation of the Fused (Fu) kinase. Col, in cooperation with Ci, controls the formation of the central intervein by activating the expression of blistered (bs), which encodes the Drosophila serum response factor (D-SRF). D-SRF activity is required for the adoption and maintenance of the intervein cell fate. Furthermore, col is allelic to knot, a gene involved in the formation of the central part of the wing. This finding completes an understanding of the sectorial organization of the Drosophila wing. It is concluded that Col, the Drosophila member of the COE family (Col/Olf-1/EBF) of non-basic, helix-loop-helix (HLH)-containing transcription factors, is a mediator of the short-range organizing activity of Hh in the Drosophila wing. These results support the idea that Hh controls target gene expression in a concentration-dependent manner and highlight the importance of the Fu kinase in this differential regulation. The high degree of evolutionary conservation of the COE proteins and the diversity of developmental processes controlled by Hh signaling raises the possibility that the specific genetic interactions depicted here may also operate in vertebrates (Vervoort, 1999).

The Hedgehog (Hh) and Epidermal growth factor receptor (Egfr) signaling pathways play critical roles in pattern formation and cell proliferation in invertebrates and vertebrates. In this study, a direct link between these two pathways is demonstrated in Drosophila. Hh and Egfr signaling are each required for the formation of a specific region of the head of the adult fruitfly. hh and vein (vn), which encodes a ligand of the Drosophila Egfr, are expressed in adjacent domains within the imaginal primordium of this region. Using loss- and gain-of-function approaches, it has been demonstrated that Hh activates vn expression. Hh activation of vn is mediated through the gene cubitus interruptus (ci) and this activation requires the C-terminal region of the Ci protein. wingless (wg) represses vn expression, thereby limiting the domain of EGFR signaling (Amin, 1999).

The Drosophila eye disc is a sac of single layer epithelium with two opposing sides: the peripodial membrane (PM) and the disc proper (DP). Retinal morphogenesis is organized by Notch signaling at the dorsoventral (DV) boundary in the DP. Functions of the PM in coordinating growth and patterning of the DP are unknown. The secreted proteins Hedgehog, Wingless, and Decapentaplegic are expressed in the PM. From there they control DP expression of the Notch ligands Delta and Serrate. Peripodial clones expressing Hedgehog induce Serrate in the DP while loss of peripodial Hedgehog disrupts disc growth. Furthermore, PM cells extend cellular processes to the DP. Therefore, peripodial signaling is critical for eye pattern formation and may be mediated by peripodial processes (Cho, 2000).

Restricted localization of Hh-, Wg-, and Dpp-LacZ+ expressing cells along the DV axis in L1 discs suggests that these signals might act upstream of N. To test this idea, Hh activity was removed using a temperature-sensitive allele; Wg was ectopically expressed using hs-wg, or Dpp activity was removed by using a heteroallelic combination of the two dpp alleles, and then the expression patterns of the N ligands Dl and Ser were visualized in the eye discs. In L2 wild-type discs, Dl is preferentially expressed in the dorsal domain, while Ser is enriched along the DV midline of the DP. Both Dl and Ser are also present in the PM at a low level. In hhts2 mutants shifted to restrictive temperature during the early L1 stage, both Dl and Ser are uniformly expressed in dorsal and ventral domains. Ubiquitous Wg overexpression causes variable defects in Dl pattern such as significant reduction in the dorsal domain except near the margin or mislocalization to the ventral domain. Wg overexpression also causes mislocalization of Ser to the dorsal DP. dppe12/dppd14 mutant discs showed similar disruption of the DV-specific Dl and Ser pattern, indicating the necessity of Hh, Wg, and Dpp in DV patterning (Cho, 2000).

The complex interplay of Hh, Wg, and Dpp signaling has been studied for initiation and progression of the morphogenetic furrow. This study has examined much earlier stages of eye development to determine whether these same molecules organize DV patterning prior to retinal differentiation; it has been demonstrated that: (1) Hh, Wg, and Dpp display distinct DV expression patterns in the PM in early discs; (2) their signals are essential for domain-specific expression of Dl and Ser in the DP, and (3) signaling from the PM to DP is important for patterning in the DP. These findings provide a novel view of how eye discs are patterned, a model suggesting Hh, Wg, and Dpp signal from the PM to the DP by means of cellular processes (Cho, 2000).

Soon after the embryo hatches, wg- and dpp-LacZ+ cells appear in the dorsal and ventral domains of the disc, respectively. This suggests that the eye disc is already subdivided into dorsal and ventral fates. Consistent with this data, analyses of genetic mosaics have indicated that the eye disc consists of dorsal and ventral compartments of different cell lineages and of different cell affinities. Subsequent to the initial appearance of wg- and dpp-LacZ+ cells, these two types of cells are juxtaposed in the DV midline of PM and seem to be mutually exclusive in later stages. Such an antagonistic interaction between Wg and Dpp is a common theme that has emerged from studies in limb disc patterning and may play a crucial role in defining domains in the PM (Cho, 2000).

In addition to DV subdivisions, the dorsal domain of L1 discs appears to be further divided into anterior-posterior subdomains. This is based on the expression of Wg in the anterior but not in the posterior dorsal domain, while Dpp is expressed in the opposite pattern. The anterior and posterior subdomains may correspond to the anlage for the head and the dorsal eye, respectively. It has been shown that Wg expressed in the vertex and gena primordia is important for head capsule formation, while Dpp is antagonistic to this process. Interestingly, many new types of wg-LacZ+ PM cells appear during the L2 stage and occupy either DV midline or anterior dorsal domain. Perhaps some of these wg-LacZ+ PM cells may play important roles in specifying head fate of the anterior dorsal domain (Cho, 2000).

The PM is an important source of inductive signals to control cell fates within the DP. According to the presented model, Hh acts differentially to localize Wg- and Dpp-expressing cells to the dorsal and ventral domains of the PM, respectively, in the L1 disc. Establishment of DV domains in the PM governs subsequent signaling from the PM to the DP for controlling the DV specificity and the level of Dl/Ser expression. This idea is supported by observations that ectopic Hh expression in the PM cells can induce Ser expression in the DP, consistent with spatiotemporal correlation of Hh and Ser expression pattern in the L1 and L2 discs (Cho, 2000).

Although this study has focused on signaling from the PM to the DP, the signaling may be bidirectional. It is conceivable that the extension of peripodial processes may depend on specific signaling cues provided from the DP. Such bidirectional signaling may be essential to coordinate DV boundary formation and disc growth in both layers. Whether the signaling molecules that are transferred from the PM cells are Hh/Wg/Dpp themselves and/or other molecules remains unanswered. Interestingly, Patched (Ptc), the receptor for Hh that is known to be upregulated transcriptionally by Hh signaling, is expressed in the DP but is more abundant in the PM. This suggests that Hh signaling may occur laterally and vertically, within the PM layer as well as between the two layers (Cho, 2000).

It has been shown that ectopic Hh+ clones generated anterior to the furrow induce ectopic furrow and retinal differentiation. Evidence presented here suggests that hh- or ectopic Hh+ clones in the PM and margin but not the DP can induce pattern changes in the DP. These data are also consistent with the observation that retinal differentiation is abolished in hh- clones induced in the disc margin but not in the middle of the eye field (Cho, 2000).

Another important question raised by the prospect of interepithelial signaling is how the signals are transmitted from the PM cells to the DP. One possibility is that signaling molecules are secreted from PM cells directly to the underlying region of the DP. Alternatively, these molecules may be transported to the DP via peripodial trans-lumenal extensions that contact specific target cells or reach in the vicinity of target areas in the DP. Hh signaling may be mediated by peripodial processes, although the former possibility cannot be excluded. Ser expression in the DP induced by ectopic peripodial Hh signaling often extends beyond the region directly underneath the Hh+ PM cells. Hh may diffuse from the processes to reach other nearby DP cells. Alternatively, PM cells may extend longer processes than what can be detected in the fixed tissues. It is also possible that the inductive event occurs earlier when the two cell populations are in closer contact and subsequently become displaced relative to one another as the epithelium grows (Cho, 2000).

Recent studies have shown that disc cells send out long and thin cytoplasmic extensions, named cytonemes. Cytonemes are actin-based extensions that grow from the apical surface of the DP cells toward the signaling center, the anterior-posterior boundary of the wing disc. Some of the peripodial extensions described in this study also show cytoneme-like long and thin processes, although it is not known whether the processes are also actin-based. The peripodial processes observed can be readily seen in fixed tissues, unlike cytonemes that cannot be detected in fixed discs. Furthermore, cytonemes extend from the DP cells and grow on the apical surface of the DP, while peripodial processes extend from the apical surface of PM cells across the disc lumen to the DP. In addition, the observation of different shapes of processes suggests that peripodial processes exist in multiple types (Cho, 2000).

Inductive signaling between two cell layers is an important mechanism of morphogenesis in vertebrate development. For instance, BMP4 signaling between optic vesicle and surface ectoderm is important for lens induction in vertebrates. Wnt signaling between the ectoderm and the mesoderm is also crucial for proper dorsoventral limb patterning. First shown to occur during Drosophila leg disc regeneration and now in the eye, peripodial signaling to the DP may be analogous to such inductive signaling in vertebrates. This study illustrates a novel mechanism of interepithelial signaling between PM and DP layers and its importance in eye disc patterning. Significantly, ablation or genetic disruption of the PM also affects development of the DP, providing additional evidence for peripodial signaling. Precise localization of receptors and downstream components for Hh, Wg, and Dpp in early eye discs will help in understanding how these signals are transmitted between the PM and the DP (Cho, 2000).

During Drosophila wing development, Hedgehog (Hh) signaling is required to pattern the imaginal disc epithelium along the anterior-posterior (AP) axis. The Notch (N) and Wingless (Wg) signaling pathways organise the dorsal-ventral (DV) axis, including patterning along the presumptive wing margin. A functional hierarchy of these signaling pathways is described that highlights the importance of the competing influences of Hh, N, and Wg in establishing gene expression domains. Investigation of the modulation of Hh target gene expression along the DV axis of the wing disc has revealed that collier/knot (col/kn), patched, and decapentaplegic are repressed at the DV boundary by N signaling. Attenuation of Hh signaling activity caused by loss of fused function results in a striking down-regulation of col, ptc, and engrailed (en) symmetrically about the DV boundary. This down-regulation depends on activity of the canonical Wg signaling pathway. It is proposed that modulation of the response of cells to Hh along the future proximodistal (PD) axis is necessary for generation of the correctly patterned three-dimensional adult wing. These findings suggest a paradigm of repression of the Hh response by N and/or Wnt signaling that may be applicable to signal integration in vertebrate appendages (Glise, 2002).

Short-range Hh signaling, partly through activation of Col function, is essential for correct AP patterning and differentiation of L3-L4 intervein tissue. N and Wg first define the DV boundary and later subdivide the region near this boundary into a number of distinct subregions that will eventually differentiate into wing margin bristles and vein tissue. These signals overlap spatially and temporally and lead to opposite fates. It is proposed that in and close to the DV boundary, N, Wg, and Hh signaling exist in a delicate balance to allow vein tissue, bristle, and sensory organ differentiation along the adult wing margin (Glise, 2002).

The Hh target genes col/kn and ptc, in contrast to en, are repressed in a wild type wing in cells corresponding to the presumptive wing margin. It has been demonstrated, using both gain- and loss-of-function experiments, that this repression is mediated by N signaling and that its inhibition results in aberrant morphogenesis of the wing. Hh signaling, achieved either by overexpression of Hh or loss of Ptc activity, is not sufficient to give maximum activation of Hh targets in cells of the prospective wing margin, suggesting that a finely tuned balance of activation and repression is required to achieve the appropriate biological output. However, overexpression of a stabilized form of Ci under the ptc-Gal4 driver results in the activation of Col in the prospective wing margin and defects in wing margin differentiation, indicating that N repression can be overcome by hyperactivity of the Hh signaling pathway. N signaling may lead to the repression of col, ptc, and dpp directly or it may act indirectly by affecting the ability of Ci to act as a transcriptional activator. Since expression of en, which requires the highest level of Hh signaling and Ci activity, appears immune to N repression, the former possibility is favored (Glise, 2002).

Analysis of Hh target gene expression close to the DV boundary in fu mutants has revealed that activation of col and dpp and up-regulation of ptc transcription is lost from the center of the wing pouch when Hh signaling is attenuated. Inactivation of fu results in a complete loss of en expression from the anterior compartment; however, the ectopic activation of en induced by submaximal Hh pathway activation (achieved by removing Pka-C1 activity) is similarly sensitive to inhibitory influences emanating from the DV boundary. In each case, the repression of Hh targets extends up to six rows of cells away from the DV boundary in the anterior compartment. This down-regulation is due to activation of the canonical Wg signaling pathway. At what level is this global inhibition of Hh targets by Wg signaling achieved? One possibility is that Hh target genes are regulated by the opposing effects of activator and repressor complexes acting directly via cis-acting sequences. Such a situation has previously been described for the stripe gene, which is directly activated by Hh signaling and repressed by Wg signaling in the Drosophila embryo. This mechanism would imply the existence of cis-acting sites for both the Ci and dTCF/pangolin proteins within the regulatory regions of each of the Hh target genes -- col, ptc, dpp, and en -- that have been analysed, a possibility that cannot currently be verified due to the absence of sequence information for all of these regions. An alternative scenario, however, is that Wg signaling modulates the activity of the Hh signaling pathway itself, perhaps at the level of the Su(fu)/Fu/Ci protein complex to affect localization and/or activity of Ci. Consistent with this possibility, recent studies have provided evidence for a direct protein-protein interaction between the vertebrate Su(fu) and ß-Catenin/Armadillo proteins. Such an interaction might mediate the attenuation of Hh signaling activity, for instance, by promoting the cytoplasmic retention of Ci. It should be noted, however, that no modulation of Ci distribution has been detected along the DV axis using the available anti-Ci antibodies in the fu mutant backgrounds, the pka-C1, or the pka-C1;wg clones (Glise, 2002).

The paradigm of modulation of Hh target gene expression by N and/or Wg may be applicable to signal integration in other systems. Hh signaling plays a remarkably similar role in patterning the Drosophila wing and vertebrate limbs. Specifically, Hh is involved in both short- and long-range signaling, elaborating upon preexisting informational cues to regulate anterior-posterior patterning, distal outgrowth, and proliferation. One of the vertebrate homologs of hh, sonic hedgehog (shh), is expressed in the zone of polarizing activity (ZPA), which is found at the posterior edge of the limb bud. Here, Shh is thought to control the expression of molecules, such as fibroblast growth factor-4 (fgf-4), which is required for limb outgrowth. Interestingly, Notch-1 expression and activity in the vertebrate limb localizes to the apical ectodermal ridge (AER), a thickening of the ectoderm along the distal tip of the limb bud, and molecular and misexpression studies have suggested that N signaling may regulate the number of AER progenitor cells (Glise, 2002).

Homologs of cut, one of the primary downstream effectors of N signaling in the wing, have also been characterised in the chick as participating in limiting the AER (Cux-1) and regulating polarizing activity derived from the ZPA (Cux-2). Furthermore, several vertebrate Wnts, homologs of Drosophila Wg, have also been implicated in the formation of the AER. It is possible that a similar signaling hierarchy exists in the vertebrate limb to allow for proximal-distal patterning and limb outgrowth (Glise, 2002).

According to the model, modulation of the expression of Shh targets near the intersection of the ZPA and AER in the developing vertebrate limb may be required for the differentiation of distal features. During mammalian tooth development, Shh and Wnt-7b are expressed in reciprocal patterns and a balance between the two activities is required for the formation of boundaries between oral and dental ectoderm. Overexpression of Wnt-7B in domains of endogenous Shh expression has been shown to repress ptch1 transcription and prevent tooth bud formation, while overexpression of Shh suppresses this effect (Glise, 2002).

The positioning and outgrowth of vertebrate epithelial appendages, such as feathers, whiskers, and hair, is another example where proper integration of positional information is essential for proximal-distal patterning. Overexpression of Wnt-7a in an in vitro feather reconstitution model leads to the inhibition of elongation along the PD axis and loss of AP asymmetry in the presumptive feather bud. These changes were correlated with diffuse Notch-1 and Shh expression, which are normally localized to the middle and posterior of the bud, respectively. The loss of localized Shh expression and loss of AP asymmetry as a consequence of ectopic expression of Wnt are reminiscent of the effect on Shh and hair follicle morphogenesis in mice expressing a stabilized form of ß-catenin in the skin (Glise, 2002).

Expression of Hh target genes requires activation of Ci by a multistep process. Initially, Ci proteolysis must be blocked. The cleavage of Ci into its truncated repressor form is promoted by Pka-C1-mediated phosphorylation. Removal of Pka-C1 leads to stabilization of Ci; however, ectopic target gene expression accomplished in such a manner does not reflect maximal Hh signaling. This is indicated by the absence of Col and Ptc expression close to the DV boundary in clones where Pka-C1 activity has been removed. In addition to preventing Ci cleavage, there is another Hh-dependent activation step that requires Fused kinase activity. Since fu is dispensable in animals where Suppressor of fused [Su(fu)] is also inactive and since Su(fu) has been shown to play a role in blocking Ci nuclear accumulation, it is hypothesised that Fu kinase activity is required in cells to facilitate nuclear localization of activated Ci (Glise, 2002).

Since fu has been implicated in localization of Ci, the down-regulation of Hh targets away from the margin described here may reflect the inability of activated Ci to enter the nucleus efficiently. A decrease in Ci activity is reflected in the decreased transcription of col, ptc, and dpp and an absence of en activation in the anterior compartment in a fu mutant background (Glise, 2002).

Recent results indicate that Hh signaling in the embryo, leading to ptc transcription and dorsal epidermal morphogenesis, occurs independently of Fu activity. However, Fu activity is indispensable in the embryo in Hh target cells expressing Wg, leading to the hypothesis that activation of Hh targets is repressed by an unknown localized factor in wg-expressing cells. It is speculated that Fu activity could be required in Wg-expressing cells in the embryo to overcome a repressive effect of Wg on Hh target gene activation. Further studies will be needed to determine whether this model can be extended to all the developmental processes involving Hh signaling (Glise, 2002).

The tissue-specific regulation of Vn signaling was investigated by examining vn transcriptional control and Vn target gene activation in the embryo and the wing. The results show a complex temporal and spatial regulation of vn transcription involving multiple signaling pathways and tissue-specific activation of Vn target genes. In the embryo, vn is a target of Spi/Egfr signaling mediated by the ETS transcription factor PointedP1 (PntP1). This establishes a positive feedback loop in addition to the negative feedback loop involving Aos. The simultaneous production of Vn provides a mechanism for dampening Aos inhibition and thus fine-tunes signaling. In the larval wing pouch, vn is not a target of Spi/Egfr signaling but is expressed along the anterior-posterior boundary in response to Hedgehog (Hh) signaling. Repression by Wingless (Wg) signaling further refines the vn expression pattern by causing a discontinuity at the dorsal-ventral boundary (Wessells, 1999).

blistered is expressed in the precursors of the terminal tracheal cells and in the future intervein territories of the third instar wing imaginal disc. Dissection of the blistered regulatory region reveals that a single enhancer element, which is under the control of the fibroblast growth factor (FGF)-receptor signaling pathway, is sufficient to induce blistered expression in the terminal tracheal cells. In contrast, two separate enhancers direct expression in distinct intervein sectors of the wing imaginal disc. One element is active in the central intervein sector and is induced by the Hedgehog signaling pathway. The other element is under the control of Decapentaplegic and is active in two separate territories, which roughly correspond to the intervein sectors flanking the central sector. Hence, each of the three characterized enhancers constitutes a molecular link between a specific territory induced by a morphogen signal and the localized expression of a gene required for the final differentiation of this territory (Nussbaumer, 2000).

Is blistered a marker for 'pro-intervein' territories? Two classes of genes involved in the late differentiation of the wing have been characterized previously: vein-promoting genes (e.g. ve/rho, vein) and vein-suppressing genes (e.g. bs, net, plexus). However, it has been proposed that blistered should be considered as an 'intervein-promoting gene' since its expression in wing cells is required to determine intervein fate. The expression of blistered is negatively regulated in vein territories by the EGF signaling pathway, whereas ve/rho expression is restricted to the future veins in the wing pouch and this expression is indirectly controlled by the long range organizers of the wing field. Therefore, the simplest mechanism for establishment of the vein-intervein network would be to spatially activate vein-specific genes, which, through their activity, would repress a wing-blade specific enhancer in the developing vein territories. Surprisingly, promoter analysis reveals that blistered expression in different presumptive intervein territories depends on distinct enhancers, which are controlled by different morphogen events. Congruent with this observation, it has been proposed that the vein-intervein network has to be considered on a stripe-by-stripe basis. Since the EGF signaling pathway represses blistered expression in the third instar wing disc, the blistered promoter could also contain DNA elements negatively regulated by EGF signaling. In summary, the expression of blistered in the future intervein cells is triggered by the integration of several signaling events, probably through distinct regulatory elements (Nussbaumer, 2000).

It has been proposed that once the vein and intervein domains have been demarcated independently, gene interactions may occur at vein-intervein boundaries to maintain and refine their respective domains. During metamorphosis, veins differentiate within the broader provein territories. The blistered expression domain within the future intervein sector C abuts the A/P boundary, although at the end of metamorphosis, this expression expands into adjacent posterior cells. Thus, since the provein territory L4 is posterior and abuts the A/P boundary, blistered in the third instar wing disc is excluded from this provein territory at the A/P boundary. It will be of great interest to determine whether blistered is excluded from the other provein domains of the third instar wing disc. In this case, blistered would have to be considered as an early marker for 'pro-intervein' territories and its expression would progressively expand during metamorphosis to the final intervein domains of the adult wing (Nussbaumer, 2000).

An enhancer (the boundary enhancer) has been identified that is activated by Hh signaling in the cells anterior to the A/P compartment boundary. In agreement with previous reports demonstrating a direct morphogenic role of Hh in the central region of the wing, this might indicate that the Hh signaling is required to trigger intervein differentiation through blistered expression in the intervein C domain. However, in contradiction to the activity of the boundary enhancer, blistered is expressed in smo mutant clones analyzed during pupal development. Noteworthy, the clones that were generated were analyzed during third instar, whereas blistered expression is detected later, 24-36 h after puparium formation. At this time, gene interactions between vein- and intervein-specific genes might be sufficient to maintain their respective, mutually exclusive expression domains. Thus, Hh would be required only for the early setting of blistered expression as a result of patterning the intervein sector C. Indeed, beta-galactosidase expression directed by the boundary enhancer is not detected in the wing of newly emerged flies. This indicates that in the presumptive intervein sector C, the early setting of blistered is controlled through the boundary enhancer, whereas the later expression might recruit another cis-regulatory element. The fact that the expression of blistered is observed in the posterior compartment of pupal wings, whereas the boundary enhancer is restricted to the anterior compartment in third instar discs, further supports this idea (Nussbaumer, 2000).

The boundary enhancer is directly regulated by Vestigial (Vg) and Scalloped (Sd) which form a complex on a 120-bp DNA sub-element. The wing-specific Vg-Sd complex restricts the activation of the boundary enhancer to the future wing, consistent with the finding that Ci can only activate it in the pouch region. Hence, the boundary enhancer integrates positional cues from the Vg-Sd transcriptional complex and the Hh signal. The gene knot/collier (kn), which encodes a putative DNA-binding protein acting downstream of the Hh signaling pathway, has been found to be required for the expression of blistered in the intervein sector C. Therefore, the Hh responsiveness of the boundary enhancer may be indirect and mediated by Kn. Alternatively, activation of the enhancer may require a molecular interaction between Ci and Kn. Therefore, it will be of prior interest to determine whether Kn and Ci directly regulate the boundary enhancer and cooperate for its activation. Further analysis of how the boundary enhancer integrates input from the Vg-Sd complex and Hh signaling will contribute to a molecular understanding of the synergistic activation of enhancers by signaling input and selector genes, a strategy that may be widely used to regulate gene expression during development (Nussbaumer, 2000).

One of the pleiotropic functions of scribbler is an effect of wing morphogenesis. This function has been addressed by Funakoshi (2001), who shows that sbb shapes the activity gradient of the Dpp morphogen through regulation of thickveins. Drosophila wings are patterned by a morphogen, Decapentaplegic, a member of the TGFbeta superfamily, that is expressed along the anterior and posterior compartment boundary. The distribution and activity of Dpp signaling is controlled in part by the level of expression of its major type I receptor, thickveins (tkv). The level of tkv is dynamically regulated by Engrailed and Hedgehog. sbb, termed master of thickveins (mtv) by Funakoshi, downregulates expression of tkv in response to Hh and En. mtv expression is controlled by En and Hh, and is complementary to tkv expression. mtv integrates the activities of En and Hh that shape tkv expression pattern. Thus, mtv plays a key part of regulatory mechanism that makes the activity gradient of the Dpp morphogen (Funakoshi, 2001).

The fact that tkv expression is repressed by hh at the A/P border and mtv is highly expressed in the same cells has prompted an examination of whether mtv mediates hh dependent tkv repression. tkv-lacZ levels were examined in clones of cells mutant for patched (ptc), which encodes the Hh receptor. Hh signal transduction is constitutively active in the absence of ptc activity. Anterior compartment ptc clones cause cell-autonomous repression of tkv. In the clones of cells mutant both for mtv and ptc, however, tkv levels are elevated as in the mtv singly mutant clones. This indicates that mtv mediates hh-dependent tkv regulation along the A/P border. mtv-lacZ expression was monitored within clones of cells mutant for smoothened (smo), which encodes a component of the Hh receptor complex and is required for Hh signaling. Within the clones located at the A/P border, mtv-lacZ expression is repressed, indicating that Hh signaling induces the high level of mtv expression along the A/P border. This was also confirmed by the fact that mtv-lacZ levels are elevated within pka mutant clones, in which Hh signaling is constitutively active. These results show that Hh represses tkv levels by upregulating its negative regulator, mtv (Funakoshi, 2001).

As described earlier, the basal tkv level in the P compartment is higher than it is in the A compartment and it is responsible for the asymmetric structure of the wing. The mtv expression pattern is complementary to the tkv expression pattern. The possibility that mtv is also responsible for regulating the basal level of tkv expression was also examined. Initially, it was asked whether the level of tkv is regulated by en in the P compartment. Within clones of cells mutant for en in the P compartment, the tkv-lacZ level is lower than that seen in the A compartment, suggesting that tkv expression is regulated by en. This is confirmed by the observation that, within clones of cells ectopically expressing en in the A compartment, the tkv level is elevated in comparison to that seen in the P compartment in an autonomous way. Within en and mtv double mutant clones, tkv transcription levels are derepressed as in mtv single mutant clones, indicating that mtv mediates en dependent tkv regulation. Ectopically expressed en downregulates mtv-lacZ in the A compartment, implying that low levels of mtv in the P compartment are under the control of en regulation. These results altogether indicate that en regulates the high basal level of tkv expression in the P compartment by downregulating mtv expression (Funakoshi, 2001).

It is concluded that the central region of the wing is patterned by Hh but not by Dpp although Dpp expression is induced by Hh in this region. This is because Dpp signaling is downregulated by Hh by upregulating mtv, which causes repression of tkv expression. The patterning by Hh appears to be ensured by lowering the Dpp signaling, which would otherwise interfere with the Hh morphogen activity, because upregulation of Dpp signaling by overexpressing tkv or by eliminating mtv activity can alter the vein patterns there. Therefore, the mtv-dependent tkv regulation is required both for Hh and Dpp morphogen activities. The patterning along the A/P border between veins 3 and 4 may be more complicated. The dorsal mtv mutant clone at the A/P border disrupts the vein pattern; vein 3 is displaced posteriorly, which is similar to the phenotype associated with sal mutant clones. The fact that mtv is downregulated in sal mutant clones might explain the phenotype if Mtv has a role in mediating Sal activity, which positions vein 3 through regulating target genes such as the iroquois gene complex. Further analysis is required to elucidate whether tkv function is linear or parallel to this regulatory cascade (Funakoshi, 2001).

The mechanism that shapes the tkv pattern and hence the Dpp morphogen gradient is unique in that it makes the mtv expression pattern that is made by integrating En and Hh signals complementary to the tkv pattern. Thus, the mtv expression pattern acts as a 'negative' for generating the tkv pattern. en positive cells initiate the cascade of the patterning along the A/P axis by expressing Hh, which both acts as short-range morphogen and induces the long-range morphogen, Dpp. Here it is proposed that not only does En induce expression of the morphogen, but it also shapes the morphogen activity gradient by regulating its receptor level via Mtv (Funakoshi, 2001).

The development of multicellular organisms requires the establishment of cell populations with different adhesion properties. In Drosophila, a cell-segregation mechanism underlies the maintenance of the anterior (A) and posterior (P) compartments of the wing imaginal disc. Although engrailed (en) activity contributes to the specification of the differential cell affinity between A and P cells, recent evidence suggests that cell sorting depends largely on the transduction of the Hh signal in A cells. The activator form of Cubitus interruptus (Ci), a transcription factor mediating Hh signaling, defines anterior specificity, indicating that Hh-dependent cell sorting requires Hh target gene expression. However, the identity of the gene(s) contributing to distinct A and P cell affinities is unknown. A genetic screen based on the FRT/FLP system has been to search for genes involved in the correct establishment of the anteroposterior compartment boundary. By using double FRT chromosomes in combination with a wing-specific FLP source, 250,000 mutagenized chromosomes were screened. Several complementation groups affecting wing patterning have been isolated, including new alleles of most known Hh-signaling components. Among these, a class of patched (ptc) alleles was identified exhibiting a novel phenotype. These results demonstrate the value of this setup in the identification of genes involved in distinct wing-patterning processes (Végh, 2003).

A total of 250,000 mutant chromosomes covering the X chromosome and both major autosomes were screened. Four complementation groups were identified that affected wing patterning similar to mutations in smo. The largest of these groups represents alleles in smo itself. Two groups exhibiting a subset of smo phenotypes represent new alleles of fused and collier/knot. Fused is a positive regulator of Hh signaling, and collier/knot is an Hh target gene required for the formation of the L3/L4 intervein region. Surprisingly, the remaining complementation group turned out to consist of novel ptc alleles with striking characteristics. Molecularly, they represent point mutations causing an amino acid substitution in either the first or the second large extracellular loop. In contrast to ptc null alleles, homozygous mutant clones failed to upregulate Hh target genes even in the presence of Hh. Together these findings suggest that the mutant proteins repress Smo constitutively, most likely because they fail to bind Hh. Animals mutant for trans-heterozygous combinations of these new ptc alleles with ptcS2 are fully viable. The ptcS2 product lacks the ability to repress Smo but is able to sequester, and hence bind to, Hh. The intragenic complementation that was observed suggests that both functions of Ptc, binding of Hh and repression of Smo, can be provided by individual proteins that possess only one of each. Recently, it was shown that a combination of two proteins, one consisting of the N- and the other the C-terminal half of Ptc, reconstitutes Ptc function. Although these experiments cannot be directly compared with the findings in this study, together they do suggest that Ptc function can be separated intramolecularly into independent modules of N- vs. C-terminal and extra- vs. intracellular domains. One possible scenario that could explain the intragenic complementation would be if Ptc proteins act in a multimeric complex (Végh, 2003).

Cadherin Cad99C is regulated by Hedgehog signaling in Drosophila

The subdivision of the Drosophila wing imaginal disc into anterior and posterior compartments requires a transcriptional response to Hedgehog signaling. However, the genes regulated by Hedgehog signal transduction that mediate the segregation of anterior and posterior cells have not been identified. The previously predicted gene Cad99C has been molecularly characterized and shown to be regulated by Hedgehog signaling. Cad99C encodes a transmembrane protein with a molecular weight of approximately 184 kDa that contains 11 cadherin repeats in its extracellular domain and a conserved type I PDZ-binding site at its C-terminus. The levels of cad99C RNA and protein are low throughout the wing imaginal disc . However, in the pouch region, these levels are elevated in a strip of anterior cells along the A/P boundary where the Hedgehog signal is transduced. Ectopic expression of Hedgehog, or the Hedgehog-regulated transcription factor Cubitus interruptus, induces high-level expression of Cad99C. Conversely, blocking Hedgehog signal transduction by either inactivating Smoothened or Cubitus interruptus reduces high-level Cad99C expression. Finally, by analyzing mutant clones of cells, it was shown that Cad99C is not essential for cell segregation at the A/P boundary. It is concluded that cad99C is a novel Hedgehog-regulated gene encoding a member of the cadherin superfamily in Drosophila (Schlichting, 2005).

The predicted Cad99C amino acid sequence contains the characteristic features common to members of the cadherin superfamily including a signal peptide, cadherin domains flanked by Ca++-binding sites, and a transmembrane domain, defining Cad99C as a member of this superfamily. A β-catenin binding site, which is present in the cytoplasmic domain of type-I and type-II cadherins providing a link to the actin cytoskeleton, was not identified. However, a conserved type I PDZ-binding sequence was identified at the C-terminus of Cad99C. PDZ domain-containing proteins commonly act as linkers that facilitate the assembly of large molecular complexes within cells, suggesting that Cad99C physically interacts with a cytosolic protein complex. Only few other cadherins have been reported to contain C-terminal PDZ-binding sequences. Cadherin 23, for example, a vertebrate cadherin lacking the consensus R1 and R2 β-catenin binding sites, associates with its C-terminal PDZ-binding site with the PDZ domain-containing protein harmonin b. Interestingly, harmonin b has been shown to be able to bundle actin filaments, providing a potential link between cadherin 23 and the actin cytoskeleton independent of β-catenin. It is therefore conceivable that Cad99C interacts with a PDZ domain-containing protein that provides a link to the actin cytoskeleton independent of β-catenin. The PDZ-binding sequence of Cad99C is adjacent to and overlapping with two predicted Ser/Thr kinase phosphorylation sites. The interaction between a PDZ-binding peptide and a PDZ domain can be disrupted by phosphorylation of the PDZ-binding sequence by Ser/Thr kinases, suggesting that the potential binding of Cad99C to a PDZ domain-containing protein could be regulated by phosphorylation (Schlichting, 2005).

Cad99C is expressed at low levels throughout the wing, haltere, and leg imaginal discs, whereas elevated levels of Cad99C expression are confined to a strip of cells along the A/P boundary of the wing imaginal disc pouch that is known to respond to the Hh signal. Even though anterior cells along the A/P boundary of haltere and leg imaginal discs as well as cells outside the pouch region of wing imaginal discs also respond to the Hh signal, no elevated level of Cad99C was observed in these cells, indicating that cad99C is a region-specific Hh target gene. The Cad99C protein profile resembles cad99C RNA levels, indicating that the elevated expression of Cad99C is mainly due to transcriptional and not translational or posttranslational regulation. High-level Cad99C expression was reduced to the low level present in cells far away from the A/P boundary in clones of cells lacking Hh signal transduction due to mutations in either smo or ci. Conversely, ectopic expression of either Hh or Ci was sufficient to increase Cad99C expression in the wing imaginal disc pouch, indicating that high-level cad99C expression is controlled by Ci-mediated Hh signaling (Schlichting, 2005).

Different Hh-regulated genes respond differently to Ci[act] and Ci[rep]. For example, the expression of dpp is regulated both by Ci[act] and Ci[rep], whereas hh and ptc only respond to one form of Ci, Ci[rep] or Ci[act], respectively. Like ptc, cad99c appears to respond exclusively to Ci[act]. This is inferred from five observations: (1) ectopic expression in posterior cells of Ci, which under the influence of Hh is converted to Ci[act], induces high-level cad99C expression; (2) misexpression of a constitutively active form of Ci, CiPKA4, also induces high levels of cad99C expression; (3) ci null mutant clones in the anterior compartment close to the A/P boundary, where Ci[act] is the predominant form of Ci, fail to upregulate cad99C expression; (4) expression of a constitutive repressor form of Ci, CiCell, does not reduce the low-level expression of cad99C; (5) ci null mutant clones in the anterior compartment away from the A/P boundary, where Ci[rep] is the prevailing form of Ci, show no increase in the expression of cad99C. Taken together, it is concluded that cad99C expression is regulated by Ci[act] and not Ci[rep] (Schlichting, 2005).

The segregation of cells at compartment boundaries is thought to depend on the differential adhesiveness (affinity) of cells on both sides of the compartment boundaries. Based on thermodynamic considerations, it has been proposed that cells will maximize the total strength of their adhesive interactions with neighboring cells by replacing weak cell–cell interactions with stronger ones. Cells with strong adhesive interactions will thus associate preferentially with one another and will segregate from less avidly adhering cells. As predicted by this model, cells expressing different levels of the same adhesion molecule segregate from one another. However, few adhesion molecules have been identified that can promote the differential adhesiveness of cells at compartment boundaries (Schlichting, 2005).

The maintenance of the A/P boundary in the developing Drosophila wing requires Ci-mediated Hh signal transduction in anterior cells. This suggests that Hh signaling may regulate the transcription of one or more genes that in turn affect the adhesiveness of anterior cells. Members of the cadherin superfamily are known to mediate adhesion between cells and several cadherins have been shown to be involved in cell segregation. Even though most cadherins implicated so far in cell segregation contain cytoplasmic β-catenin binding sites, which are absent in Cad99C, several cadherins lacking β-catenin binding sites have also been shown to mediate cell segregation. The discovery of a gene that is both regulated by Hh signaling and encodes for a cadherin, therefore, provides an attractive candidate for mediating the segregation of anterior and posterior cells. However, cad99C expression is not elevated in cells along the A/P boundary of haltere and leg imaginal discs or outside the pouch region of wing imaginal discs, indicating that if the elevated expression of Cad99C were important for cell segregation, this could not be a general mechanism for segregating anterior and posterior cells. However, since wing imaginal disc pouch cells differ in their expression profile from wing imaginal disc cells outside of the pouch, it is not inconceivable that different molecules could operate to segregate cells at the A/P boundary in different regions of the wing imaginal disc or in different imaginal discs (Schlichting, 2005).

A mutant allele of cad99C, termed cad99C57A, was generated in order to test whether Cad99C is required to segregate anterior and posterior cells. cad99C57A appears to be a null allele of cad99C based on four criteria: (1) sequencing of the genomic DNA revealed that the predicted promoter region, the transcriptional start site, and the coding sequence for the first 101 amino acid residues were deleted; (2) an RNA probe recognizing the 3′ region of the cad99C transcript, outside of the deletion present in cad99C57A, does not show detectable staining in wing imaginal discs from homozygous cad99C57A mutant larvae, indicating that the cad99C transcript levels are highly reduced; (3) an antibody directed to the C-terminus of Cad99C does not recognize a protein of the predicted size for Cad99C in extracts from wing imaginal discs of homozygous cad99C57A mutant larvae; (4) Cad99C immunoreactivity is highly reduced in homozygous cad99C57A mutant clones within wing imaginal discs (Schlichting, 2005).

The requirement for Cad99C to maintain the segregation of anterior and posterior wing imaginal disc cells was tested by clonal analysis. Using this assay, Cad99C was found not to be essential for maintaining the normal segregation of anterior and posterior cells. This result can be explained in several ways. (1) Cad99C does not play any role in the segregation of anterior and posterior cells. The Hh-dependent increase in expression of Cad99C may either be irrelevant for the function of Cad99C or might reflect an unrelated function. For example, Hh signaling is required for the patterning of the longitudinal wing veins L3 and L4. It is thus conceivable that Cad99C may play a role in this aspect of Hh signaling. However, vein patterning appeared to be normal in wings from homozygous cad99C57A mutants. (2) The activity of Cad99C in mediating the segregation of anterior and posterior cells is redundant with the activity of one or several of the remaining 16 cadherins in Drosophila. Different, partially redundant mechanisms contribute to the segregation of anterior and posterior cells. For example, one mechanism might be Cad99C-dependent whereas additional mechanisms may rely on cell surface proteins unrelated to cadherins or on cytoskeletal components (Schlichting, 2005).

The identification of cad99C as an Hh-regulated gene provides a starting point to investigate a cell biological mechanism used by Hh signaling to control the development of the Drosophila wing. It also provides a further step towards the functional characterization of all remaining members of the cadherin superfamily present in Drosophila that have so far only been predicted based on the genomic sequence (Schlichting, 2005).

Hedgehog function in legs

High levels of Hedgehog signaling activity are essential for medial-region patterning in Drosophila legs. In mid-to-late third instar leg discs, high levels of Hedgehog signals repress the transcription of pxb, a newly identified gene encoding a transmembrane protein expressed specifically in the anterior compartment (see CG14874). Misexpression experiments indicate that Pxb may serve as a Hedgehog signaling attenuator capable of acting prior to Hedgehog-Patched interactions, suggesting that Hedgehog signaling in leg discs includes a pxb-repression-mediated positive feedback loop. RNA interference and clonal analysis show that neither Wingless nor Decapentaplegic signaling is required for pxb repression but high levels of Wingless signaling activity are essential for patterning in the leg ventral medial region (Inaki, 2002).

hh9k is a temperature-sensitive allele of hh and 18°C is a permissive temperature with respect to viability and fertility. Flies homozygous for hh9k raised at 18°C show a reduction in the vein 3¯4 region area of adult wings, and unlike wild type flies, fail to express en in the wing-disc anterior compartment in mid-to-late third instar. Thus, at this temperature, Hh9k protein appears to possess lower stability and/or specific activity than wild type Hh (Inaki, 2002).

Detailed morphological examination of hh9k flies raised at 18°C indicate the presence of cuticle structure defects in the medial region of all three leg types. In wild type, each leg possesses eight bristle rows in tarsal segments 1¯5. Regions flanked by bristle rows 8 and 1 (8¯1 region) and bristle rows 4 and 5 (4¯5 region) are associated with strong ptc signals and are presumed to derive from ventral and dorsal anterior-compartment portions immediately adjacent to the A/P boundary, respectively. In the normal 8¯1 region of the distal-most tarsal segment 5, about seven hairs are present. In hh9k flies raised at 18°C, these and 8¯1 region hairs in tarsal segment 4 are almost completely abolished and the counterparts in other tarsal segments are significantly reduced in number. High levels of Hh signaling activity are thus presumed to be required for hair formation in the leg ventral medial region (Inaki, 2002).

In wild type, the distal end of the 4¯5 region is marked by a sensory bristle, hereafter referred to as the DM (distal-most) sensory bristle. This bristle was lost in 29% of legs of hh9k flies raised at 18°C, suggesting that high levels of Hh signaling activity are necessary for normal development of the peripheral nervous system (PNS) in the leg dorsal medial region. No appreciable loss of or structural defects in other bristles could be found (Inaki, 2002).

A pair of claws is situated at the distal-most end of each wild type leg. BarH1 is a homeobox gene required for normal morphogenesis of leg distal tarsal segments and pretarsus, and is expressed in claw basal cells, placing the medial Ptc-positive region between, and in cells surrounding the outer circumference of the pretarsus in late-third-instar larval and pupal stages. hh9k flies raised at 18°C always possess a pair of claws. No appreciable difference in claw-to-claw distance was detected between wild type and hh9k flies raised at 18°C when measured at the end of late third instar and normalized using the pretarsus diameter. Claws and distal tarsal segments are occasionally lost in some hh hypomorphic mutant clones. Thus, in leg discs, high levels of Hh signaling activity appear essential for determining the fate of at least a fraction of cells situated within the future medial region, while claw formation requires Hh signaling activity to a lesser extent than that of the medial-region hairs and the DM sensory bristle (Inaki, 2002).

Fused is a serine/threonine kinase involved in Hh signaling, and its kinase activity must be available for normal wing-disc expression of genes such as en and collier, all of which require high levels of Hh signaling activity for expression. In the fu kinase mutant, fu1, vein 3-4 region area of adult wings is reduced. Nearly all hairs in the 8-1 region disappear in legs hemizygous for fu1. The DM sensory bristle is lost in 16% of fu1/Y legs, but no apparent change in claw morphology or claw-to-claw distance could be detected. It follows from the present findings that, as with wing discs, Fu kinase activity in leg discs is required only for transducing high levels of Hh signals, essential for medial-region hair and DM sensory bristle formation (Inaki, 2002).

For further clarification of hh-dependent cell fate modulation along the A/P boundary in legs, search was made for enhancer trap lines in which reporter gene (lacZ) expression is positively or negatively regulated along this boundary in mid-to-late third instar leg discs and two lines, PXb41 and L71, were found. Only findings for PXb41 are presented here (Inaki, 2002).

Staining for lacZ protein and En have indicated that lacZ expression in PXb41 (pxb-lacZ expression) occurs in the anterior compartment of leg discs. In early third instar, pxb-lacZ expression abuts the En-positive domain, while in mid-to-late third instar, pxb-lacZ and en expression domains are separated from each other by a narrow gap along the A/P boundary. The posterior edge of pxb-lacZ expression in mid-to-late third instar is associated with gradation. Strong Ptc expression occurs in almost all cells, and only those cells, situated within the narrow gap, indicating that pxb-lacZ, expressed throughout the anterior compartment in early third instar, is specifically repressed along the A/P boundary in mid-to-late third instar larvae. Unlike wing discs, no En invasion into the anterior compartment is apparent in mid-to-late third instar larval leg discs. The narrow gap separating pxb-lacZ and en expression domains can be still seen in pupal stages, although the anterior en expression becomes apparent at these stages (Inaki, 2002).

pxb-lacZ expression is not restricted to leg discs. pxb-lacZ signals are also detected in anterior compartments of antennal and wing discs and embryonic segments. In antennal discs, pxb-lacZ, initially expressed throughout the anterior compartment, is repressed along the A/P boundary in mid-to-late third instar larval and pupal stages. In wing discs, no or little pxb-lacZ expression is observed except for the anterior-most region of the anterior compartment. In embryos, pxb-lacZ expression takes the form of eight stripes at stage 7. At stages 13 and 15, expression gaps between pxb-lacZ and en can be seen along the anterior and the posterior edges of each and every pxb-lacZ domain, respectively (Inaki, 2002).

In PXb41, P insertion occurred at 89A on the third chromosome. Genomic DNA fragments surrounding the P insertion site and relevant cDNA fragments showing in situ hybridization similar to pxb-lacZ expression in leg discs were cloned. The longest cDNA clone (clone a1) was 3 kb in length, this being consistent with 3 kb RNA noted in Northern blots. In situ hybridization using clone a1 DNA as a probe showed that hybridization signals distribute in the anterior compartments in a fashion very similar to pxb-lacZ expression; in wing discs, a marginal, if any, level of hybridization signals can be detected throughout the anterior compartment (Inaki, 2002).

Nucleotide sequence and 5' Race analysis disclosed P insertion to occur about 70 bp upstream of a putative RNA start (5'TTCAGTT). The longest cDNA (a1 sequence) contains an open reading frame (ORF) encoding a 636-amino-acid-long protein, referred to as pxb. Pxb protein has an N-terminal, presumptive transmembrane domain, 21 amino acids in length, and eight putative N-glycosylation sites in the C-terminal half, possibly indicating that Pxb is a type II transmembrane protein. The putative intracellular domain possesses a proline-rich stretch, a possible SH3 domain-binding site, whereas histidine-rich, acidic-amino-acid-rich, and tyrosine-rich stretches are found in the putative extracellular domain. The histidine-rich stretch, 44-amino-acid-long, includes a sequence consisting of six consecutive histidines and possible zinc finger structures, indicating that it may be a metal binding site. The acidic-amino-acid-rich region, 20 amino acids long, exhibits a marginal level of amino acid sequence similarity to Hh. No sequence homology is apparent in any other proteins and thus, Pxb constitutes a novel protein class (Inaki, 2002).

In mid-to-late third instar leg discs, pxb-lacZ and En expression domains are separated from each other by a narrow gap along the A/P boundary. Since a similar gap is also found to exist between pxb RNA and en-lacZ expression domains, gap formation is quite likely to be due to the absence of the expression of the pxb gene itself in an anterior-compartment striped region along the A/P boundary in mid-to-late third instar leg discs (Inaki, 2002).

Experments were carried out to determine whether high levels of Hh signaling activity are required for pxb (pxb-lacZ) repression along the A/P boundary, the following experiments were carried out. Leg discs were prepared from mid-to-late third instar hh9k larvae raised at 18°C and fu1 larvae, both seemingly devoid of high levels of Hh signaling activity, and stained for lacZ protein and En or pxb RNA. En/pxb-lacZ, and en-lacZ/pxb expression gaps virtually completely disappeared in both mutant animals. Ptc functions as a receptor for Hh and counteracts Hh signaling. The removal of Ptc activity significantly activates the Hh signal transduction pathway in a cell autonomous manner. Consistent with the notion that high levels of Hh signaling activity repress pxb transcription, cells within ptc mutant clones formed in the anterior compartment lack pxb-lacZ signals. Ci protein is a transcription factor in the Hh signaling cascade. In anterior-compartment cells lacking Hh signal reception, full-length Ci undergoes cleavage to become a short repressor form. In contrast, in anterior-compartment cells receiving Hh signals, such cleavage fails to occur and Ci serves as a transcriptional activator most probably after modification. Clones expressing an active form of Ci [Ci(m1-4)] were generated by the flip-out method. In these clones, pxb-lacZ signals were completely eliminated, indicating that Ci(m1-4) is capable of repressing pxb expression directly or indirectly. pxb repression along the leg A/P boundary is thus concluded to require high levels of Hh signaling activity, with consequent production of the activator form of Ci. Additional experiments show that repression by En in the posterior compartment restricts the expression of pxb to the anterior compartment throughout development, while pxb is repressed in anterior A/P boundary cells through an en-independent mechanism in mid-to-late third instar (Inaki, 2002).

To determine possible biological functions of pxb, flies lacking pxb activity were generated by imprecise P-element excision: The pxb gene in PXb4137-2 (allele name: pxb37-2) was found to be associated with a deletion uncovering the first exon. In flies homozygous for pxb37-2, no appreciable pxb RNA signals were detected, indicating that pxb37-2 is a null allele. pxb37-2 homozygotes are viable and fertile with neither obvious morphological defects nor appreciable change in the expression of putative Hh signaling targets such as ptc. pxb would thus appear dispensable for survival and morphogenesis at least in a wild type background (Inaki, 2002).

In wild type legs, pxb expression is specifically abolished in anterior-compartment cells considered to receive high levels of Hh signals and accordingly, pxb repression might be essential for normal development of these Hh-receiving cells. To test this possibility, pxb was misexpressed using the Gal4/UAS system. Flies having a UAS-pxb construct were crossed with those with either ptc-Gal4 or Dll-Gal4. pxb misexpression brings about a significant loss of 8-1 region hairs. UAS-pxb96-1 is a fly line possessing a single copy of UAS-pxb. UAS-pxb96-1/Y; Dll-Gal4 flies (flies hemizygous for the UAS-pxb transgene) were found to have 1.5 hairs on the average vs. 2.6 for UAS-pxb96-1/+; Dll-Gal4 flies (heterozygotes), indicating that hair formation is regulated in a pxb-dosage-dependent manner. Based on these results, it is concluded that effective repression of pxb along the A/P boundary is essential for normal leg morphogenesis (Inaki, 2002).

pxb misexpression along the A/P boundary has effects on PNS formation in scutella. Wild type flies possess four large sensory bristles in the posterior-most region of the scutellum, a tissue derived from a part of the anterior compartment immediately adjacent to the A/P boundary in the future notum (a part of the wing disc). There was no scutellar sensory bristle loss in hh9k flies raised at 18°C, while those raised at 20°C lost about two scutellar sensory bristles, indicating their formation to be governed by high levels of Hh signaling activity. pxb misexpression induces loss of scutellar sensory bristles. UAS-pxb96-1/Y; ptc-Gal4 males had 2.8 bristles on the average, whereas males homozygous for ptc-Gal4 showed the weak ptc mutant phenotypes and had 4.3. The fraction of UAS-pxb96-1/Y; ptc-Gal4 males possessing four scutellar bristles was only 0.25. Thus, pxb misexpression suppresses two Hh-signal-dependent developmental processes, hair formation and PNS formation in different tissues, and this may indicate that Pxb protein serves as an attenuator of Hh signaling in general (Inaki, 2002).

Pxb is a membrane protein and accordingly, may attenuate Hh signaling through modulating Hh activity prior to Hh and Ptc interactions. Thus, pxb was misexpressed in the Hh-producing but Ptc-lacking posterior compartment and hair formation was examined in the ventral medial region corresponding to the leg-disc anterior A/P boundary region receiving high levels of Hh signals. UAS-pxb was driven by hh-Gal4, a posterior-compartment-specific Gal4 driver. Hair formation in anterior A/P boundary cells is repressed significantly upon pxb misexpression in the posterior compartment. It may thus follow that Pxb is capable of modulating Hh activity before Hh binds to its receptor, Ptc (Inaki, 2002).

These results have shown that Pxb acts as a weak attenuator of Hh signaling. Pxb is a putative type II transmembrane protein, whose expression in leg discs, is repressed only in a strong Ptc-positive region situated along the A/P boundary and in this vicinity. Pxb appears to possess a marginal level of amino acid sequence similarity to Hh. Pxb may thus serve as an Hh signaling attenuator by competing with Hh for Ptc. However, this possibility appears less likely, since pxb misexpressed in the Hh-signal-producing but Ptc-negative posterior compartment results in Hh signaling reduction in the Hh-signal-receiving, anterior A/P boundary region. It is thus presumed that Pxb may attenuate Hh signaling through modulating Hh activity prior to and/or during Hh and Ptc interactions. Hh may be partly destabilized or squelched through interactions with Pxb. Hh is a zinc-containing protein and Pxb possesses a putative metal binding site such as a histidine repeat. Thus, Pxb might serve as a kind of chelating reagent, to deprive Hh of a zinc ion and eventually decrease Hh specific activity and/or stability. In vertebrates, a transmembrane protein, Hip, is capable of binding to Hh and serving as an attenuator of Hh signaling. Amino acid sequence homology does not appear to exist between Pxb and Hip, and thus they are Hh-attenuators belonging to different classes (Inaki, 2002).

The present work shows that high levels of Hh signaling activity are essential for normal cell fate determination in a narrow anterior-compartment region immediately adjacent to the A/P boundary in legs. The role of high levels of Hh signaling appears two-fold. (1) High levels of Hh signaling activity are required for the repression of pxb transcription in the anterior A/P boundary region. Since pxb encodes a putative attenuator of Hh signaling and is expressed throughout the anterior compartment in early third instar, the repression of pxb transcription in mid-to-late third instar may cause the abrupt increment of Hh signaling activity in the anterior A/P boundary region (Inaki, 2002).

(2) High levels of Hh signaling are required for the production of high levels of Wg and Dpp signals in the anterior ventral and dorsal A/P boundary region, respectively, in mid-to-late third instar. However, it should be noted that the repression of pxb transcription is not mandatory for the production of lower levels of Wg and Dpp signals. Indeed, wg and dpp are expressed in anterior A/P boundary cells with pxb expression in early third instar leg discs. Upon reduction of Wg signaling activity by RNAi, hair formation is virtually completely abolished in the leg ventral medial region, suggesting that high levels of Wg signaling activity are essential for the fate determination of leg medial cells. In contrast, Dpp signaling might be dispensable for dorsal medial cell fate determination as shown in wings (Inaki, 2002).

Misexpression of a constitutively active form of Arm and Dpp shows that neither ectopic Wg nor Dpp signaling activity is sufficient to repress pxb transcription. RNAi experiments also suggested that high levels of Wg or Dpp signaling are dispensable for pxb repression. Thus, pxb repression might be independent of Wg and Dpp signaling. wg and dpp are considered to be positively regulated by Ci, a central downstream component of Hh signaling. pxb expression is negatively regulated by the activated form of Ci. Therefore, should the activated form of Ci always serve as a transcriptional activator, this finding ought to indicate that there is a putative repressor gene (X) regulating pxb transcription, and X is activated by Ci in leg anterior A/P boundary cells in mid-to-late third instar. In wing discs, the expression of En, serving as a general repressor, begins to invade into the anterior A/P boundary region in mid-to-late third instar and regulates cell fate determination in the medial region. However, no anterior expansion of en is observed in mid-to-late third instar leg discs and no functional requirement of en for pxb repression in the anterior compartment was revealed by clonal analysis, indicating that X must not be en. Alternatively, repressor function may be directly provided by the activated form of Ci (Inaki, 2002).

The Drosophila wing is a classical model for studying the generation of developmental patterns. Previous studies have suggested that vein primordia form at boundaries between discrete sectors of gene expression along the antero-posterior (A/P) axis in the larval wing imaginal disc. Observation that the vein marker rhomboid (rho) is expressed at the center of wider vein-competent domains led to the proposal that narrow vein primordia form first, and produce secondary short-range signals activating provein genes in neighboring cells. This study examined how the central L3 and L4 veins are positioned relative to the limits of expression of Collier (Col), a dose-dependent Hedgehog (Hh) target activated in the wing A/P organizer. rho expression is first activated in broad domains adjacent to Col-expressing cells and secondarily restricted to the center of these domains. This restriction, which depends upon Notch (N) signaling, sets the L3 and L4 vein primordia off the boundaries of Col expression. N activity is also required to fix the anterior limit of Col expression by locally antagonizing Hh activation, thus precisely positioning the L3 vein primordium relative to the A/P compartment boundary. Experiments using Nts mutants further indicate that these two activities of N can be temporally uncoupled. Together, these observations highlight new roles of N in topologically linking the position of veins to prepattern gene expression (Crozatier, 2003).

In the Drosophila wing, L3 vein is decorated with campaniform sensory organs (CS). Formation of the L3 vein and sensory organs is generally thought to be topologically linked through expression of rho and the proneural genes ac and sc in overlapping A/P positions in third instar wing discs. In wild-type wings, CS overlap the posterior-most row of L3 vein cells. In heterozygous N55 mutant wings, veins are wider than in wild-type, due to defective partitioning of the provein into vein and intervein cells, but selection of sensory organ precursors (SOPs) is not affected. It was observed, however, that CS still overlap the posterior-most L3 vein cells, unlike what would be predicted if widening of L3 vein due to defective vein resolution were centered over its initial coordinate position. The position of SOP relative to the A/P border is not changed in N mutant discs, indicating that in the adult, the position of L3 vein is shifted anteriorly by one or two rows of cells. In order to determine whether this shift and the defect in vein resolution are coupled, the temperature-sensitive allele Nts2 was used. When Nts2 mutants are shifted to restrictive temperature between 104 and 128 h AEL [8-32 h after puparium formation (APF)], the L3 vein is broader, indicative of abnormal vein resolution, but, unlike N55/+ wings, the CS are now centered over the broader vein. Reciprocally, when a transient temperature shift is applied at 60-80 h AEL (mid-second/mid-third larval stages), the L3 vein retains a wild-type width but, as in N55/+, its position is shifted anteriorly relative to the CS. Nts mutant analysis thus reveals a new role of N in positioning the L3 vein relative to the A/P border, temporally uncoupled from its known role in partitioning provein into vein and intervein cells at the pupal stage. This has led to a detailed investigation of the molecular mechanisms involved in positioning L3 vein (Crozatier, 2003).

Longitudinal vein primordia can be visualized in third instar larval wing discs as a series of stripes of cells expressing provein genes, alternating with domains of D-SRF expression. col activates D-SRF expression in A/P organizer cells and positions L3 vein by limiting L3 vein competence to cells expressing iro-C but not col. Therefore col transcription was examined in N55/+ third instar wing discs; it is expanded toward the anterior by one to two rows of cells. The position of the SOPs were examined, using a neuralised (neu)-lacZ reporter gene (transgenic line A101). Whereas in wild-type, one row of cells separates SOPs from the anterior limit of Col expression, SOPs are found immediately adjacent to cells expressing high levels of Col protein in N55/+ discs. Counterstaining of discs with propidium iodide (which labels all nuclei) confirms that the position of SOPs relative the A/P border (anterior limit of hh/posterior limit of Col expression) is unchanged, leading to the conclusion that reducing N activity in third instar larvae specifically results in anterior expansion of Col expression. Col expression was then examined in clones of N mutant cells generated in a heterozygous N55/+ background and spanning the A border of Col expression; it was found not to be expanded further anteriorly. col expression is established in response to Hh in a dose-dependent manner. The present data indicate that: (1) only one or two rows of cell activate col in response to Hh in the absence of Notch signaling, and (2) the same expansion on col expression results from complete absence of N or 2-fold reduction of N signaling suggesting that col expression is very dose sensitive. Thus, the expansion of col expression observed in N55/+ discs indicates that N signaling locally antagonizes Hh activation of col transcription, to precisely position the posterior limit of the L3 vein primordium. Repression of col transcription by Notch signaling has already been reported in formation of an embryonic muscle and at the wing margin but the molecular mechanisms underlying this expression remain to be determined. iro-C expression is also expanded anteriorly in late 3rd instar larval discs in N55/+ mutants, indicating that the entire L3 vein-competence domain is shifted anteriorly. The opposite, posterior shift of iro-C expression (and consequently L3 vein position) observed in col mutant discs (this correlates with the posterior shift of L3 vein observed in these mutants) is linked to the modified range of Dpp signaling resulting from lack of col activity. Similarly, it is proposed that the anterior expansion of iro-C expression in N55/+ mutant discs reflects a modified range of Dpp signaling induced by anterior extension of Col expression. Thus, in wild-type discs the cross-regulation between Hh, N and Dpp signalling allows the positioning of the L3 vein primordium in register with CS. Next, the question of the relation between the A and P boundaries of Col expression and positions of L3 and L4 veins versus proveins was addressed. In conclusion, these observations highlight the importance of cross-talk between the Hh and N signaling pathways in assigning overlapping A/P positions to the L3 vein and associated sensory organs and the role of N in precisely positioning vein primordia, thus intimately linking prepattern to the vein resolution process (Crozatier, 2003).

Surgically fragmented Drosophila appendage primordia (imaginal discs) engage in wound healing and pattern regulation during short periods of in vivo culture. Prothoracic leg disc fragments possess exceptional regulative capacity, highlighted by the ability of anterior (A) cells to convert to posterior (P) identity and establish a novel posterior compartment. This AP conversion violates developmental lineage restrictions essential for normal growth and patterning of the disc, and thus provides an ideal model for understanding how cells change fate during epimorphic pattern regulation. Evidence is presented that the secreted signal encoded by hedgehog directs AP conversion by activating the posterior-specific transcription factor engrailed in regulating anterior cells. In the absence of hedgehog activity, prothoracic leg disc fragments fail to undergo AP conversion, but can still regenerate missing anterior pattern elements. It is suggested that hedgehog-independent regeneration within the anterior compartment (termed integration) is mediated by the positional cues encoded by wingless and decapentaplegic. Taken together, these results provide a novel mechanistic interpretation of imaginal disc pattern regulation (Gibson, 1999).

The observation that A cells switch to P identity during pattern regulation in L1 first thoracic leg disc fragments is a curious exception to the rule of compartmental lineage restriction during Drosophila appendage development. This prompted an inquiry into how AP conversion is achieved on the molecular level. Since En is normally expressed in all P cells, a test was performed to see if AP conversion correlates with activation of En in cultured disc fragments. Third instar L1 discs were cut into fragments and cultured in vivo. The anterior 1/4 (A1/4), consists of the anterior/dorsal 1/4 segment of the disc, and posterior 3/4 (P3/4) consists of the posterior 1/2 of the disc, in addition to the anterior/ventral 1/4 of the disc. Approximately 50% of A1/4 and P3/4 fragments produced novel En expression domains within 96 hours in vivo. The size and shape of nascent posterior compartments vary from small clusters of En-expressing cells to large domains almost indistinguishable from the endogenous P compartment. This suggests that stochastic variations (subtle differences in cut sites, disc morphology and precise age of the disc donor) might influence the timing and extent of pattern regulation. In a significant number of cases, however, A1/4 disc fragments regenerate a new P compartment, which restore the proportions of a whole disc, and P3/4 fragments duplicate a new En-expression domain in mirror-symmetric opposition to the endogenous P compartment (Gibson, 1999).

To directly observe the cellular origin of nascent En domains in regulating disc fragments, random GFP-labeled clones were generated in larvae 1 day prior to fragmentation and injection into host animals. Half of clone-bearing P3/4 fragments produce a duplicated En domain; many with a well-defined novel compartment boundary suitable for detailed analysis. Anterior cells are shown to be able to directly convert to P identity during pattern regulation in L1 disc fragments. The size and distribution of GFP + clones in duplicated discs provide additional insights into the dynamics of proliferation in the blastema. Nascent P compartments are never entirely composed of GFP + cells (0/12), indicating that AP conversion is a polyclonal event in duplicating disc fragments. Single GFP + clones often occupy up to 10%-20% of the whole duplicated area, suggesting that as few as 5-10 founder cells participate in duplicative growth. This agrees with a previous clonal analysis, which suggests that similar cell numbers generate disc duplications during cell-lethality-mediated pattern regulation (Gibson, 1999).

Endogenous en expression in the L1 disc is activated during embryogenesis by the transcription factor Fushi tarazu. However, ftz-lacZ expression is not detected in cultured disc fragments, suggesting an alternate mechanism for activation of en during pattern regulation. Ectopic hh activity is known to induces en in A cells of the abdominal tergites, and wing and leg imaginal discs, indicating that hh might be required to activate en in cultured disc fragments. To test this hypothesis, L1 discs from flies with temperature sensitive hh larvae raised at permissive temperature were fragmented and cultured under restrictive conditions. In the absence of Hh, cultured A1/4 fragments are small and extremely difficult to recover from hosts. Of those recovered, only a few regenerate new En domains, compared with 37% at the permissive temperature. The effect of hh loss is more pronounced in P3/4 fragments; none duplicate at restrictive temperature while 39% duplicate under permissive conditions. Loss of hh clearly blocks en-activation in both fragments. At non-permissive temperature, temperature sensitive hh disc fragments also fail to produce regenerated/duplicated posterior leg structures upon forced differentiation in host larvae. Taken together, these experiments demonstrate that hh is necessary to activate en and establish a new P compartment in cultured L1 disc fragments. Because loss of hh does affect endogenous En expression in cultured P3/4 fragments, the possibility that Hh is simply required to maintain En during culture can be ruled out (Gibson, 1999).

P3/4 fragments normally duplicate existing pattern elements, but only rarely regenerate missing structures. In the absence of hh, however, P3/4 fragments gain the ability to regenerate all missing anterior leg structures. This observation suggests a paradox: if hh is required for normal development, how do P3/4 fragments regenerate in its absence? During normal leg disc development, Hh secreted from P cells acts primarily through induction of wg and dpp in A cells along the compartment boundary. Secreted Wg and Dpp subsequently pattern both compartments through a variety of mechanisms, including activation of Distalless (Dll) at the center of the disc. Once established, late-third instar wg and dpp expression domains may be sufficient to direct anterior pattern regulation even if Hh levels are severely reduced. In intact discs, ectopic Wg interacts with Dpp to cause overgrowth and anterior pattern duplications suggesting that wg/dpp interactions might be sufficient to direct anterior compartment regeneration without direct input from hh. This assertion is supported by the fact that hh is not required for maintenance of dpp in third instar wing discs, and predicts that Wg and Dll domains are maintained in the absence of Hh in cultured disc fragments (Gibson, 1999).

To test this hypothesis, P3/4 fragments cultured in the absence of hh were immunostained with antibodies directed against Wg and Dll. Temperature sensitive hh P3/4 fragments cultured at restrictive temperature maintain Wg and Dll at reduced levels (relative to wild-type) in their endogenous domains. These fragments do not produce duplicated Wg or Dll domains. At permissive temperature Wg and Dll levels are also significantly reduced (possibly due to weak hypomorphic effects of temperature sensitive hh), but both domains are duplicated or expanded in most fragments analyzed. These data show that wild-type levels of Hh are required for activation, but not maintenance, of Wg and Dll domains in cultured disc fragments. It is concluded that P3/4 fragments regenerate missing anterior pattern elements in an hh-independent fashion and it is suggested that this occurs by a process of integration between established wg/dpp domains juxtaposed through wound healing (Gibson, 1999).

Peripodial cells contribute to a squamous epithelium that covers the columnar epithelium of the disc proper and participates in disc eversion and metamorphosis. In wing disc fragments, peripodial cells form a transient heterotypic contact with regulating columnar cells at the site of wound healing. Substantial evidence is found that a specific population of dorsolateral peripodial membrane cells act as the source of Hh in cultured L1 disc fragments. In both A1/4 and P3/4 disc fragments, peripodial cells express hh-lacZ and En prior to and during in vivo culture, and appear to fuse with anterior cells in the regenerating disc epithelium during wound healing. In A1/4 disc fragments, hh-lacZ is not expressed in nascent P cells until after a new En-domain is clearly visible. This rules out the possibility that loss of AP conversion in temperature sensitive hh mutants results from hh-dependent growth effects within the nascent P compartment and makes peripodial cells the sole potential source of Hh in regenerating A1/4 fragments. These observations indicate that Hh from peripodial cells activates en during a transient fusion between peripodial and columnar cells at the wound site. As confirmation, En-expressing peripodial cells still fuse to the wounded epithelium in the absence of Hh, but do not induce En in surrounding columnar cells. It is shown that L2 discs lack En/Hh-expressing peripodial cells and cannot regenerate posterior leg structures (Gibson, 1999).

The sensory organs of the Drosophila adult leg provide a simple model system with which to investigate pattern-forming mechanisms. In the leg, a group of small mechanosensory bristles is organized into a series of longitudinal rows, a pattern that depends on periodic expression of the hairy gene and the proneural genes achaete and scute. Expression of ac in longitudinal stripes in prepupal leg discs defines the positions of the mechanosensory bristle rows. The ac/sc expression domains are delimited by the Hairy repressor, which is itself periodically expressed. In order to gain insight into the molecular mechanisms involved in leg sensory organ patterning, a Hedgehog (Hh)- and Decapentaplegic (Dpp)-responsive enhancer of the h gene, which directs expression of h in a narrow stripe in the dorsal leg imaginal disc (the D-h stripe) has been examined. These studies suggest that the domain of D-h expression is defined by the overlap of Hh and high-level Dpp signaling. The D-h enhancer consists of a Hh-responsive activation element (HHRE) and a repression element (REPE), which responds to the transcriptional repressor Brinker (Brk). The HHRE directs expression of h in a broad stripe along the anteroposterior (AP) compartment boundary. HHRE-directed expression is refined along the AP and dorsoventral axes by Brk1, acting through the REPE. In D-h-expressing cells, Dpp signaling is required to block Brk-mediated repression. This study elucidates a molecular mechanism for integration of the Hh and Dpp signals, and identifies a novel function for Brk as a repressor of Hh-target genes (Kwon, 2004).

The D-h and V-h stripes are regulated by separate enhancers, which map between 32-38 kb 3' to the h transcription unit. ac stripes are not expressed until 6 hours after puparium formation (APF). The flanking narrow D-h stripe is positioned a few cells anterior to the compartment boundary, allowing expression of two dorsal ac stripes in the anterior compartment. V-h, however, is expressed directly adjacent to the AP boundary so that there is only one ventral ac stripe in the anterior compartment. Expression of each h stripe in its proper register is essential for positioning of the ac stripes and consequently for sensory bristle patterning in the adult leg. Focus was placed on the mechanisms that lead to expression of the D-h stripe in its precise register near the AP boundary (Kwon, 2004).

Expression of the endogenous D-h stripe is dependent on Hh signaling. In order to identify sequences that mediate Hh responsiveness, a dissection was undertaken of the D-h enhancer. The D-h enhancer maps to a 3.4 kb BamHI/EcoRI fragment located 32 kb 3' to the h structural gene. In third instar leg imaginal discs, this fragment directs lacZ expression in a dorsally restricted AP boundary-adjacent stripe. Two subfragments of the D-h enhancer were tested for the ability to drive reporter gene expression in leg imaginal discs. A 3' 2.4 kb HindIII/EcoRI subfragment of the D-h enhancer (REPE) directs no detectable reporter gene expression in leg imaginal discs. However, the complementary 5' 1.0 kb BamHI/HindIII fragment of the D-h enhancer drives expression in a stripe that is not dorsally restricted but rather traverses the entire length of the DV axis, suggesting it responds to Hh signaling in both dorsal and ventral leg cells. To determine whether Hh signals through the BamHI/HindIII fragment of the D-h enhancer, expression from a BamHI/HindIII-GFP transgene was assayed in leg clones lacking function of Smoothened (Smo), a transmembrane protein required for transduction of the Hh signal. Somatic clones lacking smo function were generated by FLP/FRT-mediated mitotic recombination. Cell-autonomous loss of GFP expression is lost in smo clones that overlap the GFP stripe. These observations imply that Hh signals through the BamHI/HindIII fragment, and therefore, this region is referred to as the D-h-Hh response element (HHRE) (Kwon, 2004).

Since the HHRE is Hh responsive, the element for the consensus binding site of the Hh pathway transcriptional effector, Ci was sought. Two potential Ci-binding sites (Ci-1 and Ci-2) were found, each of which matches the consensus, TGGG(A/T)GGTC, in a minimum of seven out of nine sites and binds the Ci zinc-finger domain (CiZn) in a electrophoretic mobility shift assay (EMSA). To determine whether the Ci-binding sites are required for HHRE-directed expression, point mutations were introduced into the Ci-1 and 2 sites. These mutations abolish Ci binding of the HHRE in vitro. Expression directed by the HHRE with a mutation in either the Ci-1 or Ci-2 sites is drastically compromised, and there is no detectable expression from an HHRE-lacZ transgene with both Ci sites mutated. Taken together, these studies indicate that D-h expression is activated primarily by the HHRE, through which Ci acts as an essential and direct transcriptional activator (Kwon, 2004).

Endogenous D-h expression is compromised in somatic clones lacking function of Mad, the transcriptional effector of Dpp signaling, and D-h-lacZ expression is severely decreased in leg imaginal discs with reduced dpp function. Furthermore, D-h-lacZ expression is ventrally expanded in wingless (wg) mutant legs, which have strong ventral dpp expression. These findings indicate a requirement for Dpp, in addition to Hh signal, for D-h expression. The most parsimonious model to explain how h integrates positive input from the Hh and Dpp signals, is that Mad acts synergistically with Ci through the D-h enhancer to activate D-h expression. However, Dpp is instead required to block REPE-mediated repression (Kwon, 2004).

Thus, the D-h activation element, HHRE, has two consensus Ci-binding sites, which bind Ci in vitro, and are required for its activity. In addition, HHRE-GFP expression is abrogated in clones lacking function of smo, a transducer of the Hh signal. These observations suggest that Ci acts directly through the HHRE to activate D-h expression. h is one of a number of genes, including dpp, patched (ptc), knot and araucan/caup (ara/caup), that have been identified as targets of Hh signaling in imaginal discs. These genes are each expressed in a stripe along the AP compartment boundary, but curiously, stripe widths among the genes varies as does register relative to the AP boundary. This has been explained in terms of differential response of Hh-target genes to the repressor and activator forms of Ci (Ci-R and Ci-A, respectively) found in anterior compartment cells. ptc, for example, has been proposed to respond only to the maximal levels of Ci-A found in cells nearest the AP boundary, while dpp responds to lower levels of Ci-A and also to Ci-R. The broad AP boundary stripe of HHRE-directed expression suggests that the HHRE is highly responsive to Ci-A. Differential response to Ci-R and Ci-A is thought to be controlled by cis-regulatory elements outside the local context (within 100 bp) of Ci binding sites in Hh responsive enhancers. Consistent with this hypothesis, an element, the REPE, has been identified that appears to modulate the response of the HHRE to Ci-A (Kwon, 2004).

Although Ci-A is an essential and important activator, which acts directly through the HHRE, it is unlikely that Ci-A function is sufficient for HHRE activity. Several studies have suggested that signal response elements in enhancers are generally not sufficient to activate gene expression. Rather, the transcriptional effectors of signals must act cooperatively with other activators to direct robust expression of target genes. This phenomenon, which has been termed 'activator insufficiency,' presumably prevents promiscuous activation of potential target genes. It is likely then, that other sites in the HHRE are required in addition to the Ci sites for expression directed by this element. For example, since the HHRE drives reporter gene expression in the wing and antennal discs as well as the leg, it might be expected that a common factor expressed in all three discs acts through the HHRE in combination with Ci. Alternatively, the enhancer might harbor sites that respond to factors specific to each disc type (Kwon, 2004).

Together, these observations are consistent with a model in which Ci, acting through the HHRE, activates D-h expression. The domain of HHRE activity can be divided into two zones, 1 and 2. The HHRE has the potential to direct expression in both zones 1 and 2, but its activity is restricted to zone 1 by Brk and perhaps a second factor, X, which binds the CRE. In zone 2 cells, Brk would bind to the CMB and repress HHRE-directed expression. It is proposed that zone 1 is defined by the overlap of Hh and high-level Dpp signaling. Dpp promotes D-h expression by repressing brk expression in zone 1. However, the presence of Mad-binding sites in the CMB suggests the potential for a more direct role for Mad in D-h regulation, perhaps in competing with Brk for binding to the CMB, or in directly mediating repression. Confirmation of a role for the Mad sites awaits further analysis of the D-h enhancer (Kwon, 2004).

back to Hedgehog Targets of Activity: part 1/2

hedgehog continued: Biological Overview | Evolutionary Homologs | Transcriptional Regulation | Protein Interactions | Developmental Biology | Effects 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.