Body structures of Drosophila develop through transient developmental units, termed parasegments, with boundaries lying between the adjacent expression domains of wingless and engrailed. Parasegments are transformed into the morphologically distinct segments that remain fixed. Segment borders are established adjacent and posterior to each engrailed domain. They are marked by single rows of stripe expressing cells that develop into epidermal muscle attachment sites. The positioning of these cells is achieved through repression of Hedgehog signal transduction by Wingless signaling at the parasegment boundary. The nuclear mediators of the two signaling pathways, Cubitus interruptus and Pangolin, function as activator and symmetry-breaking repressor of stripe expression, respectively (Piepenburg, 2000).

A cis-acting element of stripe (sr) has been identified that specifically directs gene expression in segment border cells during embryogenesis. This element was used to illuminate the molecular mechanism underlying segment border selection. The results show that Hedgehog (Hh) signaling can activate gene expression in two rows of cells, one on each side of the engrailed (en) expression domain. However, anterior Hh signaling causes the maintainance of wingless expression anterior to the PS boundary. Wg in turn antagonizes Hh-dependent gene expression and thereby prevents the formation of segment border cells anterior to the en domain. Hh and Wg activities relevant for the selection of segment border cells are mediated by functional binding sites of their nuclear mediators, Cubitus interruptus (Ci) and Pangolin (Pan), respectively within the sr cis-acing element. The data suggest that the segment border is established in response to the asymmetry of Wg signaling at the PS boundary (Piepenburg, 2000).

stripe (sr) is expressed in all precursors of the epidermal muscle attachment sites, including those marking the segment border in the Drosophila larvae. To obtain an early molecular marker for the segment border corresponding to the row of cells posteriorly adjacent to the en expression domain, a 1.9 kb enhancer element of the sr gene (sr1.9) was isolated that is both necessary and sufficient to direct transgene-dependent lacZ expression in segment border precursor cells in a dorsal and lateral position of the embryo. Expression of the reporter gene is activated in parallel with sr, which is first expressed during late stage 10. At this time, the initial equal distribution of Wg has already become asymmetric, meaning that the protein spreads anteriorly over a range of maximally five cells but is restricted to only one row of cells directly adjoining the posterior margin of the expression domain. sr acts as a transcription factor required for setting up the cell fate of the muscle attachment sites which mark the segment border of the fly. Thus, sr1.9-dependent reporter gene expression can be employed to study the transregulatory requirement for positioning the segment border cells within the PS (Piepenburg, 2000).

Subfragments of the sr1.9 element lacking sr239 failed to activate discernible gene expression, whereas sr239 directs sr1.9-like gene expression in the row of cells posterior to the en domain. Moreover, sr239-dependent gene activation could be achieved upon ectopic CiZn/C (the active form of Ci) expression within the en domain. Thus, the Ci binding site-containing sr239 element is both necessary and sufficient to activate gene expression in segment border precursor cells, and it is sufficient to mediate gene activation in response to dominant active Ci. It was next asked whether the two Ci binding sites within the sr239 element are necessary to mediate Hh-dependent gene activation. For this experiment, sr239 variants lacking either one or both Ci binding sites were generated. sr239 variants lacking only one functional Ci binding site can mediate gene expression in the correct spatial pattern, but the level of expression is strongly reduced. In contrast, lack of both Ci binding sites abolished Hh/Ci-dependent gene activation completely. In summary, these findings establish that Ci activates gene expression in segment border cells. They confirm, by direct means, that Hh signaling acts not only anteriorly and across the PS boundary to maintain Wg activity, but functions in a symmetric fashion and thereby determines the position of the segment border within the PS (Piepenburg, 2000).

Since Hh signaling appears to be symmetric, it was of interest to know why the sr1.9 element fails to mediate gene activation in the row of cells anterior to the en domain. An explanation for this phenomenon would be that signaling by Wg causes region-specific repression, preventing gene activation by Hh-dependent Ci. To test this proposal, Wg was ectopically expressed in the ptc domain and the change of sr1.9-driven gene expression was examined in such embryos. Furthermore, the sr1.9-mediated gene activation was analyzed in embryos mutant for lines (lin), sloppy paired (slp), and naked (nkd). Each of these mutant embryos express en, but the wg pattern is altered (Piepenburg, 2000 and references therein).

sr1.9-mediated gene expression was abolished in response to ectopic Wg activity. The same effect was observed in nkd mutant embryos where Wg is expressed at each side of an enlarged en domain. Conversely, in slp and lin mutants where Wg activity is not maintained, sr1.9-mediated gene expression is found in two rows of cells, one on each side of the en domain. Furthermore, sr1.9-mediated gene expression was also observed in cells anterior to the region of en expression in embryos in which late Wg activity was abolished due to a temperature-sensitive wg mutation. This establishes that the repression of Hh action in the cells anterior to the wild-type en domain is dependent on Wg activity (Piepenburg, 2000).

To explore how Wg exerts its repressing function, whether the Drosophila TCF/Lef1 homolog Pangolin (Pan), the nuclear mediator of Wg activity, can directly interact with the sr239 element was examined. Pan in vitro binding sites were found next to and partially overlapping the Ci binding sites. Deletion of one Pan binding site that leaves the Ci binding sites intact (sr239DeltaPan), resulted in gene activation anterior to the en domain. In contrast to sr239-mediated gene expression that can be suppressed by ptc-Gal4-driven Wg activation, sr239?Pan-mediated gene expression is not abolished in response to ectopic Wg activity (Piepenburg, 2000).

It is known that Pan can associate with corepressors such as dCBP or Groucho. Upon reception of the Wg signal, Pan is switched into an activator of transcription by association with Armadillo, a coactivator of Wg target genes. The findings in this study suggest an alternative mechanism since the Pan binding sites of the sr1.9 and sr239 elements mediate Pan-dependent repression in cells with high Wg activity. This repression is necessary and sufficient to antagonize Ci-dependent transcriptional activation. Pan could thereby exert this function by competing sterically for Ci binding, by short-range quenching of Ci-mediated gene activation, or by active repression. Each way, Pan would allow for the formation of only one row of segment border cells within each PS by repressing the Hh-dependent sr activation in the wg domain (Piepenburg, 2000).

Low-affinity transcription factor binding sites shape morphogen responses and enhancer evolution

In the era of functional genomics, the role of transcription factor (TF)-DNA binding affinity is of increasing interest: for example, it has recently been proposed that low-affinity genomic binding events, though frequent, are functionally irrelevant. This study investigated the role of binding site affinity in the transcriptional interpretation of Hedgehog (Hh) morphogen gradients. It is noted that enhancers of several Hh-responsive Drosophila genes have low predicted affinity for Ci, the Gli family TF that transduces Hh signalling in the fly. Contrary to an initial hypothesis, improving the affinity of Ci/Gli sites in enhancers of dpp, wingless and stripe, by transplanting optimal sites from the patched gene, did not result in ectopic responses to Hh signalling. Instead, it was found that these enhancers require low-affinity binding sites for normal activation in regions of relatively low signalling. When Ci/Gli sites in these enhancers were altered to improve their binding affinity, patterning defects were observed in the transcriptional response that are consistent with a switch from Ci-mediated activation to Ci-mediated repression. Synthetic transgenic reporters containing isolated Ci/Gli sites confirmed this finding in imaginal discs. It is proposed that the requirement for gene activation by Ci in the regions of low-to-moderate Hh signalling results in evolutionary pressure favouring weak binding sites in enhancers of certain Hh target genes (Ramos, 2013).

This study present in vivo evidence corroborating previous findings that multiple tissue-specific enhancers require low-affinity Ci binding sites for optimal activation by Hh/Ci. Most of the Hh target enhancers identified up to this point in Drosophila and mouse are regulated by degenerate Ci/Gli binding sites of low predicted affinity. The prevalence of these non-consensus sites in Hh target enhancers across species demonstrates their importance in regulating the Hh response. The transcriptional relevance of low-affinity TF binding is not limited to Hh/Ci regulated enhancers. For instance, two phylogenetically conserved low-affinity binding sites in the mouse Pax6 lens enhancer have been shown to be critical to promote gene expression at the right stage of development (Ramos, 2013).

A mechanistic explanation is provided as to why these Hh/Ci-regulated elements require low-affinity sites to activate transcription in cells with moderate signalling levels. Clusters of high-affinity sites mediate a restricted response in cells with high levels of Hh signalling, most likely as a result of cooperative interactions among Ci-Rep molecules in highly occupied Ci binding sites, whereas clusters of low-affinity sites mediate a broader response by having lower occupancy by Ci. Using synthetic enhancer reporters with high- or low-affinity Ci binding sites, this effect was confirmed in the wing, but not in embryos. This tissue-specific discrepancy may imply a context-dependent function for some non-consensus Ci binding sites. As in the Pax6 lens enhancer (Rowan, 2010), it is possible that some low-affinity binding sites are required specifically during earlier stages of development to interpret overall lower levels of Hh signalling (Ramos, 2013).

Finally, clues are provided as to additional regulatory inputs into dppD by showing a requirement for conserved consensus homeodomain (HD) binding sites. Cooperation between Glis and HD proteins has been recently shown in the mouse neural tube. In this case, HD proteins are critical to repress Hh-regulated neural tube enhancers, whereas in dppD they are critical to activate gene expression (Ramos, 2013).

The limited number of known, experimentally confirmed, direct Hh/Gli target enhancers may reflect the widespread, practical tendency to search for consensus or near-consensus motifs, and to focus on the highest peaks of TF-DNA binding, when hunting for cis-regulatory sequences. From a biochemical standpoint—for example, when mining ChIP-seq data—low-affinity DNA-binding interactions are troublesome because they are much more common, by definition, than the top 1% of peaks. It is important to note that iy is not always useful to strictly equate ChIP peak height with TF binding affinity, nor to equate in vitro binding or in silico 'motif quality' with in vivo TF occupancy, though these properties may often be roughly correlated. Separating the weak but functional binding events from weak and non-functional binding events is extremely challenging, and some have proposed that low-affinity genome-binding interactions can be categorically ignored. This certainly simplifies the problem from a computational perspective, but the findings discussed here and elsewhere suggest a risk of discarding functional sequences. Similar challenges confront in silico genomic screens to identify clusters of predicted TF binding sites: these necessarily filter out binding events of low predicted affinity, because there are many more predicted low-affinity binding motifs than consensus high-affinity motifs in any given sequence. Binding site predictions have been supported by taking evolutionary sequence conservation into account, but this risks filtering out true positives: as shown in Ci motif alignments, lower-affinity binding sites seem to be less constrained with respect to sequence variation, even in cases when the presence of the site itself is highly conserved. This is presumably because, for each non-consensus binding motif, there are multiple alternative sequences with similar affinity and thus equivalent functionality. Importantly, this type of degenerate motif conservation is easily missed: for example, some of the well-conserved Ci motifs described in this study are not properly aligned in the UCSC Genome Browser, because they do not constitute contiguous blocks of perfect sequence identity. To avoid these pitfalls, it is important to use phylofootprinting approaches that account for these alignment flaws. In contrast to most of the low-affinity binding sites discussed in this study, optimal-affinity Ci motifs in the ptc enhancer have been preserved throughout the evolution of the genus Drosophila, and perhaps much farther: GACCACCCA motifs occur in promoter-proximal regions of multiple vertebrate orthologues of ptc (Ramos, 2013).

Evolutionary enhancer sequence alignments, along with limited experimental data, also suggest that, although many predicted low-affinity sites are poorly conserved, overall TF occupancy on an enhancer may be maintained despite significant sequence turnover. This may occur either through the rapid gain and loss of individual sites, or through the maintenance of relatively weak binding affinity at a site that is unstable at the level of DNA sequence. While this last idea requires further direct testing, it is consistent with the fact that Gli sites of moderate predicted affinity have many sequence variants of similar quality, whereas the highest-affinity motifs have far fewer alternatives of similar quality. In other words, there are many more ways to be a weak binding site than a strong site. For example, among all possible 9-mer sequences, there are 654 motifs with Ci matrix similarity scores between 70 and 75 (inclusive), but only 12 motifs with scores between 90 and 95, and one motif with a score above 95. Therefore, weaker binding sites, and the enhancers containing them, have a far greater volume of sequence space in which to roam without strongly impacting transcriptional output. A thermodynamics-based simulation of enhancer evolution has shown that there is a greater number of fit solutions using weak TF sites than using high-affinity sites for a given gene expression problem (Ramos, 2013).

Equally consistent with the view of TF binding site evolution is the fact that it is much easier (that is, more likely) to create a low-affinity, non-consensus binding motif with a single mutation than a high-affinity consensus motif. An enhancer-sized DNA sequence can acquire a weak Gli motif with single-nucleotide substitutions at any of a large number of positions, as demonstrated by simulations. These arguments may help to explain why sequence conservation is not a foolproof test of the functional relevance of non-consensus TF binding sites (Ramos, 2013).

While there is no simple answer to the technical challenges facing those who hunt enhancers, the findings described in this report lead to a conclusion that low-affinity TF-DNA interactions, mediated by non-consensus and often poorly conserved sequence motifs, play important and widespread roles in developmental patterning and cis-regulatory evolution, and therefore cannot be safely ignored (Ramos, 2013).

Transcriptional Regulation

In the Drosophila embryo, the correct association of muscle cells with their specific ectodermally derived tendon cells, also known as epidermal muscle attachment or EMA cells, is achieved through reciprocal interactions between these two distinct cell types. Vein, a neuregulin-like factor secreted by the approaching myotube, activates the EGF-receptor signaling pathway within the tendon cells to initiate tendon cell differentiation. kakapo is expressed in the tendons and is essential for muscle-dependent tendon cell differentiation. Kakapo is a large intracellular protein and contains structural domains also found in cytoskeletal-related vertebrate proteins (including plakin, dystrophin, and Gas2 family members). kakapo mutant embryos exhibit abnormal muscle-dependent tendon cell differentiation. The expression of delilah, stripe, and beta1 tubulin is induced in the epidermal attachment cells as a result of the EGF-receptor pathway activation by the neuregulin-like growth factor, Vein (Yarnitzky, 1997). Vein is secreted by mesodermal cells underlying the EMA cells. Vein protein localization is restricted to the muscle-tendon junctional site in wild-type embryos. However, in kak mutant embryos, Vein protein is not localized and appears rather diffuse. This altered pattern of Vein may explain the multiple number of cells expressing delilah and stripe: since Vein is not strictly localized at a given muscle-tendon junction site, it apparently weakly activates the EGF-receptor pathway in neighboring cells as well. It is presumed that the only cells that can respond to the ectopic Vein protein are the competent population of EMA cells, defined by the early expression of stripe. These cells express stripe during early developmental stages in a muscle-independent manner and normally lose their stripe expression by stage 16 of embryonic development. When these competent EMA cells receive the muscle-derived Vein signal, the expression of stripe and delilah is reactivated. It appears that only this population of cells is capable of responding to Vein, since the pattern of the ectopic Stripe- or Delilah-expressing cells in the kak mutant embryos resembles that of the early population of Stripe-expressing cells. The reduced levels of beta1 tubulin mRNA in the mutant tendon cells may also result from the abnormal pattern of Vein localization, since lower levels of Vein may not be sufficient to induce maximal beta1 tubulin expression. It therefore appears that the primary defect in kak mutant embryos stems from the lack of Vein accumulation at the muscle-tendon junctional site (Strumpf, 1998).

Changes in the extracellular matrix (ECM) govern the differentiation of many cell types during embryogenesis. Integrins are cell matrix receptors that play a major role in cell-ECM adhesion and in transmitting signals from the ECM inside the cell to regulate gene expression. In this paper, it is shown that the PS integrins are required at the muscle attachment sites of the Drosophila embryo to regulate tendon cell differentiation. The analysis of the requirements of the individual alpha subunits, alphaPS1 and alphaPS2, demonstrates that both PS1 and PS2 integrins are involved in this process. In the absence of PS integrin function, the expression of tendon cell-specific genes such as stripe and beta1 tubulin is not maintained. In addition, embryos lacking the PS integrins also exhibit reduced levels of activated MAPK. This reduction is probably due to a downregulation of the epidermal growth factor receptor (Egfr) pathway, since an activated form of the Egfr can rescue the phenotype of embryos mutant for the PS integrins. Furthermore, the levels of the Egfr ligand Vein at the muscle attachment sites are reduced in PS mutant embryos. Altogether, these results lead to a model in which integrin-mediated adhesion plays a role in regulating tendon cell differentiation by modulating the activity of the Egfr pathway at the level of its ligand Vein (Martin-Bermudo, 2000).

Two physiologically distinct types of muscles, the direct and indirect flight muscles, develop from myoblasts associated with the Drosophila wing disc. The direct flight muscles (DFMs) are specified by the expression of Apterous, a Lim homeodomain protein, in groups of myoblasts. This suggests a mechanism of cell-fate specification by labelling groups of fusion competent myoblasts, in contrast to mechanisms in the embryo, where muscle cell fate is specified by single founder myoblasts. In addition, Apterous is expressed in the developing adult epidermal muscle attachment sites. Here, it functions to regulate the expression of stripe, a gene that is an important element of early patterning of muscle fibers, from the epidermis. These results, which may have broad implications, suggest novel mechanisms of muscle patterning in the adult, in contrast to embryonic myogenesis (Ghazi, 2000).

In examining the adult expression pattern of embryonic muscle founder markers, expression of an ap reporter was observed in a pattern that suggested specific roles in myogenesis. The DFMs, located dorsolaterally in each hemisegment of the adult mesothorax, show reporter gene activity. No staining was seen in the indirect flight muscles (IFMs), which constitute the bulk of the muscles of the dorsal mesothorax. Amongst DFMs, the most conveniently identifiable ones are muscles 49-58. Muscles 49 and 51-55, show staining for the ap lacZ reporter gene in different planes of foci (Ghazi, 2000).

Epidermal attachment sites for muscles in the adult fly are identifiable by anatomical examination and by expression of stripe (sr). sr marks all muscle attachment cells, both in the embryo and the adult. The pattern of sr expression during pupal development has been studied and the attachment sites for the IFMs identified. ap lacZ adult expression is also seen in the thoracic epidermis in the regions where muscles attach (Ghazi, 2000).

On the third instar wing imaginal disc the presumptive notum shows low ap lacZ and Ap protein expression distinct from the high levels seen in the presumptive dorsal wing. Although the presumptive notum is a dorsal structure, ap does not seem to have a selector function to define 'dorsalness' in the mesothoracic trunk as it does in the dorsal wing blade. ap expression in the pupal epidermis changes temporally beginning with an early expression in broad regions of the dorsal notal epidermis and a subsequent localization to restricted domains, including the attachment sites of the DLMs. At 18 hours after puparium formation (APF), ap lacZ expression is seen in regions that included the developing anterior attachment sites of the DLMs. This expression eventually narrows down to the anterior attachment sites and very closely abuts the posterior attachment sites of the DLMs. This can be observed by simultaneous labelling of developing pupae for ap and sr. By 36 hours APF, when a complete set of DLM fibers is in place, ap co-localization with individual muscle-attached sr-expressing tendon cells is clearly seen. This attachment site expression continues in the adult. The same pattern is seen on labelling with Ap-specific antibodies (Ghazi, 2000).

Expression of ap in epidermal attachment sites of muscles and the phenotypes noticed in ap mutants suggests a regulatory role for the gene in the development of muscle attachment sites. stripe (sr), the earliest known marker for epidermal attachment sites, was chosen as a potential target for regulation by ap in mediating its epidermal function, and its expression was examined in wing discs of ap4 homozygotes. sr is crucial for differentiation of epidermal cells into muscle-attached tendon cells. sr is expressed in discrete domains in the wing disc, which will form the attachment sites of adult thoracic muscles. sr expression in the disc commences very late in third larval instar and is consistently seen as pupation is initiated. Hence the 0 hours APF white prepupal stage was chosen for examination of sr expression in ap mutants. sr expression is either completely lost or drastically reduced in ap4 wing discs. Animals that reach pupal stages show severe reduction in sr levels. These results are further strengthened by the observation that ap and sr interact with each other genetically to affect IFM development. The ap4 mutation is completely recessive, as is a sr recessive lethal. In a transheterozygous combination, the two alleles show defective IFMs in a significant population of animals. Further, an enhancer trap insertion at the sr locus that shows a very mild recessive phenotype, enhances the IFM defects of ap4 and such animals also show a dark midnotal stripe that is characteristic of strong, viable sr alleles. This suggests that ap functions in IFM patterning by influencing attachment site development by the regulation of sr (Ghazi, 2000).

How repeating striped patterns arise across cellular fields is unclear. To address this the repeating pattern of Stripe (Sr) expression across the parasegment (PS) was examined in Drosophila. This pattern is generated in two steps. Initially, the ligands Hedgehog (Hh) and Wingless (Wg) subdivide the PS into smaller territories. Next, the ligands Hh, Spitz (Spi), and Wg each emanate from a specific territory and induce Sr expression in an adjacent territory. The width of Sr expression is determined by signaling strength. Finally, an enhancer trap in the sr gene detects the response to Spi and Wg, but not to Hh, implying the existence of separable control elements in the sr gene. Thus, a distinct inductive event is used to initiate each element of the repeating striped pattern (Hatini, 2001).

The repeating pattern of Stripe (Sr) expression across the parasegment (PS) is generated by inductive inputs from three spatially localized ligand sources. The ligands, Hh, Spi, and Wg, emitted by En, Ve, and Wg territories, respectively, control Sr expression in cells adjacent to each ligand source. There are three notable features to this regulation: (1) each ligand-producing territory induces Sr expression in the adjacent territory; (2) the induction is asymmetric, either anterior or posterior to each source; (3) the induction is initiated at the high level of signaling achieved near the source, limiting expression of Sr to a narrow row of cells. Because these same ligands act more broadly in cuticle cell fate specification, these results also suggest that the ligands and signaling territories operate in a fundamentally distinct way in order to construct a repeating striped pattern. These observations reveal a strategy used to generate a repeating striped pattern across a cellular field that may be used generally (Hatini, 2001).

Each Sr row is initiated adjacent to a different ligand source. The induction of Sr was limited to a narrow row of cells at each position. Manipulating either the ligand level, or the sensitivity of cells to a specific signaling pathway, leads to a broadened territory of Sr induction. Thus, local activity gradients of Hh, Spi, and Wg are each generated, and a threshold for activation of Sr is only surpassed in cells adjacent to each source. The gradient landscape of Spi and Wg is sculpted using the inducible antagonists Argos and Naked, respectively. Although how the activity landscape for Hh is sculpted was not specifically addressed, the Hh pathway also makes use of an inducible antagonist. It is likely that Hh spread is limited by binding to the Hh receptor Ptc, which is upregulated by Hh input (Hatini, 2001).

To generate the repeating striped tendon pattern, the Sr gene must be able to respond to each of three different ligands. To account for this, it is expected that the Sr promoter is modular, and each Sr row is induced via a separable, cis-acting response element. An enhancer trap P-insertion in the sr gene (sr03999) provides evidence for this since it detects the response to Spi and Wg, but not to Hh, implying that the P-insertion separates response elements in the sr gene. A separable Sr promoter element controlling Hh-dependent expression has been identified. Although this element operates only in dorsal and lateral epidermis, and not ventrally where Sr is expressed in repeating striped pattern, this observation strongly suggests that the control elements will be modular. Furthermore, in this dorsal/lateral element, the presence of functional, consensus Cubitus interruptus (Ci) DNA binding sites suggests direct regulation of Sr by the Hh signaling pathway. The obvious analogy is to the modularity of regulatory regions of certain pair-rule genes, which are able to integrate non-periodic information in order to generate periodicity. Note that the induction of the sr gene is limited to cells bordering each ligand source, even though each of the signals can act across several cell diameters. It is predicted that a given Sr response-element is configured to sense and respond only to a particularly high threshold level of each ligand (Hatini, 2001).

The ligands controlling Sr expression emanate from each of three territories across the PS. These territories are established by the primary organizing signals, Wg and Hh. In the earliest step, cross regulation between Wg and En/Hh-expressing cells stabilizes each ligand's expression and consolidates these two territories. In addition, through negative regulation, both Wg and Hh limit the expression of Ser to a central territory within the PS. Finally, signals from the En/Hh territory induce Ve expression in two cell rows just posterior to the En/Hh territory. The exact width of the Ve-expressing territory is adjusted as local input from the Ser territory induces a third Ve-expressing cell row. Thus, Hh and Wg act indirectly by defining and limiting each other's expression territory, as well as that of downstream ligands. All of these ligands then organize the repeating pattern. Three ligands induce Sr expression at specific positions across the PS, while the role of Ser reveals a particularly interesting spatial cue. Although the second row of Sr is induced in the anterior-most row of Ser-expressing cells, Ser expression is not necessary for this. Rather, Ser dictates the spacing between Sr row 1 and 2, because it defines the breadth of the Ve territory and thereby the position of the first non-Ve cell that can induce Sr in response to Spi-Egfr signaling (Hatini, 2001).

Sr expression is induced asymmetrically relative to each ligand source. For instance, Hh induces Sr posterior to the En territory, but not in the En territory or anterior to it. Wg imparts asymmetry to Hh/En signaling, and thereby prevents Ve expression anterior to the En/Hh territory. In exactly the same way, via antagonism of Hh signaling, Wg appears to block Sr expression anterior to the En/Hh territory, because the removal of Wg function allows Sr expression anterior to the En/Hh territory. The Wg signaling pathway imposes asymmetry to the dorsal/lateral Sr regulatory element via consensus Pangolin DNA binding sites. This principle is likely to extend to the ventral control of Sr for the generation of one element of the repeating striped pattern (Hatini, 2001).

One reason why Hh signaling cannot induce Sr expression in the En cells is that En represses expression of the Hh signal transducer, Ci. Nevertheless, it is still necessary to explain why signals from both the Ve and Wg territories cannot induce Sr in the En cells, even though each signal definitely acts on these cells to specify cuticle fate. A clue comes from the observation that when the En territory is not maintained Sr is induced symmetrically relative to the Wg or to the Ve sources. Thus, it is proposed that the En protein prevents Sr induction by Wg or Egfr inputs by repressing Sr expression. This is supported by the observation that activating Wg signaling at high levels in En cells still does not lead to Sr expression. To explain why Sr is not induced by Spi in the Ve territory, nor by Wg in the Wg territory, it is inferred that there is a specific block to autocrine signaling in each territory. Interestingly, this block is specific to Sr induction, and not to other outcomes of signaling, such as cuticle fate specification. It suggests a lack of an activator essential to induce Sr expression or expression of a repressor that blocks such a response in the Ve and Wg territories (Hatini, 2001).

The same ligands establish strikingly distinct patterns across the same cellular field. While the cuticle pattern comprises a diversity of cell types, the Sr expression pattern reflects the near-periodic specification of the same cell type. These distinct outcomes arise because the same ligands act in a fundamentally different manner in these two processes. As an example, Wg specifies smooth cuticle in a broad region anterior to the Wg territory, in the Wg territory, and posterior to the Wg territory (in anterior En/Hh cells). However, as is shown in this study, Wg induces Sr only anterior to the Wg territory in a narrow region, and not in the Wg territory or posterior to it. Also, Spi, through Egfr function, induces denticles over a broad region, both in the Ve territory and anterior to it in a subset of En/Hh cells. However, Egfr function induces Sr only in a narrow region posterior to the Ve territory. Thus, despite the broad effects of Wg and Egfr on cuticle pattern, the effect of Wg and Egfr in building the repeating striped pattern is constrained to a narrow region of cells. As a final example of the distinction between control of denticle pattern and control of repeating striped pattern, in the same row of cells, cuticle fate is specified by Spi while tendon fate is specified by Hh input. The unique effects of these ligands on Sr expression are crucial for the establishment of the repeating striped pattern. Thus, the information encoded in the signaling territories is decoded in different ways to achieve both repeating pattern and cell-type diversity across the same field (Hatini, 2001).

The generation of near-periodic Sr pattern across the PS is conceptually similar to the two-step process that is used in establishing the periodic body plan through pair-rule gene expression in syncitial embryos. Initially, primary pattern organizing centers are established at the boundaries of a field of naive nuclei or cells. In the first step, these centers establish patterned expression of secondary organizing genes across the field, subdividing the field into distinct gene expression territories. In the second step, the information encoded in these territories is used to initiate a repeating striped gene expression pattern. In the embryo, Bicoid together with Hunchback and Nanos organize expression of the gap genes. Territories of gap gene expression are then used to establish the periodic pattern of primary pair-rule gene expression. In the PS, Wg and Hh organize overall parasegmental pattern by first defining each other's territory and then the territories of secondary regulatory genes, Ser and Ve. The signaling territories are then used to establish near-periodic expression of Sr. Note that although the conceptual similarity is striking, the mechanisms generating these two repeating patterns are distinct. The pair-rule gene expression pattern is established in a unique, syncitial system by diffusion of transcriptional regulators in a common cytoplasm, whereas Sr expression is established across an epithelial monolayer by communication between cells via inter-cellular signaling systems. In addition, while a balance between diffusible activators and repressors determines pair-rule gene expression at any point along the syncitial embryo, the unique properties of the signaling territories across the PS determine Sr expression. In particular, the juxtaposition of pairs of territories, one that sends a signal with one that can initiate Sr expression in response to the signal, is utilized to initiate Sr expression adjacent to the boundaries between these territories (Hatini, 2001).

A near-periodic striped pattern of veins is produced in the developing wing disc. Emerging evidence suggests that the two-step strategy may also apply to this system, and that unique properties of putative signaling territories are used to initiate wing veins adjacent to at least two territories across the wing blade. In the first step, combined action of Hh and Dpp establishes different territories of downstream regulatory genes across the A/P axis of the future wing blade. While Hh establishes a territory of Ptc expression adjacent to the compartment boundary, Dpp acts more broadly across the wing and establishes two nested territories of Spalt and Optomotor-blind expression. In the second step, veins are induced adjacent to at least two territories, suggesting that an unknown ligand emanates from one territory and induces the vein in the adjacent territory. The possibility that the two-step process described here for the PS is used across the wing blade suggests that this may be a general strategy for creating repeating striped patterns across other cellular fields (Hatini, 2001).

In Drosophila, muscles attach to epidermal tendon cells are specified by the gene stripe (sr). Flight muscle attachment sites are prefigured on the wing imaginal disc by sr expression in discrete domains. The mechanisms underlying the specification of these domains of sr expression have been examined. The concerted activities of the wingless (wg), decapentaplegic (dpp) and Notch (N) signaling pathways, and the prepattern genes pannier (pnr) and u-shaped (ush) establish domains of sr expression. N is required for initiation of sr expression. pnr is a positive regulator of sr, and is inhibited by ush in this function. The Wg signal differentially influences the formation of different sr domains. These results identify the multiple regulatory elements involved in the positioning of Drosophila flight muscle attachment sites (Ghazi, 2003).

Much of the presumptive notum is in the anterior compartment in which there are four domains of sr expression. One of these is in the medial region (a) and gives rise to the anterior tendon cells, to which dorsal longitudinal muscles (DLMs) attach. The remaining three are in the lateral region (b, c and d) and provide dorsal attachments for dorsoventral muscles (DVMs). In the posterior compartment, sr is expressed in a narrow band that eventually forms the posterior insertion site for DLMs. The positioning of different sr domains on the wing imaginal disc have been examined with respect to prepattern and pattern forming genes expressed in this region (Ghazi, 2003).

The Wg gradient in the presumptive notum controls sr transcription differentially and keeps different sr domains distinct. The actual regulation of sr by wg appears to be very complex. Lateral domain c appears more sensitive to perturbations in wg signaling as compared to b. This is interesting since all the cells of domain b receive uniform levels of Wg whereas cells at the posterior border of c, and those bordering the posterior sr domain receive high Wg. One possibility is that the latter cells block progress of the Wg gradient and thus determine responses of cells further away. This may be brought about by targeting Wg to lysosomes and degrading it, as in the embryo. Another possibility, not exclusive of the first, could be that the domain and levels of wg transcription determine the range and gradient of Wg. The precise definition of the domain of wg transcription could be by a mechanism similar to that used in the wing margin. While the data suggest that the posterior and lateral-most domain do not receive Wg and may lie outside its purview, the formal possibility still exists that wg affects these domains in some other unknown way (Ghazi, 2003).

wg controls sr expression by in segment border cells of the Drosophila embryo. Wg signaling restricts sr activation to a single row of cells. In the presumptive notum on the wing disc, hh expression is restricted to a very narrow region, which forms the posterior compartment. Its effects in the disc are mediated by dpp, which serves multiple functions. Dpp is required for induction of wg expression, as it positively regulates pnr, which in turn activates wg. However, once wg is induced, Dpp tightly restricts its domain. This antagonism is required for correct positioning of the DC bristles. This antagonism also defines domains of sr. It is unclear if dpp directly regulates sr, or its effect is by control of other genes. The similarity between sr phenotypes observed on expansion of wg expression, and in dpp mutants, is suggestive of its effects being mediated by wg only, but it is also possible that it influences sr expression directly (Ghazi, 2003).

Pnr, a GATA-binding protein normally functions as a transcriptional activator and is antagonized by Ush in its function. Loss of function pnr mutants show no sr expression in the domain covered by pnr. This, along with sr expansion in mutants of ush, would suggest that pnr activates sr in the notum, and is inhibited by ush. However, there is also loss of sr expression in pnr `gain of function' mutants. The reason for this is not completely clear. One possibility is that since the mutation causes an increase in wg activity in the region this may cause a down-regulation of sr. This is supported by a similar effect seen on misexpression of activated armadillo in the pnr domain. Results with both pnr and ush have been taken into account to suggest that pnr positively regulates sr and is antagonized by ush (Ghazi, 2003).

Most sr expression commences in regions of low ush. Phenotypes of ush mutants, and ush misexpression experiments, also indicate that the gene inhibits sr, in keeping with the simplistic scenario that ush antagonizes pnr-mediated activation of sr. However, the medial sr domain is partially covered by ush proximally. Further, pnr is known to be required for positive induction of ush in the embryonic epidermis and in some loss of function allelic combinations of pnr, such as pnrVX6/pnrVX1, there is reduced ush expression on the disc. So how is sr initiated in the region where Pnr and Ush are co-expressed? The answer to this is not known but probably lies in levels of Ush and Pnr at that position. In loss of function ush mutants ectopic dorsocentral bristles form but post vertical (PV) bristles are missing, suggesting that the Pnr-Ush complex acts as a repressor of the DC enhancer, but as activator of the enhancer of PV bristles. Such observations have indicated complex, context dependent interactions between Pnr and Ush in determining cell fate and could explain the regulation of sr expression in the medial notal region (Ghazi, 2003).

These results indicate that each sr domain is regulated by a combination of prepattern genes and signaling molecules. But, a precise description of the 'combinatorial code' for regulation of each sr domain is beyond the scope of this work and can be achieved by generation of domain specific markers for sr. Based on expression pattern data, and existing literature, it is suggested that high levels of Pnr, low (or absence of) Ush and moderate levels of Wg determine the initial induction of domain a. The distinction between medial (a) and lateral (b-d) domains is established by the presence of very high levels of Wg (the cells where the Wg gradient originates). Lateral expression domains are probably induced in domains controlled by the lateral prepattern gene iro. The differences between different lateral domains arise as a result of expression of different genes in the region. For instance, the lateral-most domain d appears to be regulated by ush and does not encounter Wg at all. Whereas, all cells of b receive uniformly moderate levels of Wg, only cells at the borders of c receive high Wg levels, and these differences result in the distinct identities of the two domains. Dpp, either through its effects on these regulatory genes and/or through direct effects on sr influences the process (Ghazi, 2003).

The role of N as a potential regulator of sr was examined, since it is known to influence multiple events in wing disc morphogenesis from proliferation to bristle patterning. Using a temperature sensitive allele (Nts), the protein function was inactivated by growing animals at non-permissive temperatures during the third larval instar. sr expression was examined at 0 h APF. Loss of sr expression is observed in these animals. In hemizygous males, this effect is most severe and sr expression is completely abolished. Females, with one normal copy of N, showed faint sr expression. This suggested that N may be required for initiation of sr expression. A dominant negative form of N (Ndn), was expressed in the pnr domain; abolition of sr expression was found. This was observed most clearly in the anterior medial domain covered by pnr. The lateral domains showed some reduction in sr as well. In a gain of function experiment, a constitutively active form of N (Nintra) was expressed in the same region and results in an increase in sr-lacZ. The N ligand Ser is known to regulate sr expression in the embryonic segment border cells. Mis-expression of Ser in the presumptive notum region resulted in loss of sr expression. Together, these results show that the initiation of sr expression relies on N, which is antagonized by Ser in this activity (Ghazi, 2003).

Prepatterning the Drosophila notum: the three genes of the iroquois complex play intrinsically distinct roles -- Regulation of stripe expression

The Drosophila thorax exhibits 11 pairs of large sensory organs (macrochaetes) identified by their unique position. Remarkably precise, this pattern provides an excellent model system to study the genetic basis of pattern formation. In imaginal wing discs, the achaete-scute proneural genes are expressed in clusters of cells that prefigure the positions of each macrochaete. The activities of prepatterning genes provide positional cues controlling this expression pattern. The three homeobox genes clustered in the iroquois complex (araucan, caupolican and mirror) are such prepattern genes. mirror is generally characterized as performing functions predominantly different from the other iroquois genes. Conversely, araucan and caupolican are described in previous studies as performing redundant functions in most if not all processes in which they are involved. The question of the specific role of each iroquois gene in the prepattern of the notum was addressed, and it was clearly demonstrated that they are intrinsically different in their contribution to this process: caupolican and mirror, but not araucan, are required for the neural patterning of the lateral notum. However, when caupolican and/or mirror expression is reduced, araucan loss of function has an effect on thoracic bristles development. Moreover, the overexpression of araucan is able to rescue caupolican loss of function. It is concluded that, although retaining some common functionalities, the Drosophila iroquois genes are in the process of diversification. In addition, caupolican and mirror are required for stripe expression and, therefore, to specify the muscular attachment sites prepattern. Thus, caupolican and mirror may act as common prepattern genes for all structures in the lateral notum (Ikmi, 2008).

In earlier works, the iro-C genes functions were mainly assessed from studies of mutations affecting solely mirr, of deficiencies deleting the whole complex or of rearrangements susceptible to affect several genes, and on misexpression experiments. In order to unravel the respective roles of ara, caup and mirr and to analyze their possible functional redundancy in the neural prepatterning of the notum, this study has combined the analysis of LoF mutations of these genes with a functional replacement approach (Ikmi, 2008).

mirr appears to be required for the formation of four out of the seven lateral macrochaetes (PS, pSA, aPA and pPA): their loss is elicited by LoF alleles of mirr and by a dominant allele (mirrG8. The proneural clusters as well as the SOPs corresponding to the macrochaetes unaffected by mirr mutations may reside outside (aNP) or inside (pNP) the mirr domain of expression. Therefore the requirement or dispensability of mirr for the patterning of the bristles in the lateral notum appear to depend partly but not only on its domain of expression. All the phenotypes observed in this study were elicited by mild perturbations of the mirr function, compatible with viability of adults. Therefore, a role of mirr in the patterning of the other lateral macrochaetes cannot be completely excluded (Ikmi, 2008).

In the prospective lateral notum, ara and caup appear to have mostly identical domains of expression, which, at least, enclose all the SOPs for lateral macrochaetes. These authors have suggested that there is a functional redundancy of these two genes on the grounds of their similar protein-coding sequences and patterns of expression, and this is generally admitted. However no evidence of this has been reported to date. Notably no LoF mutations affecting only one of these two genes have been described previously. This study characterized mutations abolishing (or drastically reducing) caup expression without an appreciable effect on ara and mirr (e.g. iro1 or caupG3) and null (or strong hypomorph) mutations of ara that do not affect significantly caup or mirr (e.g. TSI or araG4) (Ikmi, 2008).

Six transposable element (TE) insertions were characterized in caup, and it was shown that they behave like an allelic serie of caup hypomorphic mutations: flies homozygous for these insertions are viable and display phenotypes ranging from wt or the loss of only a few macrochaetes to the loss or a strong frequency decrease of all the seven lateral macrochaetes (strong Iroquois phenotype). These phenotypes are aggravated when these TE insertions are heterozygous over an iro-C deficiency and even more over the iro1 rearrangement, which has a breakpoint in the first caup intron. Heterozygous iro1/Df-BSI L3 larvae display no caup expression (or an extremely low level) although ara and mirr levels are similar to the control conditions. Similarly, the strongest P allele of caup (caupG3) causes a drastic reduction or an absence of caup expression without affecting notably ara and mirr, neither in expression level nor in expression domain in the notum. Therefore, it can be assumed that caup is required for the patterning of all lateral macrochaetes and, presumably, for the direct or indirect activation of sc expression in the lateral notum (Ikmi, 2008).

Besides inactivating caup expression, the P[Gal4] insertions in the caup gene reproduce its expression pattern. By combining such a Gal4 driver with an UAS-caup transgene, it was shown that caup overexpression in its normal domain of expression is able to rescue the caup LoF phenotype to an almost wt situation. The frequency of occurrence of all the seven lateral macrochaetes is either normal or at least elevated as compared to the mutant situation. Taken together, loss and gain of function approaches demonstrate that caup is required for the patterning of all the lateral macrochaetes of the notum (Ikmi, 2008).

The same approach was applied to study the loss of function of ara, and the effects were analyzed of five insertions and an inversion (TSI) in the ara gene. In araG4/TSI and TSI/Df-BSI L3 larvae, ara expression is absent or strongly reduced while caup and mirr expression are not significantly affected as compared to the control situations. These larvae die as pupae, showing the requirement of ara for viability. The very rare adult escapers do not present any bristle defects. The same is true for the four other ara hypomorphic mutants studied. In summary, the sole LoF of ara is never associated with a loss of macrochaetes. Consequently, conversely to caup, ara is dispensable for the patterning of all the lateral bristles of the notum (Ikmi, 2008).

In conclusion, the Iro proteins are different for their contribution to the neural patterning of the lateral notum. Only Caup and Mirr are required for the patterning of the lateral macrochaetes and therefore to control the activation of expression of ac and sc in the corresponding proneural clusters (Ikmi, 2008).

When it is strongly overexpressed in the caup domain of expression, Ara, although not required, is sufficient to rescue the lack of lateral macrochaetes phenotype elicited by caup LoF. Therefore, unexpectedly considering the LoF results, when overexpressed Ara can carry out the function of Caup in the neural patterning of the notum. This property may play a biological role, as can be seen in the exacerbation of the phenotypic defects elicited by ara LoF in sensitized conditions (reduction of caup and/or mirr function). Although to a much less extent, Mirr is also able to perform in part the Caup function, as seen from the limited rescue of PS, aNP and aSA macrochaetae by mirr overexpression in a caup LoF background. This functional difference between Ara and Mirr correlates well to their sequence divergence with Caup. A possible explanation for the difference observed between the LoF of ara and caup is that Ara, in physiological conditions, has a lower activity than Caup to promote proneural genes expression. The increase of the amount of the Ara protein to non-physiological levels probably increases the expression of proneural genes sufficiently to compensate for the loss of Caup activity. This is in good agreement with the findings that Drosophila Iro proteins bind in vitro the same sequence but differ in their strength of binding to this site. Consequently, owing essentially to the proteins intrinsic properties, the Iro transcription factors do not regulate in vivo the same target genes (Ikmi, 2008).

It has been suggested that a common network of prepattern genes regulates all structures of the notum. Indeed, in the prospective medial notum, pnr is required for both ac-sc and sr expression, which respectively specify the development of bristles and tendon precursors. Preliminary observations indicate that pnr also regulates pigment patterns in medial notum (Ikmi, 2008).

In the wing discs of larvae mutant for caup or for mirr, two (b and d) out of the three domains of sr expression located in the prospective lateral notum are missing. Furthermore, a mirr mutant partially affects sr expression in the two remaining domains (a and c). In addition, adult flies lacking either caup or mirr display falling wings, suggesting an abnormal attachment of indirect flight muscles. Therefore, caup and mirr activate sr expression and in consequence participate to the specification of lateral flight muscle attachment sites. It had been shown that Iro proteins can form homodimers and heterodimers that bind in vitro the same palindromic site called Iro Binding Site (IBS; Bilioni, 2005). Thus, it is possible that the activity of a Caup-Mirr heterodimer controls the regulation of sr expression in regions (b) and (d). The sr expression pattern is normal in the wing discs of larvae lacking the ara function. Hence ara is not required for the expression of sr in the prospective notum. Nonetheless, a reduction of ara function, as seen in hypomorphic mutants or escapers to the lethality, leads to heldout wings. This phenotype is often observed in flies carrying mutations that affect direct flight muscles. It is thus possible that ara regulates the specification of direct flight muscle attachment sites. In summary, here again, the iro-C genes products are not functionally equivalent in their contribution to the muscular attachment sites prepattern (Ikmi, 2008).

Although there are no strong evidences that iro-C regulates pigment patterns, preliminary reports suggest that this is the case. From all these data, the iro-C genes caup and mirr appear as common prepattern genes for the specification of all the structures in the lateral notum, similarly to pnr in the medial notum (Ikmi, 2008).

In addition, pnr and iro-C domains partially overlap each other at the virtual border between the medial and the lateral notum. The DC macrochaetes and the 'a' sr domain can be affected in caup and mirr mutants. These structures are dependant on the pnr function. Therefore, both pnr and the iro-C appear to prepattern a region of the notum, at the intersection of their expression domains, which could correspond to the medial-lateral band of wg expression overlapping these domains (Ikmi, 2008).

mirr LoF leads to embryonic lethality. Although the targets of ara are yet to be identified, ara LoF causes pupal lethality, whereas flies lacking caup function are viable and exhibit developmental defects. Thus, ara, caup and mirr play distinct biological functions during development. These different roles can be attributed in part to differences in expression pattern. For instance, mirr is the first iro-C gene that is detected during embryogenesis and it is the only iro-C gene that is expressed and plays a role in oogenesis. However, this diversity in expression domains cannot account for all the observed functional differences. The overexpression of these three proteins in the same domains (here with the same caup-Gal4 driver) have different consequences: while overexpression of caup and ara to more than 10 times the normal level is compatible with viability, the overexpression of mirr to the same level is lethal. When expressed at levels compatible with viability, mirr is unable to rescue the caup LoF while ara is. Thus, the differences in the iro-C genes roles cannot be only attributed to differences in their patterns of expression but rather should also be due to differences in their coding sequences (Ikmi, 2008).

In mammals, there are two clusters of three iroquois genes (IrxA and IrxB). Expression patterns, LoF and misexpression studies reveal a similar situation where the Irx genes may have redundant or non-redundant roles, depending on the gene and/or on the developmental process (Ikmi, 2008).

The roles of the Drosophila iro-C genes have been documented in numerous other developmental processes, for instance: eye dorsal-ventral patterning, wing veins patterning, wing-body wall boundary, dorsal-ventral axis formation in oogenesis. This study has put together the tools and conditions allowing to address the question of the specific roles and/or of the functional redundancy of the iro-C gene in these other developmental processes (Ikmi, 2008).

The long-term fates for duplicated genes include inactivation, maintenance and diversification. Maintenance of duplicated developmental genes can be the result of selective pressure exerted through dosage requirements and/or contribution to the genetic and developmental robustness necessary to reproducibly elaborate correct patterning of diverse territories. Alternatively, the subfunctionalization model proposes that, after a duplication of genes, each copy may sustain deleterious mutations in different structural and/or regulatory elements. Eventually, the ancestral functions are partitioned between the copies that are both retained (Ikmi, 2008).

An interesting example is the ac-sc complex. Similarly to ara and caup, ac and sc arose from the most recent duplication event in this complex. It has been shown that ac, but not sc, is dispensable for the development of the sensory bristles. Two hypotheses have been proposed for the reason why evolution has retained ac: first, an as yet undiscovered function; second (the favored hypothesis), a contribution to genetic robustness. However the situations of the ac-sc and iro complexes are not identical: no phenotypic consequence of the ac LoF in has been found in an otherwise wt background while this study observed ara LoF cause lethality. The reasons why the three Drosophila iro-C genes are maintained appear thus different and may proceed from the two (non-exclusive) hypotheses mentioned above. mirror has clearly evolved to perform different functions from ara and caup. This can be seen in expression pattern as well as in LoF phenotypes and in the functional properties of the protein. This study has provided evidences for the functional divergence between ara and caup: first, the LoF of ara is lethal while the LoF of caup is not; second, caup is required for the neural and the muscular prepatterning of the prospective notum, while ara is dispensable. Additionally, subtle differences exist in their expression pattern in the L3 wing disc. Other differences have yet to be characterized more precisely at other stages and in other tissues and their role investigated. The three iro-C genes appear in the process of diversification and subfunctionalization, both in their expression domains and in the functions of their encoded proteins. In addition, the partial ability of ara to perform caup functions, at least in the neural patterning of the notum, may contribute to buffer this patterning against intrinsic and extrinsic perturbations. It could thus be retained by a selective pressure at work on fly populations surviving in the wild (Ikmi, 2008).

Targets of Activity

The number of epidermal cells expressing the muscle attahment markers groovin, delilah and ß1-tubulin is greatly reduced and their normally stereotyped expression patterns are strongly disturbed in sr mutants (Frommer, 1996).

Thrombospondin, acting downstream of stripe, mediates in the formation of the myotendinous junction

Organogenesis of the somatic musculature in Drosophila is directed by the precise adhesion between migrating myotubes and their corresponding ectodermally derived tendon cells. Whereas the PS integrins mediate the adhesion between these two cell types, their extracellular matrix (ECM) ligands have been only partially characterized. This study shows that the ECM protein Thrombospondin (Tsp), produced by tendon cells, is essential for the formation of the integrin-mediated myotendinous junction. Tsp expression is induced by the tendon-specific transcription factor Stripe, and accumulates at the myotendinous junction following the association between the muscle and the tendon cell. In tsp mutant embryos, migrating somatic muscles fail to attach to tendon cells and often form hemiadherens junctions with their neighboring muscle cells, resulting in nonfunctional somatic musculature. Talin accumulation at the cytoplasmic faces of the muscles and tendons is greatly reduced, implicating Tsp as a potential integrin ligand. Consistently, purified Tsp C-terminal domain polypeptide mediates spreading of PS2 integrin-expressing S2 cells in a KGD- and PS2-integrin-dependent manner. A model is proposed in which the myotendinous junction is formed by the specific association of Tsp with multiple muscle-specific PS2 integrin receptors and a subsequent consolidation of the junction by enhanced tendon-specific production of Tsp secreted into the junctional space (Subramanian, 2007).

Tsp was initially recovered in a microarray screen for genes that are downstream of Stripe by comparing the gene expression profile of embryos overexpressing Stripe in the ectoderm with that of wild-type embryos. Further analysis showed that in stripe mutant embryos Tsp protein is still detected, possibly because of earlier Stripe-independent transcriptional input. Importantly, overexpression of Stripe using the engrailed-gal4 driver leads to a significant induction of Tsp expression, confirming the microarray results and the ability of Stripe to induce Tsp expression. Because Stripe expression is greatly upregulated following muscle-tendon interaction, it is assumed that Stripe-dependent Tsp induction is linked to muscle-tendon interaction. It is concluded that Tsp distribution is dynamic and correlates with the biogenesis of adherens junction formation (Subramanian, 2007).

Drosophila tsp is expressed in all ectodermal tendon precursor cells, strongly enriched in those positioned at the segment border of the embryo. Furthermore, tsp is expressed in all differentiated tendon cells after muscle contact. Therefore, tsp is expressed in all cells that have previously been identified by the expression of stripe. stripe encodes an EGR-type Zn-finger transcription factor that is required for tendon cell differentiation. Like stripe, the initial expression of tsp is controlled by Hedgehog signaling at the segment borders and requires stripe activity only during the later stages when the tendon cells are already differentiated. These results suggest that the genes stripe and tsp are activated in parallel by Hh-dependent Ci activity, and that stripe activity maintains the expression of tsp during the later stages when Ci activity has ceased (Chanana, 2007).

Post-transcriptional Regulation

Differential RNA metabolism regulates a wide array of developmental processes. A mechanism is described that controls the transition from premature Drosophila tendon precursors into mature muscle-bound tendon cells. This mechanism is based on the opposing activities of two isoforms of the RNA binding protein How. While the isoform How(L) is a negative regulator of Stripe, the key modulator of tendon cell differentiation, How(S) isoform elevates Stripe levels, thereby releasing the differentiation arrest induced by How(L). The opposing activities of the How isoforms are manifested by differential rates of mRNA degradation of the target Stripe mRNA. This mechanism is conserved, as the mammalian RNA binding Quaking proteins may similarly affect the levels of Krox20, a regulator of Schwann cell maturation (Nabel-Rosen, 2002).

RNA binding proteins of the Signal Transduction and Activation of RNA (STAR) family may regulate gene expression at various levels, e.g., at the level of nuclear export of the target mRNA, at the level of mRNA stability, and at the translation level. The How proteins appear to exert their activity through their effect on mRNA stability. How(L) appears to induce rapid degradation of the target RNA, an activity that is tightly coupled to its nuclear retention and depends on the presence of the nuclear retention signal that is conserved in QKI-5. It has been suggested that How(L) may prevent nuclear export of its target mRNA. Indeed, in embryos overexpressing How(L), the mRNA of gfp-sr3'UTR is occasionally detected in the nucleus. Presently, it is not possible to determine whether the primary effect of How(L) is retention of the target mRNA in the nucleus followed by degradation of the target mRNA or vice versa. How(S) increases the stability of the same target RNA. The fact that How(S) is present both in the nucleus and in the cytoplasm raises the possibility that the association of How(S) with its target RNA during and following its nuclear export leads to mRNA stabilization. A number of RNA binding proteins possess both nuclear and cytoplasmic functions, e.g., proteins that affect both mRNA export and mRNA stability. Similarly, How proteins may affect both nuclear-cytoplasm shuttling and mRNA stability. The differential association of each of the How proteins with distinct protein partners presumably leads to their opposing effects on mRNA stability. A possible mechanism for the counteraction effect of How(S) may arise from its association with How(L), eliminating the repression by How(L). Indeed, How(S) and How(L) are coprecipitated from Schneider cells following their cotransfection together with the gfp-sr3'UTR (Nabel-Rosen, 2002).

A recent report suggests that a sequence (TGE) in the 3'UTR of C. elegans tra-2, essential for Gld-1 binding, mediates deadenylation and poly(A)-dependent translation repression. Poly(A) deadenylation may also lead to mRNA degradation. Since a sequence motif that is partially related to TGE is also present in the stripe and krox20 3'UTR, degradation of the target mRNA by How(L) may be based on a similar mechanism. Recently, the two cytoplasmic hnRNPs, K and E1, have been shown to inhibit translation of lipoxigenase mRNA by preventing its attachment to the 60S ribosomal unit (Ostareck, 2001). The possibility cannot be excluded that How(S), in addition to its positive effect on mRNA stability, may also facilitate translation efficacy (Nabel-Rosen, 2002).

The results suggest that the relative amount of How(L) and How(S) during different stages of embryonic or adult development regulate the switch between the premature and mature state of differentiation of tendon cells. In early stages of embryonic development, How(L) prevails, Stripe expression is downregulated, and differentiation is arrested. In later stages of embryonic development, How(S) is upregulated, overriding How(L) repression and facilitating Stripe expression. The difference in Stripe mRNA levels may be further enhanced by Stripe transcriptional autoregulation. What could be the mechanism that regulates the relative levels of How proteins during tendon cell maturation? Northern analysis suggests that the total levels of How(L) mRNA are low throughout embryonic development, relative to those of How(S) mRNA. At the protein level, the proportion of the two proteins is inverted; How(S) protein levels are low and increase only during late stages of embryonic development. This suggests that How(S) is posttranscriptionally regulated. Indeed, transgenic flies carrying How(S) with its unique 3'UTR exhibit almost undetectable levels of How(S) protein following induction by the gal4 driver. When this 3'UTR is deleted, the expression levels of How(S) are significantly higher. The expression of How(S) appears to be elevated by Vein-mediated activation of the ->F receptor pathway in tendon cells following muscle-tendon association. The molecular link between ->F receptor activation and the upregulation of How(S) has yet to be elucidated. A recent report suggesting that ERK phosphorylation of hnRNP-K drives cytoplasmic accumulation of hnRNP-K may be of relevance if, similarly to hnRNP-K, How(S) undergoes ERK-dependent phosphorylation (Nabel-Rosen, 2002).

Muscle-dependent maturation of tendon cells is induced by post-transcriptional regulation of stripeA

Terminal differentiation of single cells selected from a group of equivalent precursors may be random, or may be regulated by external signals. In the Drosophila embryo, maturation of a single tendon cell from a field of competent precursors is triggered by muscle-dependent signaling. The transcription factor Stripe induces both the precursor cell phenotype, as well as the terminal differentiation of muscle-bound tendons. The mechanism by which Stripe activates these distinct differentiation programs remained unclear. This study demonstrates that each differentiation state is associated with a distinct Stripe isoform and that the Stripe isoforms direct different transcriptional outputs. Importantly, the transition to the mature differentiation state is triggered post-transcriptionally by enhanced production of the stripeA splice variant, which is typical of the tendon mature state. This elevation is mediated by the RNA-binding protein How(S), with levels sensitive to muscle-dependent signals. In how mutant embryos the expression of StripeA is significantly reduced, while overexpression of How(S) enhances StripeA protein as well as mRNA levels in embryos. Analysis of the expression of a stripeA minigene in S-2 cells suggests that this elevation may be due to enhanced splicing of stripeA. Consistently, stripeA mRNA is specifically reduced in embryos mutant for the splicing factor Crooked neck (Crn), which physically interacts with How(S). Thus, a mechanism is generated by which tendon cell terminal differentiation is maintained and reinforced by the approaching muscle (Volohonsky, 2007).

This study demonstrates the involvement of post-transcriptional control in a cell-differentiation process that must be coupled to muscle-tendon interaction. Terminal differentiation of tendons involves a major reorganization of the microtubule and actin networks. Such processes are presumably not compatible with embryonic morphogenetic movements such as germ band retraction. Thus, it is essential to spatially and temporally restrict differentiation to single muscle-bound tendon cells. Indeed, the results show that premature overexpression of StripeA in the entire ectoderm leads to severe defects in germ band retraction (Volohonsky, 2007).

Stripe mediates both the determination of precursor cells as well as their maturation and ability to undergo specific temporal and spatial regulation. The current findings suggest both negative and positive feedback loops, based on post-transcriptional regulation of stripe splice variants that on one hand maintain non-bound tendon cells at the precursor state, and on the other hand enable irreversible differentiation of muscle-bound tendons (Volohonsky, 2007).

Whereas some tissue differentiation processes (e.g. tracheal development) initiate upon the expression of a key transcription factor, which autoregulates its own expression, thus leading to a unidirectional differentiation route, other cells (e.g. cells in the proneural region) go through an intermediate stage of a field of competent precursors, in which only additional local interactions lead to irreversible differentiation. Maturation of tendon cells follows the latter path, although the selection mechanism is based on regulation at the post-transcription level (Volohonsky, 2007).

The following model is used to explain the transition between the two phases of tendon cell development: the initial expression of stripeB is induced by segment polarity-dependent signals. StripeB defines a set of tendon precursor cells. StripeB then reinforces its own expression and in addition induces How(L) expression, which in turn suppresses stripeB mRNA levels, thus keeping StripeB levels constant throughout embryonic development. This is supported by experiments that show that StripeB overexpression leads to elevation in How(L) and in StripeB itself. Following myotube extension and adhesion to a tendon precursor cell, How(S) levels are elevated in the muscle-adherent tendon cells, presumably due to EGFR activation. How(S) associates with the splicing factor Crn and the complex shuttles into the nucleus, where it binds to stripeA intronic sequences and elevates its mRNA levels, by enhancing its splicing and maintaining the stability of the spliced mRNA. The resulting muscle-bound tendon cell expresses high StripeA levels, which further drive the expression of genes required for terminal tendon differentiation (e.g. shot, how), as inferred from StripeA overexpression experiments. This regulatory mechanism couples muscle binding and tendon cell maturation, while preventing differentiation of additional, non-bound, precursors (Volohonsky, 2007).

RNA-binding proteins can function as adaptor units promoting the assembly of large protein complexes that control the various aspects of RNA metabolism. How, together with Quaking and GLD-1, belongs to the Star family of RNA-binding proteins, the members of which often regulate more than one facet of RNA metabolism. For example, GLD-1 has been suggested to regulate mRNA stability as well as translation of some of its targets. Similarly Quaking controls mRNA stability as well as RNA splicing, and possibly also mRNA nuclear export and localization. It appears that How proteins also exhibit a wide range of activities on RNA metabolism. While the effect of How(L) and How(S) on stripe mRNA stability has been demonstrated previously, this study suggests that How(S) has an additional activity in regulating the splicing of stripeA. Consistent with this study, How has been identified in a dsRNA-based screen for alternative splicing regulators, as a protein required for specific splicing of exons within two out of five tested genes, paralytic (exons A/I), and Dscam (exon 4), in S-2 cells. Previous studies suggested that the ability of How proteins to stabilize stripe mRNA is mediated by the 3' UTR of stripe. However, the splicing of stripeA appears to be regulated by its specific intronic sequences (Volohonsky, 2007).

By contrast to How(L), which is localized specifically in the nucleus, How(S) is distributed both in the nucleus and the cytoplasm. However, when How(S) is retained in the nucleus by the addition of an NLS sequence, it loses its effect on the mRNA levels of its target. What could be the molecular explanation for the involvement of How(S) in splicing? It is suggested that How(S) binds to a cytoplasmic splicing factor and recruits it to the nucleus, where it is targeted to bind stripeA-specific intronic sequences. This may enhance the splicing of stripeA-specific exons. A candidate splicing factor is Crn. Crn is a general, well-conserved splicing factor that is expressed by a wide range of cell types and is distributed both in the nucleus and the cytoplasm. In a parallel study, it was demonstrated that crn and how mutants exhibit closely related phenotypes, affecting glial cell maturation. Importantly, both Crn and How(S) proteins [but not How(S)-NLS] coprecipitate from S-2 cell extracts, indicating that both proteins are associated in a common protein complex in the cytoplasm (Edenfeld, 2006). In addition, when Crn is myristoylated and transfected into S-2 cells together with How(S), How(S) is relocated to the membrane (Edenfeld, 2006). Furthermore, in crn mutants StripeA, but not StripeB, levels are reduced, and this is reflected in the reduction of Shot levels (Volohonsky, 2007).

These results support a model in which How(S) interacts with Crn in the cell cytoplasm, shuttles into the nucleus and facilitates stripeA splicing, and possibly mRNA stability, leading to StripeA protein elevation. A similar mechanism may operate in the Quaking-dependent facilitation of myelin-associated glycoprotein splicing (Volohonsky, 2007).

In summary, a molecular mechanism has been described that is based on post-transcriptional control, by which cell differentiation is induced and maintained by local interactions with neighboring cells (Volohonsky, 2007).

Moleskin is essential for the formation of the myotendinous junction in Drosophila

It is the precise connectivity between skeletal muscles and their corresponding tendon cells to form a functional myotendinous junction (MTJ) that allows for the force generation required for muscle contraction and organismal movement. The Drosophila MTJ is composed of secreted extracellular matrix (ECM) proteins deposited between integrin-mediated hemi-adherens junctions on the surface of muscle and tendon cells. This paper identifies a novel, cytoplasmic role for the canonical nuclear import protein Moleskin (Msk) in Drosophila embryonic somatic muscle attachment. Msk protein is enriched at muscle attachment sites in late embryogenesis and msk mutant embryos exhibit a failure in muscle-tendon cell attachment. Although the muscle-tendon attachment sites are reduced in size, components of the integrin complexes and ECM proteins are properly localized in msk mutant embryos. However, msk mutants fail to localize phosphorylated focal adhesion kinase (pFAK) to the sites of muscle-tendon cell junctions. In addition, the tendon cell specific proteins Stripe (Sr) and activated mitogen-activated protein kinase (MAPK) are reduced in msk mutant embryos. Rescue experiments demonstrate that Msk is required in the muscle cell, but not in the tendon cells. Moreover, muscle attachment defects due to loss of Msk are rescued by an activated form of MAPK or the secreted epidermal growth factor receptor (Egfr) ligand Vein. Taken together, these findings provide strong evidence that Msk signals non-autonomously through the Vein-Egfr signaling pathway for late tendon cell late differentiation and/or maintenance (Liu, 2011).

In Drosophila, the formation of a stable myotendinous junction is essential to withstand the force of muscle contraction required for larval hatching. Proper formation of the MTJ requires proper integrin heterodimer formation at the junctions between both muscle cells and tendon cells for a permanent linkage to the ECM proteins deposited between these two cell types. The precise mechanism by which the muscle cells signal to the tendon cells to form and maintain the semi-adherens junctions that comprise the stable MTJ is still being elucidated. This study shows that the canonical nuclear import protein Msk is essential for Drosophila somatic muscle attachment. Moreover, a model is provided explaining how Msk may function in non-cell autonomously from the muscle to the tendon cell for proper MTJ maintenance. Though vein mRNA is produced in the myotubes, Vein protein is secreted and is restricted to the junctions at muscle-tendon attachment sites. Msk signals through the secreted Egfr ligand Vein to mediate cross-talk between the muscle and tendon cells. The binding of Vein to the tendon-expressed Egfr activates a signaling cascade through activated MAPK. Activated MAPK translocates to the nucleus and with SrA, activates downstream genes to induce terminal differentiation in the tendon cells. An inability of the tendon cells to maintain activated MAPK and Sr activity would affect the amounts of target proteins required to maintain stable muscle-tendon adhesion. For example, a decrease in Tsp deposition into the ECM would result in smaller attachment sites and an inability to maintain a tight integrin-ECM association (Liu, 2011).

Msk plays a general role in myogenesis as defects in msk mutant embryos were observed in all hemisegments and affected all muscle groups. The variable penetrance, which was classified as either major (Class I) or moderate (Class II) muscle detachment phenotypes, present in msk mutant embryos is likely due to the presence of maternal msk transcript. Attempts to further knockdown Msk levels by removal of maternal load resulted in non-viable egg chambers, consistent with a requirement for Msk function in cell viability. It is possible that a further decrease, but not complete loss in Msk function, could result in earlier defects in myogenesis, since muscle patterning defects, predominantly characterized by missing muscles, were observed in a subset of msk mutant embryos (Liu, 2011).

After cell fate determination is established in somatic muscle development, myoblast fusion and myotube migration begin to proceed simultaneously in stage 13 embryos until the final muscle pattern is completed. As the migrating myotubes approach their target tendon cells, the αPS2ΒPS integrin heterodimer begins to accumulate at the leading edge of the muscle. This integrin complex is required for at least two separate events: (1) to serve as a transmembrane link between the internal actin cytoskeleton and the ECM components Tsp and Tig; and (2) for the proper localization and/or accumulation of Vein. The accumulation of Vein at the sites of muscle-tendon interactions is necessary for activation of the Egfr pathway and subsequent late tendon cell differentiation. In these mature, muscle-linked tendon cells, SrB expression is positively regulated. SrB also turns on the downstream transcriptional target Tsp, resulting in more Tsp secretion and subsequent strengthening of the MTJ through integrin binding (Liu, 2011).

The results suggest that Msk affects the later stages of tendon cell maturation and MTJ formation and/or maintenance. (1) No obvious defects were observed in myoblast fusion or the guidance of muscles to their correct target tendon cell. (2) Msk protein expression, visualized by both antibody immunolocalization and a fluorescently-tagged Msk fusion protein, demonstrates that enrichment of Msk protein at the future muscle-tendon attachment sites occurs after stage 15. The appearance of Msk protein localization corresponds to the timing of MTJ junction formation, but it does not rule out the possibility that Msk has a role in earlier myogenic events. (3) The muscle detachment phenotypes observed in msk mutant embryos are consistent with the myospheroid phenotype observed for other genes, including myospheroid (βPS int), inflated (αPS2 int), and rhea (Talin), which are well-characterized for their role in embryonic muscle attachment. Finally, in all msk mutants examined, regardless of the severity of the muscle detachment phenotypes, MTJ formation occurred in the correct location. Although the muscle attachment sites were smaller in msk mutants than in WT embryos, they were initially formed correctly, but not capable of reaching their mature size. These data taken together indicate that the muscle-specific αPS2βPS integrin complex initially forms an attachment to the ECM proteins Tig and Tsp. However, as mature tendon cell induction is compromised in msk mutant embryos, whereby the tendon cells are not able to produce and secrete proper levels of Tsp protein. Thus, the size of the mature MTJ is reduced and results in a decreased affinity at the muscle-tendon junctions (Liu, 2011).

Sr is a key factor in tendon cell differentiation in the embryo and fly thorax. In the embryo, SrB is essential for early tendon cell induction and SrA is activated for later tendon cell maturation. Furthermore, ectopic expression of Sr can act as a guidance cue for migrating myotubes as ectopic expression of Sr in epidermal cells, the salivary glands, or the CNS results in muscle patterning defects where muscles take the incorrect route and/or become attached to ectopic cells expressing Sr. Even though the data shows that Msk is required for nuclear Sr in the tendon cells, ectopic expression of Msk is not sufficient for tendon cell induction based upon three lines of experimentation. First, ectopic Msk expression in either the muscle or epidermis resulted in aberrant muscle attachment, but not misguided myotubes. All muscles were found to be in the correct position, regardless of Msk expression in domains outside of the normal hemisegments. Second, ectopic Msk expression was not sufficient to induce either early or elevated Sr levels. Third, expression of Msk in the salivary glands or CNS did not result in misguided muscles toward these locations (Liu, 2011).

The tissue-specific rescue experiments show that Msk is required in the muscle cell for proper muscle-tendon attachment to occur. Thus, two mechanisms that are not mutually exclusive are proposed by which Msk may be functioning in the muscle cells to exert its effect on tendon cell maturation. First, Msk may be required directly and/or indirectly for the localization and/or accumulation of secreted Vein. As antibodies against Vein were not available for testing this possibility, it was shown that reintroducing Vein in msk mutants could rescue muscle attachment defects. Second, Msk may act via an integrin-dependent mechanism to modulate adhesion, which is explained in detail below (Liu, 2011).

From these studies, Msk localizes to the ends of muscles at the sites of muscle attachment. It is proposed that this localization of Msk recruits other proteins to the sites of muscle attachment to sequester proteins near the cell periphery and/or to modulate integrin affinity at the muscle attachment site. First, the absence of pFAK localization in msk mutants strongly suggests that Msk is essential for pFAK localization to the muscle-tendon attachment site. Surprisingly, mutations in FAK do not result in embryonic muscle attachment defects. However, pFAK localization is also lost in integrin mutants, suggesting that pFAK is involved in undefined events in myogenesis. If Msk serves as a scaffold protein to localize pFAK and/or other molecules to the sites of muscle-tendon cell attachment, these proteins may play an accessory role in integrin-mediated adhesion. One idea is that a Msk-pFAK complex may serve to limit the signaling function of integrins so its adhesive role predominates in MTJ formation. It is well-established that integrins play both adhesive and signaling roles in cell migration and development. Clustering of the cytoplasmic tails of βPS-integrins initiates a downstream signaling pathway that regulates gene expression in the Drosophila midgut, but is not sufficient to induce tendon cell differentiation in formation of the MTJ. This suggests that integrin-mediated adhesion is required to assemble ECM components and influence the ability of Vein to activate the Egfr pathway. While loss of FAK activity does not result in somatic muscle defects, overexpression of FAK does. Muscles that have detached from the epidermis as a result of FAK overexpression still retain βPS2 integrins at the muscle ends. As observed in mammalian systems, this raises the possibility that pFAK may play a role in integrin complex disassembly. Excess pFAK may either displace proteins that bind to the cytoplasmic domain of integrins or excessively phosphorylate proteins resulting in integrin complex turnover and a decrease in stable adhesion. Alternatively, pFAK may exhibit redundancy with another protein at the attachment sites. There is precedence for this in the fly as pFAK functions redundantly with the tyrosine kinase Src downstream of integrins in the larval neuromuscular junctions (NMJs) to restrict NMJ growth. Furthermore, in FAK mutants, phosphotyrosine signal is still observed at the sites of muscle attachment, supporting the idea that another tyrosine kinase is functional. A viable candidate may be Src42A, as it is also expressed at the sites of muscle attachment in the embryo (Liu, 2011).

The canonical role for Msk is to import proteins, such as activated MAPK into the nucleus. However, not activated MAPK or nuclear Msk were detected in the nuclei of muscle cells. As this study has shown, strong Msk immunolocalization is detected at the MTJ in stage 16 embryos, but not in the nuclei of developing muscles in stages 13-15. It cannot be ruled out that Msk is present at low levels in the muscle, and was not detectable. It is true that the YFP-Msk fusion protein was detected in the nucleus, but this may be due to the over-expression of the protein. While it is possible that nuclear Msk and MAPK in muscles are required for some aspects of myogenesis, this requires further analysis (Liu, 2011).

Future experiments will determine the precise mechanism by which Msk influences Vein secretion, localization, and/or accumulation at muscle-tendon cell attachment sites. Is it mediated through integrins? Is phosphorylation of Msk essential for activity? Msk is tyrosine phosphorylated in response to insulin and PS integrins, although the kinase remains unknown. FAK may be an example of a kinase that can phosphorylate Msk at the muscle attachment sites (Liu, 2011).

In mammalian studies, the canonical role for Importin-7 is in nuclear import. Other roles for cytoplasmic Importin-7 have not been examined. Thus, it will be interesting to uncover new roles for Importin-7, specifically in vertebrate muscle development (Liu, 2011).



SR is not expressed in developing muscle cells. Expression in stage 14 is found in the ectodermal segment border cells that serve as attachment sites for developing myotubules. At this stage, the myotubules of sr mutants are disorganized and do not make contact with their normal atttachment sites (Lee, 1995).

The involvement of Stripe in the development of proprioceptors of Drosophila

Drosophila proprioceptors (chordotonal organs) are structured as a linear array of four lineage-related cells: a neuron, a glial cell, and two accessory cells, called cap and ligament, between which the neuron is stretched. To function properly as stretch receptors, chordotonal organs must be stably anchored at both edges. The cap cells are anchored to the cuticle through specialized lineage-related attachment cells. However, the mechanism by which the ligament cells at the other edge of the organ attach is not known. The identification of specialized attachment cells is reported that anchor the ligament cells of pentascolopidial chordotonal organs (lch5) to the cuticle. The ligament attachment cells are recruited by the approaching ligament cells upon reaching their attachment site, through an EGFR-dependent mechanism. Molecular characterization of lch5 attachment cells demonstrates that they share significant properties with Drosophila tendon cells and with mammalian proprioceptive organs (Inbal, 2004).

In an attempt to characterize the origin and fate of ch attachment cells, the distribution was examined of alpha85E-tubulin (alpha85E-tub) in ch organs. This minor alpha-tub variant is known to be expressed in the cap cells and the adjacent attachment cells, as well as in the ligament cells of lch5 organs. Close inspection of the distribution of this protein in mature embryos and first instar larvae revealed another alpha85E-tub-expressing cell in close proximity to the ventral edge of the ligament cells. Rarely, two such cells were observed. These large cells appeared to be good candidates to function in the attachment of ligament cells. Indeed, further analysis demonstrated that these cells are localized within the epidermal layer and are connected to the ventral edges of the ligament cells via Integrin-mediated adhesion, as suggested by the high concentration of the Integrin ßPS subunit in the contact site between these two cell types. In addition, these cells possess many features that are typical of other types of attachment cells. To avoid confusion, the attachment cells that anchor the cap cells are referred to as CA (cap attachment) cells and to the attachment cells that anchor the ligament cells as LA (ligament attachment) cells (Inbal, 2004).

To learn more about the structural and molecular features of ch attachment cells, tests were performed to see whether these cells share any molecular properties with tendon cells, which attach muscles to the cuticle. Tendon cells have been extensively studied, and several genes that are involved in their differentiation have been identified. Since both tendon and ch attachment cells are designed to resist mechanical strain, whether the ch attachment cells express tendon cell-typical markers was examined. The formation of tendon cells requires the expression of Stripe (Sr), an early growth response (EGR)-like transcription factor. Sr induces the expression of an array of tendon cell-specific proteins, which are required for tendon cell differentiation. Double labeling wild-type embryos for alpha85E-tub and Sr reveal that Sr is expressed in ch organs in the CA, LA, and ligament cells. Sr expression was first detected in the CA cells at stage 13. CA cells are the first to express Sr in the embryo and seem to express the highest levels of Sr throughout embryonic development. The ligament cells express lower levels of Sr from late stage 14 onward, and the LA cells express Sr in stage 16 or older embryos (Inbal, 2004).

Two other genes that are implicated in tendon cell terminal differentiation are delilah (dei), which encodes a bHLH transcription factor, and ß1-tubulin (ß1-tub). Expression of both genes has been reported in ch organs; however, their exact distribution within these organs has not been described. Double labeling wild-type embryos for alpha85E-tub and Dei reveals expression of Dei in the CA and LA cells and in the cap and ligament cells. In situ hybridization reveals that ß1-tub is expressed similarly to Dei. Very low levels of ß1-tub transcripts were observed in addition in lch5 neurons. Work done in tendon cells has shown that the expression of Dei and ß1-tub is induced by Sr in a cell-autonomous manner. The fact that in lch5 organs the expression of Dei and ß1-tub is not limited to Sr-expressing cells suggests that additional mechanisms control the expression of these genes. Thus, the differential distribution of alpha85E-tub, Sr, Dei, and ß1-tub in the cells of lch5 organs adds a new dimension of complexity to these organs and raises new questions regarding the regulation of gene expression, cell fate determination, and differentiation in each cell type (Inbal, 2004).

Despite the similarities between tendon and ch attachment cells, muscles and ch organs do not share the same attachment sites, and the CA and LA cells serve for the anchoring of lch5 organs only. One prominent difference between CA, LA, and tendon cells is the expression of the alpha85E-tub protein in ch attachment cells but not in tendon cells. This suggests that alpha85E-tub has a unique function that is required in ch organs. It has been suggested that this isoform of alpha-tubulin, which is expressed specifically in ch organ accessory cells, developing muscles, and testis cyst cells, is likely to function in cells that must elongate extensively. Thus, the contribution of the alpha85E-tub to the organization of the microtubule cytoskeleton in the ch organ accessory cells are likely to affect the elasticity of the cells and their ability to withstand tension (Inbal, 2004).

Sr functions at the top of the hierarchy to direct tendon cell differentiation. In the absence of Sr, tendon cells do not develop, and the muscles fail to attach to the ectoderm. To test the role of Sr in the formation of lch5 attachment cells, how sr loss of function affects these cells was examined. Staining sr mutant embryos with anti-alpha85E-tub reveals a loss of LA cells in the absence of Sr. The CA cells were only occasionally missing; however, their morphology appeared to be abnormal. The lch5 organs were not properly stretched and appeared to be somewhat collapsed, possibly as a result of their failure to form stable attachments to the ectoderm. Thus, Sr is required for the generation of functional lch5 organs by playing a role in the formation of LA cells and in the differentiation of CA cells (Inbal, 2004).

It is not surprising that the two types of lch5 attachment cells are affected differently by the loss of Sr function. The earliest expression of Sr in the CA cells is observed in stage 13 embryos, after all cells of lch5 organs have already formed. Thus, Sr is not expressed early enough to affect primary decisions of cell fate in the lch5 lineage, but it may represent the earliest marker of the fate acquired by CA cells. As for the LA cells, their identity is defined very late in embryonic development, and Sr expression seems to be the earliest sign of their existence. Thus, Sr is likely to play a role in their induction as well as their differentiation into attachment cells (Inbal, 2004).

Sr is a member of the EGR family of transcription factors. In mammals, EGR proteins are involved in multiple developmental processes. Egr3, which shows a significant sequence similarity to Sr, is expressed in differentiating muscle spindles, a subgroup of proprioceptors. In Egr3 null mice, these proprioceptors are missing as a result of their failure to differentiate. Thus, an intriguing molecular parallelism might exist between the formation and differentiation of Drosophila and mammalian proprioceptive organs, despite the significant differences in their structure (Inbal, 2004).

The fact that the LA cells do not belong to the ch lineage raises the question of what triggers their formation. lch5 organs are initially formed with their ligament cells in a relatively dorsal position; subsequently, these cells descend until they reach their final position in the lateral cluster. Thus, the late appearance of the LA cells presents two possibilities with regard to their induction: these cells could form at a late embryonic stage independently of the ligament cells, or they could be recruited by the approaching ligament cells. To find which of these possibilities is correct, embryos were examined in which the ligament cells were ablated, relatively late in development, by expressing in them the apoptosis-inducing gene rpr, or mutant embryos were examined in which ligament cells do not form due to mutation in the gcm or repo genes. In the absence of ligament cells, the LA cells could not be detected, suggesting that their formation depends on the presence of ligament cells. When the ligament and LA cells are missing, the lch5 organs are not fully stretched, and the cap cells appear shorter than normal. However, different types of connections between the lch5 cells and their environment (e.g., the fasciculation of the lch5 axons with the intersegmental nerve) prevent a complete collapse of these organs in the absence of their ventral anchor (Inbal, 2004).

To find whether the presence of ligament cells is sufficient to induce the formation of LA cells regardless of their position, embryos were examined in which the ligament cells were abnormally localized. Mutations in abdominal-A (abd-A), homothorax (hth), and ventral veinless (vvl) result in frequent dorsal localization of lch5 organs. lch5 organs that fail to localize to their correct position in these mutants do not have LA cells. However, since the protein products of abd-A, hth, and vvl are normally expressed in the ectoderm, it is possible that, in their absence from the ectoderm of mutant embryos, LA cells cannot develop, regardless of the positioning of ligament cells. To assess specifically the influence of ligament cell positioning, an inducible Hth antimorph (En-Hth1-430) was used that can phenocopy hth loss of function. Expression of this antimorph in ch organs under the regulation of ato-Gal4 results in a high percentage of abnormally oriented lch5 organs. Except for their abnormal positioning, lch5 organs in these embryos appear to be fully differentiated, as judged by their ability to express typical markers, such as Repo, alpha85E-tub, and Sr. In ato-Gal4/UAS-En-Hth1-430 embryos, no LA cells could be observed in abdominal segments that exhibited abnormally oriented lch5 organs. Altogether, these data suggest that lch5 ligament cells recruit their attachment cells and that this process is restricted spatially, perhaps due to competence of cells in the attachment site region (Inbal, 2004).

The recruitment of LA cells by ligament cells resembles the recruitment of tendon cells by myotubes. In the case of tendon cells, the leading edges of myotubes approach preexisting clusters of Sr-expressing cells, and upon reaching their target they induce the terminal differentiation of a single tendon cell. The expression of Sr in the tendon precursor clusters also serves to attract the myotubes. In the case of ligament cells, however, no clear expression of Sr could be detected in the prospective site of their attachment prior to the appearance of the LA cell: this occurs only when the ligament cells are in their final position. Thus, despite the high similarity between the two processes, differences seem to exist in the mechanisms that guide ligament cells and myotubes to their attachment sites (Inbal, 2004).

Tendon cells are induced by the approaching myotubes, which secrete the EGFR ligand Vein and activate the EGFR pathway in the tendon precursor cells that they contact. This activation results in the expression of tendon cell-typical markers and terminal differentiation of tendon cells. Since the approaching ligament cells seem to induce the LA cells that share many properties with tendon cells, whether the EGFR pathway plays a role in the process of LA cell induction was tested. Activation of the EGFR pathway within the developing LA cells was tested by costaining wild-type embryos for the activated form of MAP kinase (dp-ERK) and for alpha85E-tub. In stage 16 embryos, low levels of dp-ERK could be detected in the LA cells but not in any of the other lch5 cells. Higher levels of dp-ERK were detected in the LA cells of stage 17 embryos. These observations demonstrate that the MAP-kinase pathway is activated in the LA cells at the time of their formation. To establish whether this pathway is necessary for the induction of these cells and whether it is mediated through EGFR activation, the EGFR pathway was blocked specifically by expressing a dominant-negative form of the receptor (DN-DER). The DN-DER was expressed throughout the ectoderm using the 69B-Gal4 driver or, in all of the Sr-expressing cells, including the LA cells, using a sr-Gal4 driver. In both cases, the expression of DN-DER abolished the formation of LA cells, indicating that activation of the EGFR pathway is necessary for LA cell development. To establish whether activation of the pathway plays a permissive or an instructive role in the formation of LA cells, whether higher levels of EGFR activation can lead to the formation of supernumerary LA cells was tested. To elevate the level of EGFR activation locally, the EGFR ligand Vein or a secreted form of the ligand Spitz (sSpi) was expressed in the ligament cells under the regulation of repo-Gal4. This excessive activation results in the formation of increased numbers of LA cells, indicating that the EGFR pathway plays an instructive role in the induction of lch5 LA cells. Expression of sSpi throughout the ectoderm led to the induction of multiple ectopic Sr-expressing cells; however, these cells did not express the alpha85E-tub protein, suggesting that EGFR pathway activity is required but not sufficient to determine the identity of an LA cell (Inbal, 2004).

While EGFR pathway activity is clearly required for the generation of LA cells, CA cells did not seem to be affected significantly by localized blocking of EGFR signaling. Moreover, CA cells appear to be almost the only cells that continue to express high levels of Sr when the EGFR pathway is blocked, suggesting that Sr expression in these cells is controlled by a different mechanism than in LA and tendon cells (Inbal, 2004).

The data suggest that Vein is expressed in the lch5 ligament cells in late stages of embryogenesis and that Vein is the major ligand responsible for EGFR activation in the prospective LA cells. The ability of Vein to induce supernumerary LA cells when overexpressed in the ligament cells is consistent with this conclusion. The role of Vein in the induction of LA cells extends the molecular similarity between the development of lch5 organs and mammalian proprioceptors. It has been shown that Neuregulin1, a Vein homolog, is secreted from proprioceptive afferent nerve endings and is required for the expression of Egr3 and differentiation of muscle spindles in the mouse (Inbal, 2004).


SR protein is found in third-instar larvae in cells fated to become muscle attachment sites in adult epidermis. There is no staining in myoblasts (Lee, 1995).

Mutual exclusion of sensory bristles and tendons on the notum of dipteran flies

Genes of the achaete-scute complex encode transcription factors whose activity regulates the development of neural cells. The spatially restricted expression of achaete-scute on the mesonotum of higher flies governs the development and positioning of the large sensory bristles. On the scutum the bristles are arranged into conserved patterns, based on an ancestral arrangement of four longitudinal rows. This pattern appears to date back to the origin of cyclorraphous flies about 100-140 million years ago. The origin of the four-row bauplan, which is independent of body size, and the reasons for its conservation, are not known. Tendons for attachment of the indirect flight muscles are invariably located between the bristle rows of the scutum throughout the Diptera. Tendon development depends on the activity of a transcription factor encoded by the gene stripe. In Drosophila, stripe and achaete-scute have separate expression domains, leading to spatial segregation of tendon precursors and bristle precursors. Furthermore the products of these genes act antagonistically: ectopic sr expression prevents bristle development and ectopic sc expression prevents normal muscle attachment. The product of stripe acts downstream of Achaete-Scute and interferes with the development of bristle precursors. It is concluded that the pattern of flight muscles has changed little throughout the Diptera and it is argued that the sites of muscle attachment may have constrained the positioning of bristles during the course of evolution. This could account for the pattern of four bristle rows on the scutum (Usui, 2004).

In Drosophila, sr and ac-sc are expressed in spatially distinct domains on the notum. This is likely to be the case too for other cyclorraphous flies. The expression pattern of sr is conserved in at least three Drosophila species, and the expression of sc during macrochaete formation in Ceratitis capitata, Calliphora vicina, and Phormia terranovae avoids the sites of muscle attachment as it does in Drosophila. When misexpressed in D. melanogaster, Sr and Sc antagonize one another's activities. Sr does not appear to repress transcription of ac-sc but may act downstream on one or more factors required to maintain high levels of proneural protein in the bristle precursors. However, in otherwise wild-type animals, loss of the endogenous sr or ac-sc gene products does not result in ectopic bristles or tendons. Nevertheless, two observations lead to the idea that sr does have a role in repressing bristle development. (1) It is expressed early in the imaginal discs long before the tendon precursors form, and (2) it appears to act redundantly with other repressors. Macrochaetes are situated outside the sites of muscle attachment in all Diptera examined, suggesting that the spatial segregation of bristles and tendons has some significance for the flies. The mutually antagonistic properties of Ac-Sc and Sr would maintain this segregation, should the normal regulation of these genes, which involves a complex genetic network and many players, be impaired (Usui, 2004).

It is interesting to speculate that sr may have been part of an ancestral mechanism of bristle patterning. The macrochaetes on the scutum of most flies are derived from a bauplan of four rows that may have been present in a common ancestor. Remarkably, in Drosophila, sr is expressed between the inferred rows of the postulated ancestral pattern. The pattern of indirect flight muscles and their attachments appears little changed throughout the Diptera, so the function of sr is likely to be phylogenetically ancient. The ancestor of the Diptera may have had randomly distributed bristles like those of extant basal flies (Nematocera). This could have resulted from ubiquitous expression of an ac-sc homolog (ASH), as is the case for the scales of butterflies and a mosquito (Nematocera) as well as the microchaetes of higher flies. If during the evolution of macrochaetes sr acquired a new function to repress bristle development, then repression by sr, in an animal with ubiquitous expression of an ASH, would have generated a pattern of rows (Usui, 2004).

The very precise positioning of macrochaetes in Drosophila is achieved by spatially restricted transcriptional activation of ac-sc in small proneural clusters that prefigure the sites of each bristle. Studies in Calliphora vicina, however, suggest that the four rows in a common ancestor of higher, cyclorraphous flies may have been generated from four stripes of sc expression. If expression of an ASH was ubiquitous in the Dipteran ancestor, then this would imply a change in the transcriptional regulation of ASHs. The proneural clusters of Drosophila result from the activity of shared cis-regulatory enhancer sequences that respond to local transcriptional activators. Nevertheless, a number of different repressors, such as the products of extramacrochaetae, hairy, and u-shaped, are also required to prevent levels of Ac-Sc accumulating outside the proneural clusters. So bristle patterning in this species relies on both activation and repression of the activity of the ac-sc genes. It is postulated that transcriptional activation may be a more recently derived patterning mechanism. If so, the cis-regulatory modules for activation of the AS-C may be of recent origin. The addition of these modules would have enhanced both the precision and robustness of the pattern. At least one of the AS-C enhancers present in cyclorraphous flies appears to be absent in Anopheles gambiae, a basal species. The number of genes at the AS-C has increased throughout the Diptera by duplication, and it is conceivable that this may have provided material for the evolution of these modules (Usui, 2004).

Flies have remarkable powers of flight and the conservation of the pattern of flight muscles probably results from strong selective pressures. Sites of muscle attachment to the epidermis are also conserved, indicating a similar location of tendons. In contrast, the positions of macrochaetes vary considerably throughout the higher flies. A survey of more than 300 species indicates, however, that this variation occurs only within the limits imposed by muscle patterning. Macrochaete patterns may therefore have been constrained during evolution by the sites of flight muscle attachment, thus accounting for the bauplan of four longitudinal rows at the origin of most patterns. The concept of developmental constraint has been discussed extensively. It proposes that certain phenotypic traits are not seen because the genetic mechanisms underlying development do not allow their formation. The alternative is that such traits are simply not favored by selection. Here, it is argued that bristle patterns may be constrained by the sites of muscle attachment (Usui, 2004).

Apart from the fact that they are mechanosensory organs, the function of macrochaetes is unknown. If the segregation of tendons and bristles is important for the function of either one, then one might expect their separation to be maintained by selection. One cannot know what selective pressures have operated in the past, and in an ancestor of the cyclorraphous flies, bristles and tendons may have been kept separate by the forces of selection. Subsequently, however, the genetic circuitry required for development could have evolved to an extent that in extant species they do not allow the development of bristles over the muscle attachment sites. One argument in favor of this comes from the study of Drosophila lines artificially selected in the laboratory. Selection for an increased number of macrochaetes on the scutum gives rise to flies with rather specific bristle patterns. Additional DC bristles form and some bristles situated on the lateral scutum. Several of these lines were examined and it was ascertained that the bristles are not located over the sites of muscle attachment that are situated on either side of the ectopic DC bristles. This suggests that artificial selection for ectopic bristles does not readily overcome the mechanism that prevents formation of bristles over muscle attachment sites. Therefore, Sr may limit the variation to generate different bristle patterns. A further observation consistent with an Sr-induced constraint is that in flightless ectoparasitic flies with divergedbristle patterns not arranged into longitudinal rows, the macrochaetes are nevertheless consistently excluded from the muscle attachment sites (Usui, 2004).

In addition to macrochaetes, higher flies have microchaetes, small mechanosensory bristles that are generally not patterned. Basal flies do not have macrochaetes (long, stout, thick bristles), and their thin, flimsy bristles may be located anywhere on the scutum. Puzzlingly, microchaetes are not excluded from the sites of muscle attachment. It is not known whether the two classes of bristles have different functions, but they differ in morphology and, at least in Drosophila, mode of development: (1) formation and maintenance of macrochaete precursors (the probable point of intervention by Sr) requires a specific regulatory sequence not used for microchaete development; (2) the microchaetes develop later, when a second Sr isoform, SrA is coexpressed with SrB (Usui, 2004).

Macrochaetes seem to have arisen in the Brachycera, and their appearance may have been caused by the acquisition of an additional, earlier phase of ASH expression. The macro- and micro-chaetes of cyclorraphous flies arise from two temporally distinct phases of sc expression, whereas all notal sensory organs of Anopheles gambiae, a basal species, arise from a single, late phase of AgASH expression. Among the derived taxa, however, there are species at scattered phylogenetic positions, devoid of macrochaetes. So it is not clear whether these structures have arisen many times or whether they arose once and have been lost in a number of lineages. A common ancestor of the monophyletic cyclorraphous flies is likely to have existed more than 100-140 myr ago, so if macrochaetes evolved only once, the four row bauplan must have been strongly selected for. In contrast, if any early accumulation of ASH were to be antagonized by Sr, then the bristles would consistently be restricted to non-sr-expressing areas, and macrochaetes arranged in similar patterns could have arisen many times independently (Usui, 2004).

In addition to being restricted to areas outside the muscle attachment sites, in many species the number as well as the position of individual macrochaetes is highly stereotyped. Amongst acalyptrate flies there has been a tendency to reduce the number of macrochaetes to just a few. This means that even at some locations devoid of sr expression, macrochaetes do not develop. Many stereotyped patterns are phylogenetically ancient; for example, the pattern in the Drosophilidae has been conserved for at least 40 myr. This suggests that in addition to exclusion from muscle attachment sites by Sr, the precise positioning of bristles may be maintained by selection. Studies of Drosophila hybrids have provided evidence of stabilizing selection for the identical bristle pattern seen between these two species. Given their scattered phylogenetic locations, the reduction/loss of wings and flight muscles in ectoparasitic species is almost certainly a result of convergence. The fact that these modifications are associated with bristle patterns that have diverged from those of winged species again suggests the patterns common to flying Diptera are subject to selective pressures. It is proposed that stereotyped macrochaete patterns may be the result of two independent forces: (1) a constraint induced by flight muscle attachment may restrict the bristles to certain locations that form the basis of the four-row bauplan; (2) selective pressures may operate to maintain precise positions of individual bristles (Usui, 2004).

Coordinated development of muscles and tendons of the Drosophila leg

A set of GFP markers and confocal microscopy has been used to analyse Drosophila leg muscle development, and all the muscles and tendons present in the adult leg are described. Importantly, evidence is provided for tendons located internally within leg segments. By visualising muscle and tendon precursors, it was demonstrated that leg muscle development is closely associated with the formation of internal tendons. In the third instars discs, in the vicinity of tendon progenitors, some Twist-positive myoblasts start to express the muscle founder cell marker dumbfounded (duf). Slightly later, in the early pupa, epithelial tendon precursors invaginate inside the developing leg segments, giving rise to the internal string-like tendons. The tendon-associated duf-lacZ-expressing muscle founders are distributed along the invaginating tendon precursors and then fuse with surrounding myoblasts to form syncytial myotubes. At mid-pupation, these myotubes grow towards their epithelial insertion sites, apodemes, and form links between internally located tendons and the leg epithelium. This leads to a stereotyped pattern of multifibre muscles that ensures movement of the adult leg (Soler, 2004).

A common feature of all Drosophila muscles is that they arise from twi-expressing non-differentiated cells. Leg muscles originate from a restricted subpopulation of such cells (5-10 myoblasts) associated with the embryonic leg disc primordia. These cells start to proliferate in the second instar larvae to form a population of about 500 myoblasts that are randomly deployed on the disc epithelium and also are known as adepithelial cells. Unlike the embryonic promuscular cells, they do not seem to be organised into clusters of cells from which progenitors of individual muscles segregate, but rather they follow the segmental subdivision of the leg disc within the proximodistal axis. This leads to the early loss of twi expression in adepithelial cells from the tarsal segments. The main feature of all Drosophila muscles that form de novo, including the larval body wall and the adult direct flight muscles, is that they develop from the specialised myoblasts named muscle founder cells. The leg muscles belong to this category of muscle, and this study shows that their formation is preceded by the specification of cells expressing the muscle founder marker duf-lacZ. How the duf-lacZ-expressing cells segregate from the population of adepithelial cells and how they become muscle founders remains unclear, but their association with sr-positive tendon progenitors suggests that interactions between these two cell types may promote their differentiation (Soler, 2004).

Interestingly, in third instar leg discs, duf-lacZ cells segregate in around only one out of five sr-expressing epithelial domains. This domain, termed the 'a' domain, is located in the dorsal Dpp-dependent portion of the disc, suggesting that Dpp signalling may be involved in eliciting this group of presumptive founders. Similar to the leg tendon precursors described in this study, sr-expressing domains have been reported in the notum of the third instar wing discs. These sr-positive domains have been reported to be involved in flight muscle patterning (Soler, 2004).

In spite of all the similarities, marked differences in appendicular versus flight and larval body wall musculatures exist that can be explained by the specific properties of leg tendons. As demonstrated by analyses of Stripe-GFP-expressing leg discs, at the end of third instar, concomitant with disc evagination, the epithelial domains of tendon progenitors start to invaginate inside the disc. This leads to the formation of internal tendons that have not been described in other body parts of the adult fly. Importantly, the presumptive founder cells associated with the invaginating tendon precursors are vectored and deployed throughout the proximodistal axis of the leg segments. Such a system provides an effective way to generate multifibre muscles in an invertebrate leg devoid of internal skeleton (Soler, 2004).

The mechanisms governing the formation of internal tendons remain to be elucidated; however, the co-expression of sr with odd in invaginating tendon precursors suggests a potential involvement of Notch. odd was previously described as an important element of the Notch-dependent cascade that controls the invagination of segmental joints. Thus, it is possible that a similar set of genes controls the different epithelial invagination events that occur in the developing leg (Soler, 2004).

Using transgenic lines that express GFP in tendon precursors (Stripe-GFP), in myoblasts and in tendons (1151-GFP), and in developing myotubes (MHC-tauGFP), it was possible to monitor appendicular myogenesis during pupa metamorphosis. At 20 hours APF, a large number of myoblasts are associated with the internal tendons, suggesting that the founder cells that are initially linked to tendons have attracted fusion-competent myoblasts to form prefusion complexes. Five hours later muscle precursors can be discerned composed of 5 to 10 nuclei, indicating that the first wave of fusion takes place between 20 and 25 hours APF. Shortly after, at 35 hours APF, the second fusion wave occurs, giving rise to the multinucleated myotubes that are attached on one side to the internal tendons. The timing of the observed fusion events is comparable to that reported previously for the de novo forming direct flight muscles. The next myogenic steps, including myotube growth, recognition of cognate sr-expressing epithelial attachment sites and induction of expression of myofibrillar proteins, are similar to the previously described events that lead to the formation of the flight and body wall muscles. The most important, unique, feature of leg muscle fibres that makes them different from other Drosophila muscles is their association with the internal tendons (Soler, 2004).

The appendicular muscle pattern revealed by this study consists of two principal muscles (levator and depressor) in each leg segment. The organisation of the muscle fibres composing levators and depressors reveals that they are attached to internal tendons. The long tendon of the tarsus extends to the femur and harbours two previously undescribed muscles, which have been designated ltm1 and ltm2 (Soler, 2004).

Overall, the computer-assisted reconstruction of the leg musculature enabled all the appendicular muscles and tendons to be identified, their anteroposterior, dorsoventral and proximodistal positions to be defined, and the number of muscle fibres that compose the individual muscles to be determined. Since this is the first reported systematic analysis of the Drosophila leg musculature, designations and their corresponding abbreviations have been proposed for all the identified muscles and tendons. In general, the proposed designations reflect the muscle and tendon functions. For example, muscles located in the femur that ensure movements of the adjacent tibia are named tibia levator (tilm) and tibia depressor (tidm) muscle (Soler, 2004).

The observations also indicate that the general pattern of appendicular muscles is invariant in males and females. However, muscle fibres that contribute to depressors and levators display distinct characteristics, suggesting differences in the genetic programme that ensures their specification. Most specifically, they differ at the ultrastructural level, displaying variations in sarcomere size and number of mitochondria. As determined by the analyses of dissected appendicular muscles, the number of nuclei that contribute to the mature fibres differs in the different types of muscle, but is relatively invariant when the same muscles from two different legs are compared. This suggests a precise control mechanism that sets up the complex events of appendicular myogenesis in Drosophila (Soler, 2004).

The association of muscle and tendon precursors in the imaginal leg discs of Drosophila reported here resembles the temporally and spatially linked development of avian tendons and muscles described in the chick hind limb, the specification of tendon progenitors in vertebrate embryos takes place very early in development, in a compartment immediately adjacent to the myotome. Thus it seems that conserved mechanisms may control the co-ordinated development of muscles and tendons in both the Drosophila leg and vertebrate embryos. An attractive possibility is that the muscle and tendon progenitors mutually promote each other's specification. The existence of such a mechanism could be easily tested in the future using Drosophila as a model system (Soler, 2004).

The muscle pattern of the Drosophila abdomen depends on a subdivision of the anterior compartment of each segment

In the past, segments were defined by landmarks such as muscle attachments, notably by Snodgrass, the king of insect anatomists. This study shows how an objective definition of a segment, based on developmental compartments, can help explain the dorsal abdomen of adult Drosophila. The anterior (A) compartment of each segment is subdivided into two domains of cells, each responding differently to Hedgehog. The anterior of these domains is non-neurogenic and clones lacking Notch develop normally; this domain can express stripe and form muscle attachments. The posterior domain is neurogenic and clones lacking Notch do not form cuticle; this domain is unable to express stripe or form muscle attachments. The posterior (P) compartment does not form muscle attachments. In vivo films indicate that early in the pupa the anterior domain of the A compartment expresses stripe in a narrowing zone that attracts the extending myotubes and resolves into the attachment sites for the dorsal abdominal muscles. The tendon cells were mapped precisely, and it was show that all are confined to the anterior domain of A. It follows that the dorsal abdominal muscles are intersegmental, spanning from one anterior domain to the next. This view is tested and supported by clones that change cell identity or express stripe ectopically. It seems that growing myotubes originate in posterior A and extend forwards and backwards until they encounter and attach to anterior A cells. The dorsal adult muscles are polarised in the anteroposterior axis: the hypothesis is disproved that muscle orientation depends on genes that define planar cell polarity in the epidermis (Krzemien, 2012).

All the dorsal muscles of a typical segment of the abdomen, including the larval persistent muscles and the MOL, attach only to a small subregion: part of the anterior (A) compartment of each adult segment. This discussion first considers the mechanisms responsible for the muscle pattern and then the general significance of what was found (Krzemien, 2012).

It appears that the pupal A compartment contains two distinct types of cells even before Hh arrives to pattern it. In an anterior domain of the A compartment, the cells develop normally without N and do not form bristles and, above a certain level, Hh specifies a1 cuticle. By contrast, when presumptive a3 cells of the posterior domain of the A compartment receive Hh, they now make a4-a6 cuticle as well as the appropriate types of bristles. Here, it was confirmed that epidermal cells of this posterior domain absolutely depend on N activity as N- clones do not form cuticle in that territory and instead make clusters of neurons. More evidence is added for the two domains: the anterior domain appears to be competent to express sr and to form muscle attachments whereas the posterior domain does neither. Also, in the thorax of Drosophila, there is earlier evidence that sensory bristles and tendon cells cannot be both formed by the same progenitors. If sr is overexpressed in the thorax, bristle formation is inhibited. Reciprocally, expression of achaete-scute in the sr domain leads to impairment of the flight muscles. This antagonism between achaete-scute and sr is conserved in many dipteran groups. These findings provide a genetic correlate with the two domains of the A compartment (Krzemien, 2012).

In general, the pattern of attachment of muscles is related to the patterned expression of sr in the ectoderm. In the embryo, as soon as the myotubes arrive nearby, they become attracted to the sr-expressing cells and contact them. This is crucial; only the epidermal cells that establish contact with a myotube can maintain a high level of sr expression and can mature properly. In the pupal abdomen, as only the anterior domain of the A compartment can express sr, one would expect muscle attachments to be limited to those cells, as was found. The P compartment is a special case; N- clones form normally showing the cells to be non-neurogenic, consistent with an absence of bristles and sensilla in P compartments. However, muscle attachments do not form there either, perhaps because the action of en on cell identity abrogates the expression of sr (Krzemien, 2012).

Consistently, it was find that ectopic expression of either srA or srB bypasses the original positional identity, even P compartment cells will attach muscles if they express ectopic sr. In these circumstances, muscles may lose their normal anteroposterior orientation and bend to reach the clones expressing sr. This is more evidence that the orientation of the dorsal abdominal muscles is a response to an Sr-dependent signal and not a response to polarity information in the epidermis (Krzemien, 2012).

The pattern of muscle attachments therefore depends on information specified in the ectoderm, raising the question of how far the mesodermal founder cells are also programmed. In pupal myogenesis, the myoblasts and histoblasts develop in close association and the movies illustrate this in vivo suggesting that they interact. The most concrete evidence for some interaction between the ectoderm and the mesoderm comes from the 'muscle of Lawrence' (MOL). The MOL is longer than other adult muscles, extending further both at the front and at the back, even though both ends attach within the tendon-competent zones. The peculiar attachment sites and size of the MOL depend entirely on the intervention of a neuron that is active only in the fifth abdominal segment of males; without this intervention, these myotubes form the same attachment sites as the normal dorsal abdominal muscles (Krzemien, 2012).

This study presents some additional evidence that signals are sent from the epidermis to the developing myotubes. Clones in the epidermis with changed cell identities were induced in third instar larvae, before the formation of myotubes. These clones affected the growing myotubes: if the tendon-competent cells of the anterior domain of the A compartment were replaced by tendon-incompetent cells of the posterior domain, the muscles passed over the transformed cells to attach to anterior domain cells further away, suggesting that a signal emanating from these tendon-competent (sr-expressing) cells had attracted the extending myotubes (Krzemien, 2012).

Films show that precise anteroposterior orientation of the developing myotubes is maintained from the beginning. How is this achieved? One possibility is that polarity information originating in the epidermis might orient the muscle migration but this is shown to be not correct, at least with regard to the PCP genes. The films also show that, just as the myotubes are forming, a band of cells in the anterior region of each A compartment expresses sr. Probably, as in the embryo, these sr-expressing cells of the pupa attract and orient the growing myotubes. Later still, sr expression is increased at the actual sites of attachment, presumably as a result of short-range interaction between the specified myoblasts and the epidermal cells; again, as in the embryo (Krzemien, 2012).

For many decades, the definition of a segment was contentious -- scientists argued using anatomical criteria about the homology of parts of one species with similar parts of another. Most had been content with Snodgrass' perspective that the integumentary plates of harder cuticle (tergites in the dorsal abdomen of Drosophila) form the essential segments and these are separated by more flexible cuticle he called the intersegmental membranes (Snodgrass, 1935). Muscle attachments were traditionally used as landmarks for the borders of segments, particularly in juvenile and primitive soft-bodied insects. Using these criteria, it was concluded that most of the longitudinal muscles were 'intrasegmental'. But, in adults or in more advanced arthropods, Snodgrass noted that most muscles spanned from one integumentary plate to the next, and he therefore described those muscles as 'intersegmental'. He concluded that the muscle attachments could shift during development and evolution. Following Brenner, one could call this rather vague picture an 'external description' of a segment. However, a different way of defining a metamere came from studies of cell lineage during development and from mapping domains of gene expression, and these methods redefined a segment as the sum of two precisely defined group of cells: the A and the P compartments. This definition amounted to an 'internal description' (Brenner, 1999; Krzemien, 2012 and references therein)

This study shows that the dorsal abdominal muscles of the abdomen can attach to a limited anterior domain of the A compartment at the front of one segment and, as they extend, they cross over the posterior domain of the A compartment as well as the P compartment to reach the anterior domain of the next segment. It follows that all the three types of muscles of the dorsal abdomen that were studied are intersegmental in extent. It is likely that these same fundamental subdivisions of the segments are conserved in other insects and, if so, would confine most muscle attachment sites to particular locations, thereby constraining the range of muscle patterns that can be built by evolution (Krzemien, 2012).

The bristle patterning genes hairy and extramacrochaetae regulate the development of structures required for flight in Diptera

The distribution of sensory bristles on the thorax of Diptera (true flies) provides a useful model for the study of the evolution of spatial patterns. Large bristles called macrochaetes are arranged into species-specific stereotypical patterns determined via spatially discrete expression of the proneural genes achaete-scute (ac-sc). In Drosophila ac-sc expression is regulated by transcriptional activation at sites where bristle precursors develop and by repression outside of these sites. Three genes, extramacrochaetae (emc), hairy (h) and stripe (sr), involved in repression have been documented. This study demonstrates that in Drosophila, the repressor genes emc and h, like sr, play an essential role in the development of structures forming part of the flight apparatus. It was found that, in Calliphora vicina a species diverged from D. melanogaster by about 100Myr, spatial expression of emc, h and sr is conserved at the location of development of those structures. Based on these findings it is argued, first, that the role emc, h and sr in development of the flight apparatus preceded their activities for macrochaete patterning; second, that species-specific variation in activation and repression of ac-sc expression is evolving in parallel to establish a unique distribution of macrochaetes in each species (Costa, 2013).

Effects of mutation or deletion

stripe gene function is required for normal muscle development. Some mutations disrupt embryonic muscle development and are lethal. Other mutations cause total loss of only a single muscle in the adult (Lee, 1995). In strong mutants scattered myotubules do not form contacts with their normal attachment sites in the epidermis. In other mutants, defects are found in specific muscles including the indirect flight muscles and the primary jump muscles (Lee, 1995). stripe was initially identified by Bridges and Morgan in 1923 through a weak mutation which showed a longitudinal "stripe" covering the dorsal roof of the adult thorax (cited in Frommer, 1996).

The Egr-type zinc-finger transcription factor encoded by the Drosophila gene stripe (sr) is expressed in a subset of epidermal cells to which muscles attach during late stages of embryogenesis. Loss-of-function and gain-of-function experiments indicate that sr activity provides ectodermal cells with properties required for the establishment of a normal muscle pattern during embryogenesis and for the differentiation of tendon-like epidermal muscle attachment sites (EMA). To interfere with the activity of both longer variants (Stripe a) and shorter variants (Stripe b) of Stripe, a dominant-negative Stripe variant was generated: the unique Stripe DNA-binding domain was fused to the repressor domain of the transcription factor Engrailed. This turns out to be a transcriptional repressor that acts from the Stripe DNA-binding domain. The fusion protein, Striperep, causes mutant stripe phenocopies, strongly disrupting muscle patterns. Striperep expression already specifically perturbs muscle pattern formation during an early stage when endogenous stripe is first expressed in a subset of ectodermal cells. The phenocritical period covers the time window when the myotubes normally undergo their oriented growth along the inner surface of the epidermis. Levels of expression of groovin, delilah and beta1-tubulin are altered in response to Striperep activities. Ectopic stripe induces groovin, delilah and beta1-tubulin only in epidermal cells. Ectopic Stripe b expression in ventral midline cells interfers with the orientation of myotubes, and the effects on the muscle pattern is restricted to muscles of the ventral half of the embryos. Thus, sr encodes a transcriptional activator that acts as an autoregulated developmental switch gene. sr activity controls the expression of EMA-specific target genes in cells of ectodermal but not of mesodermal origin. sr-expressing ectodermal cells generate long-range signals that interfere with the spatial orientation of the elongating myotubes (Vorbruggen, 1997).

Broad-Complex transcription factors regulate muscle attachment in Drosophila. In Brc mutants of the rbp complementation group, dorsoventral indirect flight muscles (DVM) are largely absent and the dorsal longitudinal indirect flight muscles, tergotrochanteral muscles (TTM), and remaining DVM often select incorrect attachment sites. One striking aspect of the rbp phenotype is that dorsal attachments are defective while ventral attachments seem normal. This phenomenon has also been observed in the TTM of bendless and myospheroid mutant adults. It is also noted that all the susceptible attachment sites are derived from the wing imaginal disc, while the unaffected sites are derived from leg discs. Analysis of the Derailed receptor tyrosine kinase suggests that muscles use different molecules for making attachments at dorsal and ventral ends of their fibers. Derailed is localized to the ventral ends of a subset of embryonic muscles; derailed mutations cause ectopic ventral attachment of these muscles. The stripe dorso-longitudinal indirect flight muscle phenotype resembles the rbp DVM phenotype, in that normal initial development is followed by degeneration and disappearance. Although stripe does not appear to be under rbp control at head eversion, the two genes may regulate overlapping subsets of downstream targets (Sandstrom, 1997).

Analysis of mutants of two gene pairs stripe and erect wing, and erect wing and vertical wings reveals that these loci exhibit a synergism. In addition, a dosage effect is apparent between ewg and sr. The ewg phenotype is similar to that of stripe These interactions suggest the existence of a functional relationship between the three loci (de la Pompa, 1989).

Stripe provides cues synergizing with Branchless to direct tracheal cell migration

The Drosophila tracheal system is an interconnected tubular respiratory network, which is formed by directed stereotypic migration and fusion of branches. Cell migration and specification are determined by combinatorial signaling of several morphogens secreted from the ectoderm. A group of ectodermal cells, marked by Stripe (Sr) expression coordinates tracheal cell migration in the dorsoventral axis. Sr, an EGR family transcription factor, is known to regulate muscle migration. Sr ectodermal cells also provide signals that are utilized for tracheal migration. These cues are separated in the time course of embryonic development. Initially, tendon-precursor cells are in close proximity to the tracheal cells, and later, when tracheal migration is complete, the muscles displace the trachea and attach to the tendon cells. sr-mutant embryos exhibit defects in migration of all tracheal branches. Although the FGF ligand Branchless (Bnl) is expressed in a subset of tendon-precursor cells independent of Sr, Bnl functions cooperatively with proteins induced by Sr in attraction of tracheal branches (Dorfman, 2002).

The pattern of Sr expression was followed with respect to tracheal cell migration. The enhancer trap sr l(3)03999 marks four groups of ectodermal cells at stage 12. One group (I) dorsally abuts and covers the dorsal part of the tracheal pit. The dorsal group of sr-positive cells at stage 13 becomes more elongated, and at stage 14 a line of Sr cells is aligned above the dorsal branch and the transverse-connective part and the lateral anterior trachea branch. All the cells expressing Sr are in intimate contact with tracheal cells that are positioned underneath. Until stage 14, the muscle precursors are positioned between the tracheal pits at the same distance from ectoderm and are not in contact with tendon precursors. At the middle of stage 14, the laterally migrating muscles displace the tracheal cells and only the terminal cell of the dorsal branch remains in contact with ectodermal cells. The muscle-precursor cells start expressing myosin at stage 14, just prior to displacement of the trachea. The displacement of trachea appears to be a result of an active process of muscle migration toward the tendon cells that coincides with myosin expression (Dorfman, 2002).

The second group of tendon-precursor cells (II) is positioned anterior to the dorsal trunk-forming cells. At stage 13, this group decreases to two cells, which disappear at midstage 13, when migration of dorsal trunk cells is almost complete. The third group of Sr-expressing cells (III) is positioned above the migrating ganglionic branches. With progression of embryonic development, concurrent with tracheal migration, the number of Sr cells increases and the cells are aligned in the dorsoventral axis above the ganglionic branches that remain in close proximity until stage 16 (Dorfman, 2002).

At stage 14, the fourth group of Sr-positive cells appears (IV), forming a line in the anterior-posterior axis just above the visceral branch. However, the tendon-precursor cells in this group are in direct contact with only two to three cells of the visceral branch at the point where visceral branch emanates from the transverse connective branch (Dorfman, 2002).

Later, at stage 15, additional rows of Sr-positive cells appear, but since the tracheal migration is complete, they are unrelated to formation of tracheal branches. However, these Sr-expressing cells may play a role in attachment of the terminal tracheal cells to the ectoderm. Terminal tracheal cells require induction by Bnl for expression of target genes, like SRF, and the formation of long extensions. The terminal cells of the dorsal branch, visualized by SRF expression, are attached to the group of ectodermal cells that express low levels of Sr. Similarly, terminal cells of the lateral branches are attached to ectoderm and tend to align along Sr-expressing cells (Dorfman, 2002).

In order to assess the role of tendon cells in regulation of tracheal development, the tracheal phenotypes of sr mutant embryos were studied. sr mutants exhibit a general failure of the tracheal branches to migrate, which is detected at early stages of tracheal migration. In sr155/DG4 embryos, which are sr null mutants, most of the dorsal branches fail to form. The initial formation of the dorsal branches and the fine cytoplasmic extensions proceeded normally, showing proper response to Bnl signaling; nonetheless, the cells of dorsal branch failed to migrate. Similarly, the dorsal trunk cells do not migrate and fuse properly, and fewer cells than normal form the lateral and ganglionic branches (Dorfman, 2002).

Likewise, in the hypomorphic background of srG11 homozygous mutants, a reduced number of cells is found in the dorsal and the ganglionic branches, and the dorsal trunk branches fail to fuse. Tracheal terminal cell differentiation is not affected in the extended branches, indicating a normal position of Bnl expression. From the mutant analysis, it appears that the dorsal branch phenotype may be functionally correlated with absence of group I Sr-expressing cells, the dorsal trunk phenotype with group II, and the ganglionic branch with group III expression patterns. Thus, distinct aspects of tracheal branch migration may be correlated to different aspects of Sr expression (Dorfman, 2002).

In sr mutant embryos, both muscle patterning and tracheal migration are impaired. However, the tracheal phenotype appears to be independent of the muscle phenotype detected in these 12453464 allelic combinations: (1) tracheal defects are detected already at stage 12, in which the muscle pattern is still normal in sr mutants; (2) in heartless (htl) mutant embryos, where most of the dorsal muscles are not formed properly, the dorsal branch appeared to be normal in segments lacking the dorsal muscles or exhibiting abnormal dorsal muscle pattern. Similarly, no tracheal defects were observed in vein mutants, which exhibit abnormal muscle migration and attachment. While Sr is essential for specifying the fate of the tendon cells, the cuticles of sr mutant embryos show no significant ectodermal patterning defects, again implying that the tracheal phenotype is not caused by malformation of the ectoderm per se. These findings show that migration of the cells in the dorsal branch is coordinated by ectodermal signals and is independent of muscles (Dorfman, 2002).

In order to test whether the highly restricted expression of Sr provides an instructive cue for tracheal migration, Sr was misexpressed in the ectoderm with 69B-Gal4. The overall tracheal migration pattern was not altered. The only deviation observed was in the number of cells allocated to the dorsal branch. Normally five to six cells form the dorsal branch. Ectopic Sr results in an increase in the number of cells forming the dorsal branch (up to 10 cells). The specification of terminal and fusion tracheal cell fates is not affected as detected by normal SRF expression in a single terminal cell. These experiments demonstrate that regulated Sr expression does not provide positional cues for migration, but rather may facilitate the activity of other restricted signals. Enhancement of such a restricted signal by ectopic Sr may account for the recruitment of extra cells to the dorsal branch (Dorfman, 2002).

Expression of Sr by the tracheal driver btl-Gal4 gives rise to a similar increase in the number of cells allocated to the dorsal branch. This result demonstrates that the ectopic effects of Sr following ubiquitous ectodermal expression are not due to a secondary effect of ectodermal patterning, but rather due to a specific effect on the dorsal tracheal branch. Furthermore, it indicates that the proteins induced by ectopic Sr, which are relevant for tracheal migration, may be produced either by the ectodermal cells or the tracheal cells and are thus likely to be functional outside of the producing cell (on the cell surface, the ECM, or as secreted proteins) (Dorfman, 2002).

Sr mutant and misexpression experiments imply that Sr induces an essential, yet noninstructive, signal for tracheal migration. To assess whether Sr is able to modulate or enhance the chemoattractive ability of Bnl, Sr or Bnl were expressed alone or simultaneously, in the salivary glands of larvae, which normally lack any trachea. Tracheal branches are attracted to the salivary glands upon ectopic expression of Bnl. Misexpression of Sr in the salivary glands attracts muscles. However, misexpression does not attract tracheal branches. In contrast, misexpression of both Sr and Bnl leads to dramatic tracheal sprouting in the salivary glands, much more pronounced than following Bnl misexpression alone. This experiment demonstrates that Sr can enhance the chemoattractive activity of Bnl. One possible way by which Sr cells may facilitate Bnl activity is through expression of an extracellular component by the tendon-precursor that is able to trap secreted Bnl, thus increasing the local concentration of this potent chemoattractant. Similarly, the activity of Bnl has been shown to depend on heparan sulfate proteoglycans. This mode of action is also employed by tendon cells via the activity of Kakapo to concentrate Vein in tendon/muscle junctions. An alternative way for Sr to provide synergizing cues to the tracheal cells is through induction of cell-cell contact or cell-matrix interaction between the tendon and the tracheal cells. However, the putative Sr-target genes that may mediate the tendon-tracheal interaction remain to be identified (Dorfman, 2002).

In conclusion, it has been shown that ectodermal cells expressing Sr provide sequential signals for migration of tracheal and muscle cells. While the signal for muscles is instructive, the cue for tracheal migration synergizes with the restricted and dynamic Bnl signal. Sr and Bnl functioned cooperatively to attract trachea to the salivary glands upon ectopic expression. Based on this result, it is tempting to speculate that, in a similar manner, the tendon cells normally mark the correct path to the tracheal cells. This may be achieved by expression of cell surface molecules that restrict the diffusion of Bnl and thus necessitate tight association between the trachea and ectoderm for proper migration (Dorfman, 2002).

Non-cell-autonomous control of denticle diversity in the Drosophila embryo

Certain Drosophila embryonic epidermal cells construct actin-based protrusions, called denticles, which exhibit stereotyped, column-specific differences in size, density and hook orientation. This precise denticle pattern is conserved throughout all drosophilids yet studied, and screening for mutations that affect this pattern has been used to identify genes involved in development and signaling. However, how column-specific differences are specified and the mechanism(s) involved have remained elusive. This study shows that the transcription factor Stripe is required for multiple aspects of the column-specific denticle pattern, including denticle hook orientation. The induction of stripe expression in certain denticle field cells appears to be the primary mechanism by which developmental pathways assign denticle hook orientation. Furthermore, the cytoskeletal linker protein Short stop (Shot) functions both cell-autonomously and non-autonomously to specify denticle hook orientation via interaction with the microtubule cytoskeleton. It is proposed that stripe mediates its effect on hook orientation, in part, via upregulation of shot (Dilks, 2010).

The denticle field within most abdominal parasegments consists of seven columns of epidermal cells that can each produce several denticles. Although the posterior-most two columns of cells produce small, non-descript denticles, cell columns 1-5 produce denticles that are uniquely patterned from column to column. These cell columns differ from one another in the size, shape, number and hook orientation of the denticles they produce, as well as in the expression of certain cell fate markers and signaling components. Of note, cell columns 1 and 4 are the only columns that produce denticles that hook towards the anterior. This anterior hooking is unique, in the wild-type denticle field as well as in situations in which the field is artificially expanded. This prompted an examination of what it is that is unique about columns 1 and 4 that allows them to produce denticles that have a reversed hook orientation. To date, no specific cell fate determinant has been found to be common to these two cell columns that might provide a clue to this process (Dilks, 2010).

Screening for mutations that affect the cuticle pattern has been a powerful approach to identify genes involved in developmental decisions. Some years ago, it became clear that the column-specific differences in denticle shape and hook orientation occurred as a result of differential activation of the signaling pathways that pattern the epidermis. However, since then, no downstream targets have been identified that selectively affect column-specific denticle diversity, and the mechanisms involved have remained elusive. This study shows that stripe integrates signaling information and positional cues to specify denticle density and anterior hook orientation. Furthermore, stripe is shown to govern hook orientation, in part via the spectraplakin shot. Thus, the stripe-shot circuit has the potential to link the unpatterned blastoderm to the fully patterned cuticle (Dilks, 2010).

stripe is a genetic target of both the Hh and Egfr pathways, and is regulated by Hh (and Wingless) directly via distinct cis-regulatory elements. Thus, these data suggest that the signals responsible for the elaborate denticle pattern act by determining stripe expression in the appropriate cell columns (Dilks, 2010).

This work significantly advances the current understanding of how pattern diversity is produced within the cuticle. The allocation of cells into the smooth and denticle fields occurs when competitive signaling interactions between the Wingless and Egfr pathways result in restricted expression of the transcription factor Shavenbaby (Svb; Ovo -- FlyBase). Svb then regulates target genes that participate in actin filament, extracellular matrix and membrane morphogenesis to construct denticles. However, whereas svb is sufficient for denticle production, stripe is required for denticle diversity by regulating target genes that modulate denticle morphogenesis. How the Stripe targets interact with, and modify, the output of Svb targets in specific denticle columns will be interesting to explore (Dilks, 2010).

Taking the data together with previously published work, stripe has at least three functions in this epidermis: the specification of the muscle attachment sites, the reversal of denticle hooking orientation, and the determination of denticle density. At least some of these functions are separable, as it was shown that muscle attachment is not required for hook orientation. However, the ventral denticle pattern is conserved throughout all the drosophilids studied so far, which suggests that these three biological characteristics are functionally coupled. This, in turn, suggests an explanation for why select denticle columns are anteriorly hooked. As muscle attachments are required for larval movement, perhaps a reversal in hook orientation at the attachment site provides for more efficient traction. Similarly, the denticles produced by tendon cells, although fewer, are often more robust and might be better able to transmit tension. It is suggested that stripe could integrate patterning information to correlate muscle attachment with denticle number and shape, the latter via cytoskeletal remodeling or modulation of cuticle secretion components. It should be possible to design biophysical approaches to test these inferences (Dilks, 2010).

One Stripe target that this study has implicated is the spectraplakin shot. Shot acts intrinsically in denticle hooking, but is also important for the non-autonomous influence that one cell column exerts on another. The phenotypic similarities between stripe and its target shot strongly suggest that shot acts as part of a non-autonomous, stripe-controlled circuit, rather than in some parallel pathway. In turn, this suggests that the smaller, stripe-dependent Shot isoform functions in the non-autonomous circuit. In addition, although it was found that muscle tension is not required, the involvement of Shot suggests that the non-autonomous signal from tendon cells to the responding cells might still be mechanical in nature. This is the first work to demonstrate a functional role for either spectraplakins or microtubules in the shaping of actin-based protrusions (Dilks, 2010).

Although Shot is required intrinsically for columns 1 and 4 to construct properly hooked denticles, there are two pools of Shot to consider in this context. Thus, it is not possible to differentiate whether Shot is required intrinsically to respond to stripe-dependent cues or for a separate structural purpose. Of course, these possibilities are not mutually exclusive, and it seems likely that Shot is intrinsically required at multiple steps (Dilks, 2010).

Shot, actin filaments and microtubules are all found within the protrusion. Since Shot has the ability to bind both actin and microtubules simultaneously, the intrinsic role of Shot might be to physically link these cytoskeletal networks within the protrusion itself. In addition, it was recently demonstrated that MACF, a vertebrate homolog of Shot, possesses actin-induced ATPase activity, which is hypothesized to move microtubules along actin filaments (Wu, 2008). Thus, Shot could function within the protrusion by maintaining physical tension between the cytoskeletal networks (Dilks, 2010).

Conversely, as Shot is also enriched at the posterior cortex of the responding cells, this pool of Shot sits in close proximity to the denticles it influences. Thus, it is tempting to speculate that this pool of Shot is required to respond to the hooking cues given by the tendon cell. One unresolved issue is how Shot in column 1 becomes enriched along the 1-2 boundary. Since Shot protein fails to localize to the posterior cortex of columns 1 and 4 in the absence of stripe, it is tempting to speculate that stripe-dependent Shot stabilizes intrinsic Shot at select column boundaries (Dilks, 2010).

Although this study demonstrated that Shot is required in tendon cells for denticles in the adjacent column to hook properly, the specific molecular requirement for Shot in the tendon cell remains to be determined (Dilks, 2010).

The column 1-2 and 4-5 interfaces are enriched for a number of cytoskeletal components (Walters, 2006; Simone, 2010), and late stage embryos show dramatic enrichment of Shot protein at these interfaces. stripe does not control the special nature of these interfaces, and raises the wider question of what initially differentiates these interfaces from other column boundaries. It is speculated that the same signaling cues that initiate stripe expression (Hh and Spi) might also be responsible for the specialization of the 1-2 and 4-5 interfaces, and serve to coordinate cytoskeletal enrichment, muscle attachment and reversal of denticle hook orientation. It is speculated that stripe-dependent Shot could be required to stabilize this enrichment along the 1-2 and 4-5 column boundaries (Dilks, 2010).

Notably, in response to Stripe, a specialized microtubule array is formed within the tendon cell that runs from the basal to the apical cell surface. Shot localizes along this array in two places - the basal cell surface, where the EMA junction will form, and the apical portion of the cell, where the microtubule array links to the cortical actin cytoskeleton. It is not known whether the apical-basal microtubule array is responsible for the localization of Shot, or whether this array is required for denticle hooking, although this would be consistent with the requirement observed for microtubules (Dilks, 2010).

It appears that the non-autonomous stripe-shot circuit culminates in the localization of Shot protein across the boundaries where denticle hooks reverse. As spectraplakins can stabilize, localize and bundle microtubule arrays, as well as create specialized membrane domains via membrane protein clustering , a likely hypothesis is that Shot organizes a specialized microtubule array or other cytoskeletal complex at these interfaces (Dilks, 2010).

Since denticle hooks manifest themselves at the level of the cuticle and not in bending of the actin filaments themselves, it is speculated that this complex could control polarized secretion, extracellular matrix deposition or membrane-matrix attachment. Alternatively, since a role was discovered for microtubules in denticle hooking, perhaps Shot remodels the cytoskeleton via membrane-microtubule attachment, as it does in tracheal cells (Dilks, 2010).


Becker, S., et al. (1997). Reciprocal signaling between Drosophila epidermal muscle attachment cells and their corresponding muscles. Development 124(13): 2615-2622. PubMed ID: 9217003

Biesiada, E., Razandi, M., and Levin, E. R. (1997). Egr-1 activates basic fibroblast growth factor transcription. Mechanistic implications for astrocyte proliferation. J. Biol. Chem. 271(31): 18576-18581. PubMed ID: 8702507

Bilioni, A., Craig, G., Hill, C. and McNeill, H. (2005). Iroquois transcription factors recognize a unique motif to mediate transcriptional repression in vivo. Proc. Natl. Acad. Sci. 102(41): 14671-6. PubMed ID: 16203991

Brenner S. (1999). Theoretical biology in the third millennium. Philos. Trans. R. Soc. Lond. B Biol. Sci. 354: 1963-1965. PubMed ID: 10670016

Chanana, B., Graf, R., Koledachkina, T., Pflanz, R. and Vorbruggen, G. (2007). αPS2 integrin-mediated muscle attachment in Drosophila requires the ECM protein Thrombospondin. Mech Dev. 124(6): 463-75. PubMed ID: 17482800

Chen, H. H., Tourtellotte, W. G. and Frank, E. (2002). Muscle spindle-derived neurotrophin 3 regulates synaptic connectivity between muscle sensory and motor neurons. J. Neurosci. 22(9): 3512-9. PubMed ID: 11978828

Copertino, D. W., Edelman, G. M. and Jones, F. S. (1997). Multiple promoter elements differentially regulate the expression of the mouse tenascin gene. Proc. Natl. Acad. Sci. 94 (5): 1846-1851. PubMed ID: 9050867

Costa, M., Calleja, M., Alonso, C. R. and Simpson, P. (2013). The bristle patterning genes hairy and extramacrochaetae regulate the development of structures required for flight in Diptera. Dev Biol. [Epub ahead of print] PubMed ID: 24384389

Criado, M., et al. (1997). Differential expression of alpha-bungarotoxin-sensitive neuronal nicotinic receptors in adrenergic chromaffin cells: a role for transcription factor Egr-1. J. Neurosci. 17(17): 6554-6564. PubMed ID: 9254668

de la Pompa, J. L., Garcia, J. R. and Ferrús, A., (1989). Genetic analysis of muscle development in Drosophila melanogaster. Dev. Biol. 131: 439-454. PubMed ID: 2492244

Delbridge, G. J. and Khachigian, L. M. (1997). FGF-1-induced platelet-derived growth factor-A chain gene expression in endothelial cells involves transcriptional activation by early growth response factor-1. Circ. Res. 81(2): 282-288. PubMed ID: 9242190

Dilks, S. A. and DiNardo, S. (2010). Non-cell-autonomous control of denticle diversity in the Drosophila embryo. Development 137(8): 1395-404. PubMed ID: 20332154

Dorfman, R., Shilo, B. Z. and Volk, T. (2002). Stripe provides cues synergizing with Branchless to direct tracheal cell migration. Dev. Bio. 252: 119-126. PubMed ID: 12453464

Dunin-Borkowski, O.M., Brown, N.H. and Bate, M. (1995). Anterior-posterior subdivision and the diversification of the mesoderm in Drosophila. Development 121: 4183-93. PubMed ID: 8575318

Edenfeld, G., et al. (2006). The splicing factor crooked neck associates with the RNA-binding protein HOW to control glial cell maturation in Drosophila. Neuron 52(6): 969-80. PubMed ID: 17178401

Frommer, G., et al. (1996). Epidermal egr-like zinc finger protein of Drosophila participates in myotube guidance. EMBO J. 15: 1642-49. PubMed ID: 8612588

Ghazi, A., Anant, S. and VijayRaghavan, K. (2000). Apterous mediates development of direct flight muscles autonomously and indirect flight muscles through epidermal cues. Development 127: 5309-5318. PubMed ID: 11076753

Ghazi, A., Paul, L. and VijayRaghavan, K. (2003). Prepattern genes and signaling molecules regulate stripe expression to specify Drosophila flight muscle attachment sites. Mech. Dev. 120: 519-528. PubMed ID: 12782269

Groupp, E. R. and Donovan-Peluso, M. (1997). Lipopolysaccharide induction of THP-1 cells activates binding of c-Jun, Ets, and Egr-1 to the tissue factor promoter. J. Biol. Chem. 271(21): 12423-12430. PubMed ID: 8647847

Harris, D. T. and Horvitz, H. R. (2011). MAB-10/NAB acts with LIN-29/EGR to regulate terminal differentiation and the transition from larva to adult in C. elegans. Development 138(18): 4051-62. PubMed ID: 21862562

Hatini, V. and DiNardo, S. (2001). Distinct signals generate repeating Striped pattern in the embryonic parasegment. Mol. Cell 7: 151-160. PubMed ID: 11172720

Ikmi, A., Netter, S. and Coen, D. (2008). Prepatterning the Drosophila notum: the three genes of the iroquois complex play intrinsically distinct roles. Dev. Biol. 317(2): 634-48. PubMed ID: 18394597

Inbal, A., Volk, T. and Salzberg, A. (2004). Recruitment of ectodermal attachment cells via an EGFR-dependent mechanism during the organogenesis of Drosophila proprioceptors. Dev. Cell 7: 241-250. PubMed ID: 15296720

Isalan, M., Choo, Y. and Klug, A. (1997). Synergy between adjacent zinc fingers in sequence-specific DNA recognition. Proc. Natl. Acad. Sci. 94(11): 5617-5621. PubMed ID: 9159121

Jacobson, C., Duggan, D. and Fischbach, G. (2004). Neuregulin induces the expression of transcription factors and myosin heavy chains typical of muscle spindles in cultured human muscle. Proc. Natl. Acad. Sci. 101(33): 12218-23. PubMed ID: 15302938

Jain, N., et al. (1996). Casein kinase II associates with Egr-1 and acts as a negative modulator of its DNA binding and transcription activities in NIH 3T3 cells. J. Biol. Chem. 271(23): 13530-13536. PubMed ID: 8662759

Khachigian, L. M., et al. (1996). Egr-1-induced endothelial gene expression: a common theme in vascular injury. Science 271(5254): 1427-1431. PubMed ID: 8596917

Klein, T. and Campos-Ortega, J. A. (1997). klumpfuss, a Drosophila gene encoding a member of the EGR family of transcription factors, is involved in bristle and leg development. Development (16): 3123-3134. PubMed ID: 9272953

Klug, A., and Schwabe, J. R. (1995). Zinc fingers. Faseb J. 9597-604 . PubMed ID:

Krzemien, J., Fabre, C. C., Casal, J. and Lawrence, P. A. (2012). The muscle pattern of the Drosophila abdomen depends on a subdivision of the anterior compartment of each segment. Development 139(1): 75-83. PubMed ID: 22147953

Lee, J. C., et al. (1995). Identification of a Drosophila muscle development gene with structural homology to mammalian early growth response transcription factors. Proc. Natl. Acad. Sci. 92: 10344-10348. PubMed ID: 7479781

Lin, J. X. and Leonard, W. J. (1997). The immediate-early gene product Egr-1 regulates the human interleukin-2 receptor beta-chain promoter through noncanonical Egr and Sp1 binding sites. Mol. Cell. Biol. 17(7): 3714-3722. PubMed ID: 9199305

Liu, C., Adamson, E., and Mercola, D. (1996). Transcription factor EGR-1 suppresses the growth and transformation of human HT-1080 fibrosarcoma cells by induction of transforming growth factor beta 1. Proc. Natl. Acad. Sci. 93(21): 11831-11836. PubMed ID: 8876223

Liu, C., et al. (1996). EGR-1, the reluctant suppression factor: EGR-1 is known to function in the regulation of growth, differentiation, and also has significant tumor suppressor activity and a mechanism involving the induction of TGF-beta1 is postulated to account for this suppressor activity. Crit. Rev. Oncog. 7(1-2): 101-125. PubMed ID: 9109500

Liu, Z. C., Geisbrecht, E. R. (2011). Moleskin is essential for the formation of the myotendinous junction in Drosophila. Dev. Biol. 359(2): 176-89. PubMed ID: 21925492

Martin-Bermudo, M. D. (2000). Integrins modulate the Egfr signaling pathway to regulate tendon cell differentiation in the Drosophila embryo. Development 127: 2607-2615. PubMed ID: 10821759

Maltzman, J. S., Carmen, J. A., and Monroe, J. G. (1996). Transcriptional regulation of the Icam-1 gene in antigen receptor- and phorbol ester-stimulated B lymphocytes: role for transcription factor EGR1. J. Exp. Med. 183(4): 1747-1759. PubMed ID: 8666932

Muthukkumar, S., et al. (1997). Role of Egr-1 gene expression in B cell receptor-induced apoptosis in an immature B cell lymphoma. J. Biol. Chem. 272: 27987-27993. PubMed ID: 9346950

Nabel-Rosen, H., et al. (2002). Two isoforms of the Drosophila RNA binding protein, How, act in opposing directions to regulate tendon cell differentiation Dev. Cell 2: 183-193. PubMed ID: 11832244

Nair, P., et al. (1997). Early growth response-1-dependent apoptosis is mediated by p53. J. Biol. Chem. 272(32): 20131-20138. PubMed ID:

Piepenburg, O., Vorbruggen, G., Jackle, H. (2000). Drosophila segment borders result from unilateral repression of hedgehog activity by wingless signaling. Mol. Cell 6(1): 203-9. PubMed ID: 10949042

Ramos, A. I. and Barolo, S. (2013). Low-affinity transcription factor binding sites shape morphogen responses and enhancer evolution. Philos Trans R Soc Lond B Biol Sci 368: 20130018. PubMed ID: 24218631

Rowan, S., Siggers, T., Lachke, S. A., Yue, Y., Bulyk, M. L. and Maas, R. L. (2010). Precise temporal control of the eye regulatory gene Pax6 via enhancer-binding site affinity. Genes Dev 24: 980-985. PubMed ID: 20413611

Sandstrom, D. J., et al. (1997). Broad-Complex transcription factors regulate muscle attachment in Drosophila. Dev. Biol. 181: 168-185. PubMed ID: 9013928

Shao H., et al. (1997). Induction of the early growth response (Egr) family of transcription factors during thymic selection. J. Exp. Med. 185(4): 731-744. PubMed ID: 9034151

Simone R. P. and DiNardo S. (2010). Actomyosin contractility and Discs large contribute to junctional conversion in guiding cell alignment within the Drosophila embryonic epithelium. Development 137: 1385-1394. PubMed ID: 20332153

Soler, C., et al. (2004). Coordinated development of muscles and tendons of the Drosophila leg. Development 131: 6041-6051. PubMed ID: 15537687

Strumpf, D. and Volk, T. (1998). Kakapo, a novel cytoskeletal-associated protein is essential for the restricted localization of the neuregulin-like factor, vein, at the muscle-tendon junction site. J. Cell Biol. 143(5): 1259-70. PubMed ID: 9832554

Subramanian, A., Wayburn, B., Bunch, T. and Volk, T. (2007). Thrombospondin-mediated adhesion is essential for the formation of the myotendinous junction in Drosophila. Development 134(7): 1269-78. PubMed ID: 17314133

Svaren, J., et al. (1996). NAB2, a corepressor of NGFI-A (Egr-1) and Krox20, is induced by proliferative and differentiative stimuli. Mol. Cell. Biol. 16(7): 3545-3553. PubMed ID: 8668170

Svaren, J., et al. (1997). The Nab2 and Stat6 genes share a common transcription termination region. Genomics 41(1): 33-39. PubMed ID: 9126479

Swirnoff, A. H., et al. (1998). Nab1, a corepressor of NGFI-A (Egr-1), contains an active transcriptional repression domain. Mol. Cell. Biol. 18(1): 512-524. PubMed ID: 9418898

Tourtellotte, W. G. and Milbrandt, J. (1998). Sensory ataxia and muscle spindle agenesis in mice lacking the transcription factor Egr3. Nat. Genet. 20: 87-91. PubMed ID: 9731539

Tourtellotte, W. G., Keller-Peck, C., Milbrandt, J. and Kucera, J. (2001). The transcription factor Egr3 modulates sensory axon-myotube interactions during muscle spindle morphogenesis. Dev. Biol. 232: 388-399. PubMed ID: 11401400

Tournay, O. and Benezra, R. (1997). Transcription of the dominant-negative helix-loop-helix protein Id1 is regulated by a protein complex containing the immediate-early response gene Egr-1. Mol. Cell. Biol. 16(5): 2418-2430. PubMed ID: 8628310

Usui, K., Pistillo, D. and Simpson, P. (2004). Mutual exclusion of sensory bristles and tendons on the notum of dipteran flies. Curr. Biol. 14: 1047-1055. PubMed ID: 15202997

Vainio, S., et al. (1993). Identification of BMP-4 as a signal mediating secondary induction between epithelial and mesenchymal tissues during early tooth development. Cell 75(1): 45-58. PubMed ID: 8104708

Volk, T. and VijayRaghaven, K. (1994). A central role for epidermal segment border cells in the induction of muscle patterning in the Drosophila embryo. Development 120: 59-70. PubMed ID:

Volohonsky, G., Edenfeld, G., Klambt, C. and Volk, T. (2007). Muscle-dependent maturation of tendon cells is induced by post-transcriptional regulation of stripeA. Development 134(2): 347-56. PubMed ID: 17166919

Vorbruggen, G. and Jackle, H. (1997). Epidermal muscle attachment site-specific target gene expression and interference with myotube guidance in response to ectopic stripe expression in the developing Drosophila epidermis. Proc. Natl. Acad. Sci. 94(16): 8606-8611. PubMed ID: 9238024

Walters, J. W., Dilks, S. A. and DiNardo, S. (2006). Planar polarization of the denticle field in the Drosophila embryo: roles for Myosin II (zipper) and fringe. Dev. Biol. 297: 323-339. PubMed ID: 16890930

Wang, D., et al. (1997). Basic fibroblast growth factor transcriptional autoregulation requires EGR-1. Oncogene 14(19): 2291-2299. PubMed ID: 9178905

Wu X., Kodama A. and Fuchs E. (2008). ACF7 regulates cytoskeletal-focal adhesion dynamics and migration and has ATPase activity. Cell 135: 137-148. PubMed ID: 18854161

Yao, J., et al. (1997). Lipopolysaccharide induction of the tumor necrosis factor-alpha promoter in human monocytic cells. Regulation by Egr-1, c-Jun, and NF-kappaB transcription factors. J. Biol. Chem. 272(28): 17795-17801. PubMed ID: 9211933

Yarnitzky, T., Min, L. and Volk, T. (1997). The Drosophila neuregulin homolog Vein mediates inductive interactions between the myotubes and their epidermal attachment cells. Genes Dev. 11: 2691-2700 . PubMed ID: 9334331

stripe: Biological Overview | Evolutionary Homologs | Regulation | Developmental Biology | Effects of Mutation

date revised: 10 February 2014 

Home page: The Interactive Fly © 1997 Thomas B. Brody, Ph.D.

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