Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Dev Biol. Author manuscript; available in PMC 2007 March 20.
Published in final edited form as:
PMCID: PMC1828914

Tissue Remodeling During Maturation Of The Drosophila Wing.


The final step in morphogenesis of the adult fly is wing maturation, a process not well understood at the cellular level due to the impermeable and refractive nature of cuticle synthesized some 30 hours prior to eclosion from the pupal case. Advances in GFP technology now make it possible to visualize cells using fluorescence after cuticle synthesis is complete. We find that, between eclosion and wing expansion, the epithelia within the folded wing begin to delaminate from the cuticle and that delamination is complete when the wing has fully expanded. After expansion, epithelial cells lose contact with each other, adherens junctions are disrupted, and nuclei become pycnotic. The cells then change shape, elongate, and migrate from the wing into the thorax. During wing maturation, the Timp gene product, Tissue Inhibitor of Metalloproteinases, and probably other components of an extracellular matrix are expressed that bond the dorsal and ventral cuticular surfaces of the wing following migration of the cells. These steps are dissected using mutants of the batone and Timp genes and ectopic expression of αPS integrin, inhibitors of Armadillo/β-catenin nuclear activity and baculovirus caspase inhibitor p35. We conclude that an epithelial-mesenchymal transition is responsible for epithelial delamination and dissolution.

Keywords: Drosophila, wing, epithelial-mesenchymal, migration, p35, Armadillo/β-catenin, Timp, αPS integrin, batone


The depth of our knowledge concerning the development and genetics of Drosophila melanogaster has made it the premier model system for animal development. Embryogenesis and growth of the larva, proliferation and patterning of the imaginal tissues within the larva, and differentiation and tissue remodeling after pupariation are well known steps in the morphogenesis of the adult fly. Following eclosion of the adult, the final step in this morphogenesis is wing maturation.

Wing maturation is preceded by well defined stages of wing development. Upon larval pupariation the imaginal wing discs evaginate. Subsequent epithelial cell expansion, without further cell proliferation, causes the wings to become compactly folded within the confines of the pupal case, prior to secretion of the adult wing cuticle. Upon eclosion this cuticle is pale and pliable, and soon an increase in blood pressure forces the wings to expand. Within approximately an hour the dorsal and ventral cuticular panels of each wing have expanded and bonded. Subsequent tanning over a period of several hours forms a strong, flexible flight organ.

The cellular development of the wing before and during early pupal development is well documented until the point at which the wing cuticle is synthesized (Cohen, 1993; Fristrom and Fristrom, 1993; Murray et al., 1995; Brabant et al., 1996). The impermeable and refractive nature of the cuticle has impeded cellular studies of the wing during later stages of pupal development and following eclosion. Electron microscopy studies of fixed and sectioned wings suggested that, at eclosion, the epithelial cells are in a state of dissolution (Johnson and Milner, 1987) and this point has been codified in influential reviews (Fristrom and Fristrom, 1993; Ashkenas et al., 1996). In addition, these studies have shown that cells are absent from the wing some hours after opening (Johnson and Milner, 1987; Roch and Akam, 2000). The advent of the Gal4/UAS system has made it possible to visualize cells of the intact wing by GFP fluorescence after cuticle synthesis, permitting more dynamic observations (Kiger et al., 2001; Kimura et al., 2004).

We have previously described many precisely arrayed fluorescent cells in newly open wings using the enhancer detector strain Gal4-30A to drive UAS-GFP expression. Gal4-30A-driven ectopic expression of Ricin A in these cells reduces their numbers and prevents bonding of dorsal and ventral wing surfaces. Ectopic expression of Protein kinase A catalytic subunit (PKAc) also prevents bonding of the wing surfaces causing hemolymph, populated with large numbers of free floating cells, to fill the wing. Eventually, this hemolymph and many of the cells are absorbed into the thorax of the fly leaving a collapsed wing with no bonding of the wing surfaces (Kiger et al., 2001). For a number of reasons these cells were tentatively identified as hemocytes. Hemocytes are known to mediate extracellular matrix (ECM) formation between dorsal and ventral epithelia during prior pupal appositions of the wing epithelia (Murray et al., 1995; Brabant et al., 1996). Thus it was reasonable to propose that they might also secrete an ECM within the wing after eclosion. Also, hemocytes might be expected to phagocytose dead or dying epithelial cells reported to be present (Johnson and Milner, 1987; Fristrom and Fristrom, 1993; Ashkenas et al., 1996). Supporting evidence that Gal4-30A might be expressed in hemocytes at this time was provided by the observation that melanotic masses, as well as disrupted wing maturation, are produced by Gal4-30A driven expression of PKAc or the dominant-negative transcription factor dTCF/Pangolin (Pan)ΔN (Kiger et al., 2001), or of Glycogen synthase kinase 3/Shaggy. Ectopic expression of these three proteins is known to block Wingless/Wnt signal transduction in many cell types (van de Wetering et al., 1997; Cavallo et al., 1998; Kiger et al., 1999; J. A. K., unpublished observations), including hemocytes, as has recently been confirmed by use of the Hemese-Gal4 driver (Zettervall, C.-J. et al., 2004). Hemese-Gal4 driven expression of PanΔN or of Shaggy causes lamellocyte differentiation and melanotic mass formation by one type of hemocyte.

Here we show that during late pupal development the Gal4-30A driver is expressed in wing epithelial cells, leading to the surprising conclusion that these putative wing hemocytes do not stem from a hematopoietic lineage but arise from wing epithelial cells. Employing mutants and ectopic gene expression studies we dissect wing maturation at the cellular level. These observations present a new view of the final stages of wing morphogenesis and indicate that an epithelial-mesenchymal transition is responsible for epithelial delamination and dissolution.

Materials and Methods

Fly Strains

The construct ywing-GFP (Wittkop et al., 2002), located on chromosome III, was a gift of S. B. Carroll. The Arm-GFP fusion transgene (McCartney et al., 2001), located on chromosome II, and the apterous-Gal4 stock ( P{GawB}apmd544) were obtained from the Bloomington Stock Center. UAS-pygo strains (Parker et al., 2002) were a gift of K. M. Cadigan. UAS-Δarm (Tolwinski and Wieschaus, 2001) was a gift of E. Wieschaus. The Timp deletion, syn28 (Godenschwege et al., 2000) was a gift of E. Buchner. The stock y w Ubx-flp; P[ry+ FRT}82B P[w+hs-πM]87E Sbsbd e P[y+]/TM6B, Tb, used to construct the stocks for clonal analysis of Timp, was a gift of S. Younger and Y.-N. Jan. The stock Hemese-Gal4, UAS-GFP.nls was a gift of D. Hultmark. The stock Gal4-684, UAS-αPS2m8 (Brabant et al., 1996) was a gift of D. Brower. We have used Gal4-30A (P{GawB}30A) in our studies here to target ectopic gene expression to wing epithelial cells prior to and after eclosion. While it is not exclusively expressed in the wing (see Flybase FBti 0002091), its effective expression in the wing blade begins after deposition of adult wing cuticle and, importantly, it is not strongly expressed in any vital cell type, permitting adults to eclose even when expressing toxic UAS transgenes. It is not obviously expressed in embryonic, larval, and pupal hemocytes as determined by UAS-GFP patterns. All other strains were described in Kiger et al. (2001), Kiger and Ho (2001), or obtained from the Bloomington Stock Center.


Standard epifluorescence microscopy of adult wings was carried out with a Zeiss Axioplan fitted with a Kodak digital camera. Wings were removed at the hinge and generally mounted dry under a coverslip taped to a slide. Wings mounted in PBX provided images similar to wings mounted dry. While wing cuticle is highly refractive if viewed by dark-field microscopy, much of the GFP fluorescence emitted from within the wing comes straight through the cuticle without refraction to create the image viewed by epifluorescence and confocal microscopy. Refraction of fluorescence within the wing cuticle (which is not itself fluorescent) is also quite apparent, because light scattered from the cuticle makes the cuticle weakly visible even in areas distant from the source of fluorescence.

Three-dimentional epifluorescence deconvolution microscopy was carried out with a DeltaVision Restoration Microscopy System (Applied Precision, Inc., Issaquah, WA). Newly open wings were removed at the hinge and mounted in PBX without fixation under a raised coverslip sealed with rubber cement. Using a mechanical stage, serial projections of 0.2 μm-spaced optical sections were collected beginning at the surface of the wing and delving progressively deeper into the wing. These individual optical sections are qualitatively similar to the images of fluorescent cells seen with standard epifluorescence microscopy. The DeltaVision System uses an iterative deconvolution algorithm to reassign out-of-focus fluorescence back to its original location computationally, in each optical section, and permits a large stack of optical sections to be rotated in space to create a 3-dimentional image.

For confocal microscopy, pupal wings were dissected and mounted in PBS without fixation. Pupae were staged using the method of Bainbridge and Bownes (1981). Adult wings were dissected and fixed for 20 minutes in 0.5 ml of 4% formaldehyde in PBS/0.95 ml heptane in 1.5 ml screw cap microfuge tubes rolled on an orbital shaker platform. Wings were then washed in PBX, cleared in a series of glycerol/PBS solutions, counter-stained with DAPI, and mounted in 70% glycerol/PBS with 2.5% DABCO. Scanning confocal fluorescence microscopy was carried out with either a Leica TCS-NT or Olympus FV-1000 system.

Fixation and staining of wings for LacZ detection was carried out as described by Montagne et al. (1996).

Clonal Analysis Of The Timp Mutant

Clonal analyses were carried out using FLP-mediated mitotic recombination (Golic and Lindquist, 1989), and a construct expressing FLP from a sequence upstream of the Ubx gene. The efficacy with which Ubx-flp can induce mitotic clones in the wing was first tested. Ubx-flp was used to induce clones expressing Gal4-30A driving UAS-PKAc, which produces wing blisters or collapsed wings. This was accomplished by removing from the clone a mutant P repressor gene (SalI ) whose product represses Gal4 transcription (Kiger and Ho, 2001). Males of genotype y w/Y; Gal4-30A UAS-PKAc 15.3/+; P[ry+ FRT]82B P[w+hs-πM]87E P[ry+ SalI]89D P[ry+ y+]96E/+ were crossed to females of genotype y w Ubx-flp/FM6; P[ry+ FRT]82B/TM3 Sb. Progeny that were Gal4-30A UAS-PKAc 15.3/+ (identified by eye color) and Sb+ y+(except for y clones) were scored for wing phenotype; 23/47 flies had normal wings and the other 24 (51%) had small to large blisters or collapsed wings. The most extreme examples approached the phenotype of Gal4-30A UAS-PKAc 15.3/+ flies, confirming that, at high frequency, Ubx-flp produces clones of PKAc-expressing cells that prevent bonding of wing surfaces.

Males of genotype y w f36a/Y ; P[ry+ FRT]82B Timp/TM3 Sb were then crossed to females of genotype y w Ubx-flp f36a; P[ry+ FRT]82B y+ f+. As a control, males of genotype y w f36a/Y ; P[ry+ FRT]82B/TM3 Sb were crossed to the above females. Sb+ progeny were scored for any type of wing defect: among the P[ry+ FRT]82B Timp/P[ry+ FRT]82B y+ f+progeny 39/652 flies exhibited wing defects; among the control P[ry+ FRT]82B/P[ry+ FRT]82B y+ f+progeny 46/695 flies exhibited wing defects. These numbers do not differ significantly (χ2 p ≈ 0.65). Normal wings from P[ry+ FRT]82B Timp/P[ry+ FRT]82B y+ f+progeny exhibited large clones marked by y and f36a bristles or f36a wing hairs. In cases where clones on the dorsal and ventral surfaces overlap, in no instance did it appear that the two surfaces had not bonded properly.


RNA was isolated from the wings of 100 Oregon-R wild type flies, eclosed for less than 30 minutes, using the Ambion Totally RNA Total RNA kit. Reverse-transcriptase PCR was performed using the Invitrogen SuperScript One-Step RT-PCR kit with Platinum Taq. Thermocycling parameters were designed according to the kit protocol with an annealing temperature of 56 °C. PCR primers were designed to span at least one intron to differentiate between RNA and DNA templates. The primers for RP49 were: DKOG52 TCCTTCCAGCTTCAAGATGACC and DKOG53 ACGTTGTGCACCAGGAACTTCT. The primers for Timp were: DKOG80 TGGGTTTATTGACGCTCCTC and DKOG81 TTAATGTCGGCATCCTACCC.

PCR products were run on a 1% agarose gel with an Invitrogen low-range mass DNA marker. The PCR product generated from the Timp RNA sequence was cut from the gel and purified using the BioRad prep-a-gene DNA purification kit. DNA sequencing was carried out with the Timp primers.

UV Irradiation

Flies of genotype y w/y w or Y; apterous-Gal4/UAS-GFP were collected in batches 1–15 minutes after eclosion, immobilized on ice, and placed dorsal side up on an ice-chilled glass plate in a Stratalinker with 254 nm bulb. Doses of 5, 50 and 500 mJ/cm2 were given to different batches. After irradiation flies were placed into vials inside a dark incubator at 25 °C. Flies receiving the two lower doses had flat, normally adhered wings. Those receiving the highest dose exhibited blisters after control flies had produced normal wings. Control flies were treated in the same way but were not irradiated.


The Spatial Arrangement Of Wing Cells

The newly open wing of a Gal4-30A, UAS-GFP fly exhibits an ordered array of fluorescent cells, the putative hemocytes (Figure 1, top). During pupal development, each epithelial cell produces a wing hair and surrounding cuticle at its apical surface, marking its location in the wing (Mitchell et al., 1983). The difference in focal plane between the hairs and the fluorescent cells made it difficult to determine their relative spatial arrangement using standard epifluorescence microscopy. We therefore turned to deconvolution microscopy to produce a three dimensional image of the wing to clarify the relationship between the Gal4-30A-labeled cells, the epithelial cells, and the wing hairs they produce. Deconvolution epifluorescence microscopy of the GFP fluorescence in optical sections of such a newly open wing reveals a honey comb organization (Figure 1, bottom). A stack of optical sections rotated 90° to present a view in the plane of the wing (Figure 1G) shows the fluorescent cells to be located between wing hairs, outlining dark areas within the wing subapical to each hair, and apparently within the plane where the epithelial cells would reside in the pupal wing (Mitchell et al., 1983).

Figure 1
Top. Newly open wing of a fly of genotype Gal4-30A UAS-GFP. Bar = 400 μm. Bottom (A–F). Serial projections in the XY plane (viewed up the Z axis) of optical section stacks of increasing depth, prepared by epifluorescence deconvolution ...

Based on the prevailing model of adult wing maturation, the interpretation of this picture would be that fluorescent Gal4-30A-labeled hemocytes have infiltrated the layer of unlabeled dying epithelial cells, ready to engulf them. This interpretation would be consistent with speculation that hemocytes might be responsible for secreting ECM components required to bond the two surfaces of the wing together (Johnson and Milner, 1987; Fristrom and Fristrom, 1993; Ashkenas et al., 1996; Kiger et al., 2001). As an alternative, the ordered array of GFP fluorescence in the open wing might indicate that Gal4-30A is expressed within the wing epithelium itself, with the observed pattern of interspersed fluorescent and dark areas indicating a change in the close cellular packing of the normal epithelial structure. This alternative model would imply that the unattached fluorescent cells observed later in the wing (Kiger et al., 2001) are derived directly from the wing epithelia.

To test these interpretations by observing the epithelial cells directly would require a bono fide fluorescent marker for wing epithelial cells. The yellow gene is expressed in cuticle and bristle secreting epithelial cells where it is required for melanization (Walter et al., 1991). Wittkopp et al. (2002) created a construct in which the yellow gene wing enhancer drives expression of GFP in wing epithelial cells of pharate adults (ywing-GFP). The ywing-GFP transgene (green fluorescence) was combined in flies with Gal4-30A driving the expression of UAS-AUG-DsRedGFP to label the putative hemocytes (red fluorescence). Figure 2 shows epifluorescence microscopy (A–C) of ywing-GFP fluorescence (Fig. 2A), Gal4-30A-driven DsRedGFP fluorescence (Fig. 2B), and their merger (Fig. 2C) in a newly open adult wing. Note that all red cells exhibit green fluorescence (epithelial cell marker), but that some green cells exhibit very little or no red fluorescence (putative hemocyte marker). Thus both GFPs can be found in the same cell, but there are no cells expressing only red fluorescence.

Figure 2
Labeling of wing cells by ywing-GFP (green) and Gal-4-30A driven UAS-AUG-DsRed (red). (A–C) Cells in a newly open adult wing viewed using epifluorescence: green (A); red (B); merger (C). Bar = 50 μm. (D–O) Cells in a pupal wing ...

Similarly, examination of pupal wings prior to eclosion demonstrated only simultaneous expression of both markers within the same cell population. From the time that Gal4-30A expression of DsRed GFP becomes evident in the wing blade during stage P9 (43–25 hours before eclosion), red fluorescence is coincident with the green fluorescence of ywing-GFP. Confocal optical sections of a stage P11 (25–22 hours before eclosion) wing are seen in Figure 2D–O.

We conclude that the cells imaged with Gal4-30A as a GFP driver in newly open wings (Figure 1) are epithelial cells that have undergone a spatial rearrangement. This conclusion coupled with a reexamination of Figure 2(A–C) leads to the further conclusion that some green cells must exhibit very little or no red fluorescence due to variability in Gal4-driven transgene expression (red) compared to that of yellow gene wing enhancer-driven expression (green).

The above evidence for a spatial rearrangement of epithelial cells in the expanded wing has been confirmed using standard epifluorescence microscopy. In order to reduce the complexity of the fluorescence pattern, we used apterous-Gal4, expressed only in dorsal wing epithelial cells, to drive UAS-GFP. A newly open wing of a fly expressing apterous-Gal4-driven GFP is shown in Figure 3(A, B); note that the wing hairs are located between the fluorescent cells (Figure 3B). Delamination from the cuticle and rearrangement of epithelial cells can be seen clearly by comparing the open wing with the folded wings of a newly eclosed fly (Figure 3C, D). In the unopened wing the wing hairs are located right above the fluorescent cells (Figure 3C), which are arranged in an even contiguous layer. Refraction of light in the cuticle makes it possible to see that the fluorescent cells still nestle within cuticular ridges (Figure 3C arrow) outlining the original locations of the cells when the cuticle was deposited. The majority of cells in the unopened wings have the appearance of the cells in Figure 3C, however, a group of cells in one of the two wings can be seen to have delaminated (Figure 3D) suggesting that delamination progresses as the wing is opening.

Figure 3
Delamination of epithelial cells from the wing cuticle in normal and batone mutant flies of genotype apterous-Gal4 UAS-GFP/+. (A) Wing (10 minutes after opening) of a batone/+ female. (B) Magnified view of A. Note that the wing hairs are centered between ...

We have also made a comparison of wings from normal and batone mutant flies employing apterous-Gal4-driven GFP. The batone gene is required in a non-autonomous manner for wing expansion. The mutant focus of batone is located in the dorsal anterior region of the embryonic fate map, far from the wing primordia, where it marks the primordia of cells that will produce the hormonal signal inducing wing expansion (Kiger et al., 2001). A newly open normal wing of a heterozygous batone/+ female (Figure 3A) is compared with an unopened wing of a batone/Y male of the same age (Figure 3E). Interestingly, delamination does not occur in batone mutant wings. Even a whole day after eclosion, a batone/Y wing shows no delamination (Figure 3F); as in Figure 3C, it is possible to see that the fluorescent cells nestle within the cuticular ridges (Figure 3F arrow) marking the locations of the cells when the cuticle was deposited. Thus, the batone mutation blocks both wing expansion and epithelial cell delamination.

Epithelial Cells Migrate From The Wing

An absence of cells within the wing several hours after eclosion has been noted (Roch and Akam, 2000), and Kimura et al. (2004) have presented observations indicating that GFP-labeled disaggregated epithelial cells flow out of the wing through the veins into the thorax. To examine this, flies with Gal4-30A driving UAS-lacZ were allowed to open their wings; wings were then removed from individual flies after successively longer periods, fixed, and stained. A time-series shows that epithelial cells disappear from the wing by one to two hours after expansion beginning with the cells in the posterior compartment most distant from the wing hinge, suggesting a regulated movement that does not appear to be confined to the veins (Figure 4).

Figure 4
Epithelial cells disappear from the wing following delamination from the cuticle. Flies of genotype Gal4-30A UAS-lacZ/UAS-GFP were allowed to open their wings fully and then, at recorded times, wings were removed, fixed, and stained for LacZ activity ...

The initial stages of migration can be followed by epifluorescence microscopy of detached wings. A wing photographed within 30 minutes of opening (Figure 5A) is shown 124 minutes later (Figure 5B). The precise array of cells in Figure 5A is maintained by cell to cell contacts (see below). The separation between adjacent round cells in the central portion of the wing in Figure 5B is much more pronounced than in Figure 5A, and their distribution has become more random. A transitional wave of changing cell shape, passing from proximal to distal (early to later), also can be observed in Figure 5B (compare Figure 5C and D). In Figure 5E, the cells in the distal anterior portion of the wing have elongated and appear to be organized lengthwise as if they were forming streams. Comparable changes occur in wings that are removed from flies at intervals following wing expansion and viewed immediately indicating that these changes are not artifacts, e.g. due to dehydration. The extreme fragility of wings at this stage and the difficulty of immobilizing the fly without introducing possible artifacts make it difficult to film the actual migration of cells from the living wing. Frequent grooming (which would have to be suppressed) of the wings by the hind legs accompanies wing maturation and may be an integral part of the process.

Figure 5
Epithelial cells break contacts with each other, change shape, and change position following opening of the wing. (A) Wing from a fly of genotype apterous-Gal4/UAS-GFP. Note the close contacts of the cells. (B) The same wing viewed 124 minutes later. ...

Cell to cell contacts at adherens junctions can be monitored during wing maturation by using a GFP-tagged Armadillo/β-catenin transgene. Arm-GFP fluorescence in wings fixed at 2, 40, and 60 minutes after eclosion is seen in Figure 6(A–C). The wings fixed at 2 (A) and 40 (B) minutes had not expanded while the wing fixed at 60 (C) minutes was fully expanded. Arm-GFP is distinctly localized to the cell membranes in Figure 6(A, B) where the cells assume the shape of stars and interdigitate. In Figure 6(C) the cells are round, Arm-GFP has moved to the cytoplasm, and cell contacts have been lost. Note also the pycnotic nuclei in the cells at this stage suggesting that these cells will die after leaving the wing.

Figure 6
The effect of age on Arm-GFP fluorescence in wings. Wings are from flies of genotype y w f36a/Y; arm-GFP/+. The time at which wings were fixed following eclosion is (A) 2 minutes, (B) 40 minutes, and (C) 60 minutes. Note the DAPI-stained nuclei (false-colored ...

Epithelial Cells Are Active Participants In Wing Maturation

Previous studies have proposed the death of epithelial cells following eclosion (Johnson and Milner, 1987; Kimura et al., 2004) without clarifying how the wing actually achieves maturation. To examine the consequences of induced cell death in the wing after eclosion, we employed UV irradiation. UV irradiation can be a valuable tool in developmental biology, from reprogramming cell fates (Kalthoff, 1971) to fate mapping embryos (Lohs-Schardin et al., 1979a,b). We reasoned that if epithelial dissolution and disappearance of cells is simply the consequence of programmed cell death, UV irradiation of the wing after eclosion would have little effect on the normal course of events.

Newly eclosed flies were irradiated dorsally with doses up to 500mJ/cm2 UV, a dose range used to study apoptosis induction in embryos (Zhou and Steller, 2003). After UV irradiation, flies opened their wings normally, but at the upper dose tight bonding of the wing surfaces did not occur, whereas wings of flies of comparable age and treatment, but unirradiated, bonded their surfaces normally (compare Figure 7A and B). In some areas of the irradiated wing, epithelial cells remain in contact with each other and have not delaminated from the cuticle, while in other areas the cells have migrated. We presume cells in the latter areas received a lower UV dose than those in the former due to protection from the maximum dose within folds of the unopened wing. Regions of intact epithelium persist in the UV-treated wings indefinitely (data not shown). Thus, UV irradiation interferes with cellular functions within the wing to prevent, rather than acquiesce in or accelerate, the breakdown of the epithelial organization and the bonding of wing surfaces (Figure 7C). It should be noted that UV irradiation did not interfere with wing expansion, a process under the control of hormones produced outside the wing and carried out by abdominal muscle contractions that increase hemolymph pressure, forcing wing expansion.

Figure 7
The effect of UV irradiation on epithelial cell behavior in flies of genotype apterous-Gal4/UAS-GFP. (A) Wing of a control fly that was not irradiated removed 1 hr after opening. (B) Wing of an irradiated fly removed 1 hr after opening. Although it is ...

Matrix metalloproteinases and their inhibitor proteins play important roles in both synthesis and degradation of ECM during tissue remodeling. The Drosophila genome encodes a single Tissue-Inhibitor-of-Metalloproteinases (Timp) gene. Homozygous deficiency for Timp prevents bonding of dorsal and ventral cuticle surfaces, producing a fluid filled wing (Godenschwege, et al., 2000). Adult wing blisters, in which the cuticle layers do not bond, can be caused by abnormal Integrin gene activities during the prepupal apposition of dorsal and ventral epithelia (Brower and Jaffe, 1989; Brabant et al., 1996). Loss of Timp function during these earlier stages might likewise produce the wing blister phenotype, without active involvement of epithelial cells after eclosion. To determine whether the Timp gene is active in the wing at the time of eclosion, we turned to RT-PCR. Primers were designed to span introns so that products amplified from mRNA can be distinguished from DNA (Figure 8A). A product of the appropriate size is strongly amplified when a template prepared from wings of newly eclosed flies is provided (Figure 8B). This product was sequenced and found to align with 100% identity to the Timp sequence CG6281-RB (Flybase), thus also confirming the predicted intron splice sites. The wing of a fly homozygous for Timp deficiency and ywing-GFP, removed approximately 22 hr after eclosion, shows that loss of Timp does not interfere with epithelial cell mobilization, since virtually all epithelial cells have left the wing (Figure 8C). Viewed in bright field the wing exhibits fluid filled blisters caused by failure of dorsal and ventral cuticle to bond (Figure 8D).

Figure 8
The Timp gene is active in unopened wings and required for bonding of the dorsal and ventral wing surfaces. (A) A schematic of the Timp gene showing the start and stop codons, intron splice sites, 5' and 3' UTRs, and RT-PCR primer placement. (B) Gel image ...

The cellular autonomy of the Timp deficiency was tested using the FLP/FRT system (see Materials and Methods). Wings with multiple mitotic clones of epithelial cells homozygous for the deficiency were normal. Because Timp must function outside of cells to form ECM, it is expected that the mutant effect of Timp deficiency might be non-autonomous due to heterozygous cells within the wing supplying normal Timp function to the entire wing. While it is reasonable to believe that epithelial cells are the source of the Timp mRNA, the presence in the wing of cells other than epithelial cells makes this assignment somewhat uncertain (see below and Figure 11A).

Figure 11
Hemese-Gal4 drives UAS-GFPnls in a small population of true hemocytes in the newly expanded wing. (A) Wing of a Hemese-Gal4 UAS-GFPnls fly. (B) Wing of a UAS-shaggy/+; Hemese-Gal4 UAS-GFPnls/+ fly. (C) Wing of a UAS-shaggy/+; Hemese-Gal4 UAS-GFPnls/+ ...

Dissecting Wing Maturation With Ectopic Gene Expression

Gal4-30A has proven to be a good tool for targeting ectopic gene expression to wing epithelial cells prior to and after eclosion (see Materials and Methods). Gal4-30A-driven expression of baculovirus caspase inhibitor p35 (Clem et al., 1991) prevents migration of most cells and blocks bonding of the wing surfaces (Figure 9) but does not appear to affect delamination from the cuticle. The cells remain within the wing suspended in hemolymph in a round state [elongated cells like those in Figure 5(E) are not seen]. In flies heterozygous for UAS-p35, presumably with a lower level of p35 expression, many fewer cells remain within the wing. These cells are round and tightly bound within the wing due to bonding of the wing surfaces; they eventually form brown spots in the wing as the fly ages (data not shown).

Figure 9
The effect of the caspase inhibitor p35 on epithelial cell behavior. Wing of a fly of genotype Gal4-30A UAS-GFP; UAS-p35 removed 18–24 hr after eclosion. (A) Bar = 400 μm. (B) Bar = 100 μm. (C) Bar = 50 μm.

We have previously demonstrated that Gal4-30A-driven expression of the dominant negative UAS-pan N5 prevents bonding of the wing surfaces, suggesting that Wingless target gene activity is important for wing maturation (Kiger et al. 2001). Evidence for a direct involvement of Armadillo/β-catenin is provided by the following observations.

First, Gal4-30A-driven ectopic expression of Pygopus (Pygo) prevents delamination of large portions of epithelium (Figure 10A, B), and dorsal and ventral cuticle surfaces do not bond. Ectopic expression of Pygo, a nuclear PHD-finger protein that anchors Arm in the nucleus (Townsley et al., 2004), is known to block expression of Arm target genes in a variety of tissues (Parker et al., 2002). The fact that some epithelial cells are able to delaminate is most probably a consequence of the variability of Gal4-driven transgene expression as commented upon earlier.

Figure 10
Epithelial cell delamination requires both normal Armadillo/β-catenin and PS integrin functions. (A) Wing from a fly of genotype Gal4-30A/UAS-pygo3–8; ywing-GFP/+ removed >24 hr after eclosion. Many fluorescent cells remain within ...

Second, ectopic expression of Shaggy/Glycogen synthase kinase 3 also causes failure of epithelial cells to delaminate from the cuticle (Figure 10C), and again dorsal and ventral cuticle surfaces do not bond. Shaggy blocks expression of Arm target genes by phosphorylating cytoplasmic Arm, promoting its degradation through the proteasome, and depleting nuclear Arm (Bourouis, 2002).

Third, ectopic expression of stabilized forms of Arm not subject to Shaggy phosphorylation (Tolwinski and Wieschaus, 2001, 2004) prevents epithelial cells from delaminating and blocks bonding of dorsal and ventral cuticle. The effect of ΔArm, a stabilized form with an N-terminal deletion that removes Shaggy phosphorylation sites, disrupts α-catenin binding and contains an N-terminal HA-tag and myristoylation site, is seen in Figure 10D; many cells fail to delaminate from the cuticle. The effect of ArmS10, a form with a smaller deletion lacking the Shaggy phosphorylation sites, is seen in Figure 10E; it also causes cells to fail to delaminate. Ectopic expression of wild type ArmS2, however, has no apparent effect; all cells delaminate, leave the wing, and wing surfaces bond normally (Figure 10E Inset).

We earlier referred to the fact that disturbance of PS Integrin activity at the time of prepupal apposition of dorsal and ventral epithelia leads to blisters in the adult wing. Brabant et al. (1996) have elegantly shown that prepupal apposition within a mitotic clone mutant for an integrin gene causes failure of apposition (a blister) between the epithelia at the position of the clone at the time of the next pupal apposition, leading to a blister in the adult wing cuticle. It was also shown that Gal4-684-driven expression of a wild type αPS-integrin gene at the time of the prepupal apposition causes blisters in the adult wing. We examined adult wings from flies of genotype Gal4-684, UAS-αPS2m8/ywingGFP (Figure 10F) and found that a significant number of cells have failed to delaminate.

The Hemese gene encodes a hemocyte-specific transmembrane protein (Kurucz et al., 2003). Use of the Hemese promoter to drive Gal4 expression (Zettervall et al., 2004), shows that a small population of true hemocytes is present in the wing during wing maturation. A newly open wing from a fly in which Hemese-Gal4 drives UAS-GFP.nls expression is shown in Figure 11A. Similarly, a newly open wing in which Hemese-Gal4 drives both UAS-GFP.nls and UAS-shaggy expression is shown in Figure 11B. The hemocytes have left the wing by the time the wing cuticle layers bond normally (Figure 11C). In contrast to Gal4-30A-driven expression, Hemese-Gal4-driven expression of Shaggy has no effect on removal of the epithelial cells from the wing or the bonding of the wing surfaces. Similar results were observed for Hemese-Gal4-driven expression of Pygo in a background containing the ywing-GFP epithelial marker, which clearly showed normal exit of epithelial cells from the wing (data not shown).


This work defines an active role for wing epithelial cells during the final stages of wing morphogenesis. Epithelial cell GFP markers have allowed us to observe the sequence of events that transforms the folded pupal wing into the flattened wing blade and to clarify that the population of non-adherent cells we observed previously (Kiger et al., 2001) is derived from the wing epithelia. We have used UV irradiation to incapacitate epithelial cells and demonstrate their active role in wing maturation, uncovering a developmental program completing wing morphogenesis.

We propose the following outline of that program based upon cell behavior: delamination and severing contacts; changing cell shape; and migration and ECM synthesis.

Stage 1, Delamination and Severing Contacts

Brabant et al. (1996) have proposed a signaling role for Integrins during the prepupal apposition that prepares cells for Integrin-based adhesion of the epithelia at the pupal apposition. Our observation that wing epithelial cells persist in the blistered regions produced by ectopic αPS-Integrin expression suggests that the Integrin interaction also prepares cells to respond to the later signal that induces epithelial delamination and dissolution. This signal is also blocked in the mutant batone, which prevents wing expansion (Kiger et al., 2001). Some cells begin to delaminate from the cuticle before wing expansion has begun (Figure 3D), and all have delaminated by the time expansion is complete. Delamination must involve severing of ECM contacts. The precision of the cellular array in a newly open wing (Figure 1, top) must derive from cell-cell contacts between stretched cells that are maintained following delamination (note the honey-comb organization in Figure 1, bottom and Figure 3B). Each cell then compacts and becomes round (as judged by the increase in fluorescence intensity; compare Figure 5A and B). The round cells have evidently severed their junctions with adjacent cells because the precise array of cells begins to break up (Figure 5C) and Arm-GFP moves from the cell membrane to the cytoplasm (Figure 6).

It would appear that disturbing the normal state of Arm/β-catenin signaling activity in epithelial cells blocks delamination. Delamination is blocked by ectopic expression of Pygo in the epithelial cells, which blocks expression of Arm target genes in a variety of tissues (Parker et al., 2002), and by ectopic expression of Shaggy, which blocks expression of Arm target genes by phosphorylating cytoplasmic Arm, promoting its degradation and depleting nuclear Arm (Bourouis, 2002). Ectopic expression of stabilized forms of Arm not subject to Shaggy phosphorylation evidently has a dominant-negative effect on Arm signaling activity in the maturing wing, blocking delamination of epithelial cells. This interpretation is supported by the following observations. First, no effect is produced by ectopic expression of wild type Arm using the same Gal4-30A driver, consistent with other reports (Tolwinski and Wieschaus, 2001), very likely indicating the efficiency with which wild type Arm is eliminated by phosphorylation and degradation through the proteasome. Second, a very low level of nuclear Arm is sufficient for target gene expression. The Arm-GFP fusion protein used here is fully active and completely covers homozygosity for a null arm allele, yet nuclear Arm-GFP cannot be detected in cells receiving a Wingless signal (D. G. McEwen and M. Peifer, personal communication). Thus, it is reasonable that non-physiologically high levels of stable forms of Arm could have a dominant-negative effect, not unlike the inhibitory effect of over-expression of Pygo on Arm-directed transcription. Herskowitz (1987) has examined in detail ways in which over-expression of a gene might cause functional inactivation of that gene’s product.

Arguing against an interpretation that the effects of ectopic gene expression might be non-specific, note that Gal4-30A-driven expression of p35 (Figure 9) does not block delamination. Nor does Gal4-30A-driven expression of either αPS Integrin or wild type transcription factor Pangolin/dTCF/LEF-1 (Kiger et al., 2001), or a dominant-negative form of CREB have any effect on wing maturation (Eresh et al., 1997; J.A.K., unpublished).

Stage 2, Changing Cell Shape

The round cells then begin to change shape, extending thin cytoplasmic filaments (Figure 5D), and elongate into spindles that associate with similarly shaped cells forming streams (Figure 5E). The fact that p35 expression interrupts developmental progression at the round cell stage clearly separates Stage 1 from the changes in cell shape, cell migration, and ECM synthesis events that follow. In some cellular contexts caspase inhibition prevents cell migration independently of blocking apoptosis (reviewed in Algeciras-Schimnich et al., 2002). Kimura et al. (2004) have shown that the nuclei of wing cells cease to retain nuclear-targeted GFP and begin to fragment their DNA at what appears to be the round cell stage, consistent with our observation of pycnotic nuclei at this stage (Figure 6).

Stage 3, Migration and ECM Synthesis

The cells migrate toward the hinge and into the body of the fly (Figure 4), leaving behind components, perhaps including Tissue Inhibitor of Metalloproteinases (Figure 8), of an ECM that will bond dorsal and ventral cuticular surfaces. It is noteworthy that Timp deficiency does not interfere with cell migration. ECM assembly must be the final step in the developmental program. The non-autonomous action of Timp in bonding cuticle secreted by mutant Timp clones suggests that Timp is present in abundance and diffuses over large distances in the wing to participate in ECM formation.

Precisely how ectopic expression of the various UAS transgenes studied here produces wing blisters or collapsed wings is not wholly clear. It seems doubtful that cells that fail to delaminate during early phases of tissue remodeling would secrete ECM components normally. Yet a variable number of cells in these wings do delaminate and leave the wing, presumably because of variation in the level of Gal4-30A expression. These cells might be expected to secrete the necessary ECM components, although the level of critical component(s) may be insufficient for normal bonding to occur in some cases. Blister formation might also be caused by the presence of numbers of undelaminated cells physically preventing ECM from bonding the underlying cuticle. Note that when ectopic p35 expression is limited, a moderate number of round, delaminated cells can become bound in the wing without producing blisters.

The presence of true hemocytes in the wing raises the question of whether these cells play a role in wing maturation. If Gal4-30A were to be expressed in these cells, as well as in epithelial cells, interpretation of ectopic expression studies would be complicated. We failed to detect any cells expressing DsRedGFP fluorescence that did not express ywing-GFP fluorescence (Fig. 2A B), suggesting that Gal4-30A is not expressed in true hemocytes. The observations that Hemese-Gal4-driven expression of Shaggy or of Pygo has no effect on wing maturation strongly suggest that the effects of Shaggy and Pygo on wing maturation are not mediated by true hemocytes exclusively, if at all. While we cannot rule out the possibility that Timp and/or other ECM components are supplied by true hemocytes, the bulk of the evidence supports an active role for epithelial cells in bonding the wing surfaces. Precocious death of epithelial cells induced by Gal4-30A-driven expression of Ricin A (Kiger et al., 2001) in late pupal epithelial cells prevents bonding of dorsal and ventral cuticle after eclosion. Because the wing cuticle is fully formed, the induced cell death must have occurred after cuticle deposition but before eclosion. UV irradiation after eclosion blocks both epithelial cell delamination and bonding of the wing surfaces (Figure 7). In addition, it is clear that mitotic clones of defective epithelial cells affect bonding of the wing surfaces. Mitotic clones mutant for an integrin gene produce blisters in the wing cuticle (Brabant et al. 1996) as do mitotic clones ectopically expressing PKAc (see Materials and Methods and Kiger and Ho, 2001).

Concluding Remarks

These studies describe for the first time the developmental program that completes morphogenesis of the adult fly Drosophila melanogaster. The requirement for a normal state of Arm/β-catenin signaling activity suggests that an epithelial-mesenchymal transition (EMT) transforms epithelial cells into mobile fibroblasts in the wing.

The best known example of an EMT in Drosophila is neuroblast delamination. In embryonic central nervous system formation, Wingless signaling has been shown to induce nonautonomously the delamination of specific neuroectoderm cells to form S2 neuroblasts (Chu-LaGraff and Doe, 1993). In peripheral nervous system formation, Wingless signaling is required for bristle formation at the wing margin, and ectopic expression of Wingless induces ectopic bristles in the wing blade (Axelrod et al., 1996). The ability of Wingless to induce neuroectoderm cells to form neuroblasts is tightly regulated by Notch in both the central and peripheral nervous systems (Struhl et al. 1993; Axelrod et al., 1996). Evidence that Notch modulates Wingless signaling by associating directly with Arm/β-catenin to regulate its transcriptional activity is presented by Hayward et al. (2005).

Arm/β-catenin signaling appears to be characteristic of EMTs. Translocation of Arm/β-catenin into the nucleus precedes gastrulation in Drosophila (Farge, 2003), the sea urchin (McClay et al.,2000), and zebrafish (Kelly et al., 2000). EMTs occur in the vertebrate neural crest when cells delaminate from the neural epithelium and migrate throughout the embryo. In the avian neural crest, dominant-negative forms of β-catenin and LEF/TCF inhibit delamination of cells from the epithelium, G1/S transition, and transcription of target genes (Burstyn-Cohen, et al., 2004). β-catenin and LEF/TCF proteins are observed to translocate to the nuclei of avian neural crest cells only during delamination and to be absent during advanced stages of migration (de Melker et al., 2004). EMTs are also a characteristic of cancer formation and can be initiated in some cancers by aberrant β-catenin activity (for review see Martinez Arias, 2001).

Multiple ways of activating Arm/β-catenin signaling exist. There are two independently regulated pathways that can target Arm/β-catenin to the proteasome, the Shaggy/Glycogen synthase kinase 3 degradation complex and the Seven in Absentia Homologue/ubiquitin ligase degradation complex (Jang et al., 2005). Multiple G protein-coupled receptors target the Shaggy/Glycogen synthase kinase 3 degradation complex for inhibition (Castellone et al., 2005; Katanaev et al., 2005). Further studies are necessary to identify the hormone(s), receptor(s) and signal transduction mechanisms acting in the wing maturation program and to relate this work to the extensive studies of the hormonal signals controlling wing expansion and cuticle tanning (McNabb et al., 1997; Baker et al., 1999; Baker and Truman, 2002; Dewey et al., 2004; Luo et al., 2005).


We thank the Bloomington Stock Center and all those mentioned who have contributed fly strains. We acknowledge and thank Mr. Onur Unal for preparing the wings in Figure 4. We are grateful to Don McEwen and Mark Peifer for sharing their knowledge of the arm-GFP transgene and to two reviewers who provided excellent suggestions for improving the manuscript. This work was supported by funds of the Agricultural Experiment Station at UC Davis to J. A. K. and J. E. N., and an NIH grant RO1GM61458 to D. A. K.


Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.


  • Algeciras-Schimnich A, Barnhart BC, Peter ME. Apoptosis-independent functions of killer caspases. Curr Opin Cell Biol. 2002;14:721–726. [PubMed]
  • Ashkenas J, Muschler J, Bissell MJ. The extracellular matrix in epithelial biology: shared molecules and common themes in distant phyla. Dev Biol. 1996;180:433–444. [PMC free article] [PubMed]
  • Axelrod JD, Matsuno K, Artavanis-Tsakonas S, Perrimon N. Interaction between Wingless and Notch signaling pathways mediated by Dishevelled. Science. 1996;271:1826–1832. [PubMed]
  • Bainbridge SP, Bownes M. Staging the metamorphosis of Drosophila melanogaster. J Embryol Exp Morph. 1981;66:57–80. [PubMed]
  • Baker JD, McNabb SL, Truman JW. The hormonal coordination of behavior and physiology at adult ecdysis in Drosophila melanogaster. J Exp Biol. 1999;202:3037–3048. [PubMed]
  • Baker JD, Truman JW. Mutations in the Drosophila glycoprotein hormone receptor, rickets, eliminate neuropeptide-induced tanning and selectively block a stereotyped behavioral program. J Exp Biol. 2002;205:2555–2565. [PubMed]
  • Behrens J, von Kries JP, Kuhl M, Bruhn L, Wedlich D, Grosschedl R, Birchmeier W. Functional interaction of beta-catenin with the transcription factor LEF-1. Nature. 1996;382:638–642. [PubMed]
  • Bourouis M. Targeted increase in Shaggy activity levels blocks Wingless signaling. Genesis. 2002;34:99–102. [PubMed]
  • Brabant MC, Fristrom D, Bunch TA, Brower DL. Distinct spatial and temporal functions for PS integrins during Drosophila wing morphogenesis. Development. 1996;122:3307–3317. [PubMed]
  • Brower DL, Jaffe SM. Requirement for integrins during Drosophila wing development. Nature. 1989;342:285–287. [PubMed]
  • Burstyn-Cohen T, Stanleigh J, Sela-Donenfeld D, Kalcheim C. Canonical Wnt activity regulates trunk neural crest delamination linking BMP/noggin signaling with G1/S transition. Development. 2004;131:5327–5339. [PubMed]
  • Castellone MD, Teramotto H, Williams BO, Druey KM, Gutkind JS. Prostaglandin E2 promotes colon cancer cell growth through a Gs-Axin-β-catenin signaling axis. Science. 2005;310:1504–1510. [PubMed]
  • Cavallo RA, Cox RT, Moline MM, Roose J, Polevoy GA, Clevers H, Peifer M, Bejsovec A. Drosophila Tcf and Groucho interact to repress Wingless signalling activity. Nature. 1998;395:604–608. [PubMed]
  • Chu-LaGraff Q, Doe CQ. Neuroblast specification and formation regulated by wingless in the Drosophila CNS. Science. 1993;261:1594–1597. [PubMed]
  • Clem RJ, Fechneimer M, Miller LK. Prevention of apoptosis by a baculovirus gene during infection of insect cells. Science. 1991;254:1388–1390. [PubMed]
  • Cohen SM. Imaginal disc development. In: Bate M, Martinez Arias A, editors. The Development of Drosophila melanogaster. II. Cold Spring Harbor Laboratory Press; 1993. pp. 747–841.
  • de Melker A, Desban N, Duband J-L. Cellular localization and signaling activity of β-catenin in migrating neural crest cells. Dev Dyn. 2004;230:708–726. [PubMed]
  • Dewey EM, McNabb SL, Ewer J, Kuo GR, Takanishi CL, Truman JW, Honegger H-W. Identification of the gene encoding Bursicon, an insect neuropeptide responsible for cuticle sclerotization and wing spreading. Curr Biol. 2004;14:1208–1213. [PubMed]
  • Eresh S, Riese J, Jackson DB, Bohmann D, Bienz M. A CREB-binding site as a target for decapentaplegic signaling during Drosophila endoderm induction. EMBO J. 1997;16:2014–2022. [PubMed]
  • Farge E. Mechanical induction of Twist in the Drosophila foregut/stomodeal primordium. Curr Biol. 2003;13:1365–1377. [PubMed]
  • Fristrom D, Fristrom JW. The metamorphic development of the adult epidermis. In: Bate M, Martinez Arias A, editors. The Development of Drosophila melanogaster. II. Cold Spring Harbor Laboratory Press; 1993. pp. 843–897.
  • Godenschwege TA, Pohar N, Buchner S, Buchner E. Inflated wings, tissue autolysis and early death in tissue inhibitor of metalloproteinases mutants of Drosophila. Eur J Cell Biol. 2000;79:495–501. [PubMed]
  • Golic KG, Lindquist S. The FLP recombinase of yeast catalyzes site-specific recombination in the Drosophila genome. Cell. 1989;59:499–509. [PubMed]
  • Hayward P, Brennan K, Sanders P, Balayo T, DasGupta R, Perrimon N, Martinez Arias A. Notch modulates Wnt signalling by associating with Armadillo/β-catenin and regulating its transcriptional activity. Development. 2005;132:1819–1830. [PMC free article] [PubMed]
  • Herskowitz I. Functional inactivation of genes by dominant negative mutations. Nature. 1987;329:219–222. [PubMed]
  • Jang KL, Shackelford J, Seo SY, Pagano JS. Up-regulation of β-catenin by a viral oncogene correlates with inhibition of the seven in absentia homolog 1 in B lymphoma cells. Proc Natl Acad Sci USA. 2005;102:18431–18436. [PubMed]
  • Johnson SA, Milner MJ. The final stages of wing development in Drosophila melanogaster. Tissue Cell. 1987;19:505–513. [PubMed]
  • Kalthoff K. Position of targets and period of competence for UV-induction of the malformation “double abdomen” in egg of Smittia spec. (Diptera, Chironomidae) Wilhelm Roux's Arch. 1971;168:63–84. [PubMed]
  • Katanaev VL, Ponzielli R, Sémériva M, Tomlinson A. Trimeric G protein-dependent Frizzled signaling in Drosophila. Cell. 2005;120:111–122. [PubMed]
  • Kelly C, Chin AJ, Leatherman JL, Kozlowski DJ, Weinberg ES. Maternally controlled β-catenin-mediated signaling is required for organizer formation in the zebrafish. Development. 2000;127:3899–3911. [PubMed]
  • Kiger JA, Jr, Eklund JL, Younger SH, O'Kane CJ. Transgenic inhibitors identify two roles for Protein Kinase A in Drosophila development. Genetics. 1999;152:281–290. [PubMed]
  • Kiger JA, Jr, Ho CI. P repressor as a tool to analyze GAL4/UAS enhancer trap phenotypes in Drosophila melanogaster. D I S. 2001;84:188–199.
  • Kiger JA, Jr, Natzle JE, Green MM. Hemocytes are essential for wing maturation in Drosophila melanogaster. Proc Natl Acad Sci USA. 2001;98:10190–10195. [PubMed]
  • Kimura K, Kodama A, Hayasaka Y, Ohta T. Activation of the cAMP/PKA signaling pathway is required for post-ecdysial cell death in wing epidermal cells of Drosophila melanogaster. Development. 2004;131:1597–1606. [PubMed]
  • Kurucz E, Zettervall CJ, Sinka R, Vilmos P, Pivarcsi A, Ekengren S, Hegedüs Z, Ando I, Hultmark D. Hemese, a hemocyte-specific transmembrane protein, affects the cellular immune response in Drosophila. Proc Natl Acad Sci USA. 2003;100:2622–2627. [PubMed]
  • Lohs-Schardin M, Sander K, Cremer C, Cremer T, Zorn C. Localized Ultraviolet laser microbeam irradiation of early Drosophila embryos: fate maps based on location and frequency of adult defects. Dev Biol. 1979;68:533–545. [PubMed]
  • Lohs-Schardin M, Cremer C, Nüsslein-Volhard C. A fate map for the larval epidermis of Drosophila melanogaster: localized cuticle defects following irradiation of the blastoderm with an Ultraviolet laser microbeam. Dev Biol. 1979;73:239–255. [PubMed]
  • Luo CW, Dewey EM, Sudo S, Ewer J, Hsu SY, Honegger HW, Hsueh AJW. Bursicon, the insect cuticle-hardening hormone, is a heterodimeric cystine knot protein that activates G protein-coupled receptor LGR2. Proc Natl Acad Sci USA. 2005;102:2820–2825. [PubMed]
  • Martinez Arias A. Epithelial mesenchymal interactions in cancer and development. Cell. 2001;105:425–431. [PubMed]
  • McClay DR, Peterson RE, Range RC, Winter-Vann AM, Ferkowicz MJ. A micromere induction signal is activated by β-catenin and acts through Notch to initiate specification of secondary mesenchyme cells in the sea urchin embryo. Development. 2000;127:5113–5122. [PubMed]
  • McNabb SL, Baker JD, Agapite J, Steller H, Riddiford LM, Truman JW. Disruption of a behavioral sequence by targeted death of peptidergic neurons in Drosophila. Neuron. 1997;19:813–823. [PubMed]
  • Mitchell HK, Roach J, Petersen NS. The morphogenesis of cell hairs on Drosophila wings. Dev Biol. 1983;95:387–398. [PubMed]
  • Montagne J, Groppe J, Guillemin K, Krasnow MA, Gehring WJ, Affolter M. The Drosophila Serum Response Factor gene is required for the formation of intervein tissue of the wing and is allelic to blistered. Development. 1996;122:2589–2597. [PubMed]
  • Murray MA, Fessler LI, Palka J. Changing distributions of extracellular matrix components during early wing morphogenesis in Drosophila. Dev Biol. 1995;168:150–165. [PubMed]
  • Parker DS, Jemison J, Cadigan KM. Pygopus, a nuclear PHD-finger protein required for Wingless signaling in Drosophila. Development. 2002;129:2565–2576. [PubMed]
  • Roch F, Akam M. Ultrabithorax and the control of cell morphology in Drosophila halteres. Development. 2000;127:97–107. [PubMed]
  • Struhl G, Fitzgerald K, Greenwald I. Intrinsic activity of the Lin-12 and Notch intracellular domains in vivo. Cell. 1993;74:331–345. [PubMed]
  • Tolwinski NS, Wieschaus E. Armadillo nuclear import is regulated by cytoplasmic anchor Axin and nuclear anchor dTCF/Pan. Development. 2001;128:2107–2117. [PubMed]
  • Tolwinski NS, Wieschaus E. A nuclear function for Armadillo/β-catenin. PLoS Biol. 2004;2(E95):486–493. [PMC free article] [PubMed]
  • Townsley FM, Cliffe A, Bienz M. Pygopus and Legless target Armadillo/β-catenin to the nucleus to enable its transcriptional co-activator function. Nature Cell Biol. 2004;6:626–633. [PubMed]
  • van de Wetering M, Cavallo R, Dooijes D, van Beest M, van Es J, Loureiro J, Ypma A, Hursh D, Jones T, Bejsovec A, Peifer M, Mortin M, Clevers H. Armadillo coactivates transcription driven by the product of the Drosophila segment polarity gene dTCF. Cell. 1997;88:789–799. [PubMed]
  • Walter MF, Black BC, Afshar G, Kermabon A-Y, Wright TR, Biessmann H. Temporal and spatial expression of the yellow gene in correlation with cuticle formation and DOPA Decarboxylase activity in Drosophila development. Dev Biol. 1991;147:32–45. [PubMed]
  • Wittkopp PJ, Vaccaro K, Carroll SB. Evolution of yellow gene regulation and pigmentation in Drosophila. Curr Biol. 2002;12:1547–1556. [PubMed]
  • Zettervall CJ, Anderl I, Williams MJ, Palmer R, Kurucz E, Ando I, Hultmark D. A directed screen for genes involved in Drosophila blood cell activation. Proc Natl Acad Sci USA. 2004;101:14192–14197. [PubMed]
  • Zhou L, Steller H. Distinct pathways mediate UV-induced apoptosis in Drosophila embryos. Dev. Cell. 2003;4:599–605. [PubMed]