Cell protrusions are outward extensions of the cell plasma membrane made by many different cell types and organisms. Important cellular processes involve cell protrusions, including nutrient resorption (Louvard et al.,1992), mechanosensing (Frolenkov et al.,2004), photosensing (Corbeil et al.,2001; Pellikka et al.,2002), establishment of cell adhesion (Vasioukhin et al.,2000), cell migration (Fulga and Rorth,2002), fusion of epithelial sheets (Jacinto et al.,2000; Martin-Blanco et al.,2000), wound healing (Wood et al.,2002), axon guidance (Ritzenthaler et al.,2000), and cell-to-cell communication (reviewed in Rorth,2003). Intercellular communication is important for the growth and patterning of tissues and organs and cell protrusions may mediate this communication by one of several mechanisms. Signaling mediated by cell protrusions can occur by means of transport and local release of signalling molecules, shedding of vesicles (Marzesco et al.,2005), display of receptors (Tomlinson et al.,1987; Hsiung et al.,2005) and membrane-tethered ligands on the protrusion (De Joussineau et al.,2003), or by establishing direct organelle exchange between connected cells (Rustom et al.,2004; reviewed in Demontis,2004).
The wing imaginal discs of Drosophila melanogaster, which will give rise to the wings and parts of the body wall of adult flies, provide an attractive system for studying cell protrusions within a developing tissue. In wing imaginal discs, a monolayer of columnar and squamous epithelial cells is arranged in a sac-like structure with the apical membranes facing an internal lumen (Fig. 1A; reviewed in Cohen,1993). Signaling among columnar cells as well as between columnar and squamous cells is important for growth and patterning of wing imaginal discs (reviewed in Ramirez-Weber and Kornberg,2000; Gibson and Schubiger,2001). Several protrusions have been identified in imaginal discs. The apical side of imaginal disc cells is decorated with microvilli (Poodry and Schneiderman,1970; Ursprung,1972) and filopodia have been described at their basal side (Eaton et al.,1995). In addition, three more kinds of protrusions have been observed, all of which have been implicated in exchanging signaling molecules important for the growth and patterning of wing imaginal disc cells. Cells at the periphery of the columnar cell sheet have long, planar filopodia-like protrusions, termed cytonemes, which connect to cells within the center of the columnar cell sheet (Ramirez-Weber and Kornberg,1999). Columnar cells also display apical cell protrusions that are a few cell diameters long and that are involved in the process of lateral inhibition (De Joussineau et al.,2003). Furthermore, a subpopulation of squamous cells forms microtubule-containing protrusions that extend through the lumen toward columnar cells (Cho et al.,2000; Gibson and Schubiger,2000).
A comprehensive analysis of the types of cell protrusions present in the wing imaginal disc has not been reported. It is therefore conceivable that we are currently underestimating the number and sort of cell protrusions present in this tissue. The identification and characterization of additional cell protrusions could be hampered by fixation of the tissue, which is known to result, at least in some cases, in the profound alteration or destruction of cell protrusions (Ramirez-Weber and Kornberg,1999).
Here, we have identified cellular protrusions of wing imaginal discs by expressing a membrane-bound form of green fluorescent protein (GFP) in columnar cells. Live, unfixed imaginal discs were then analyzed by laser scanning microscopy and three-dimensional (3D) rendering. We have identified and characterized two kinds of previously unreported protrusions. Protrusions of the apical plasma membrane that extended through the imaginal disc lumen and apparently contacted the squamous cells, and protrusions of the lateral plasma membrane that extended in between neighboring columnar cells. Both kinds of protrusions could be identified with GFP-actin, a GFP-tagged version of the Dpp receptor Thickveins (Tkv), and the endosomal marker GFP-Rab5. Implications for signaling among wing imaginal disc cells are discussed.
Apical Protrusions of the Columnar Epithelium Contact the Apposing Squamous Epithelium in Live Drosophila Wing Imaginal Discs
To identify and analyze cellular protrusions of columnar wing imaginal disc cells, we used the ap-GAL4 driver to express CD8-GFP, a transmembrane protein routinely used to mark plasma membranes (Lee and Luo,1999), in columnar cells of the dorsal compartment of wing imaginal discs. ap-GAL4 is not active in the squamous wing imaginal disc cells, limiting CD8-GFP expression to the columnar cells. Third instar larvae were dissected, transferred to culture medium containing the lipophilic dye FM4-64 to stain plasma membranes of all wing imaginal discs cells, and imaged by confocal microscopy. Stacks of XY confocal sections were collected and rendered in three dimensions (see the Experimental Procedures section). We have identified protrusions extending from the apical plasma membrane and protrusions extending from the lateral plasma membrane of columnar cells. We will first describe the apical protrusions and in subsequent paragraphs the lateral protrusions.
The apical surface of the columnar epithelium displayed many cell protrusions, ranging in length from 1 to 10 μm (Fig. 1B–D and Supplementary Movie S1, which can be viewed at http://www.interscience.wiley.com/jpages/1058-8388/suppmat). These apical protrusions appeared to be a common attribute of all wing imaginal disc columnar cells, with an average density of 2.4 ± 1.0 (S.D.) apical protrusions per cell (n = 48). Apical protrusions projected through the lumen of the wing imaginal disc toward the squamous epithelium (scheme in Fig. 1A).
To investigate in more detail the morphology of these apical protrusions, we expressed CD8-GFP in clones of cells. The 3D rendering showed that these cell protrusions arose from CD8-GFP–expressing cells in the columnar epithelium (Fig. 1E,E′), passed through the lumen that separated the columnar from the squamous epithelium (Fig. 1G,G′), and reached and apparently contacted the squamous epithelium (Fig. 1F,F′,H,H′ and Supplementary Movies S2,3). Most of these protrusions were linear, elongated structures; however, some protrusions were branched, especially in the terminal tracts. Sometimes, conversely, two protrusions initially distinct came in close contact with one another and apparently coalesced to form a single structure (Fig. 1I). Remarkably, most of these protrusions had an enlarged and roughly spherical terminal tip (Fig. 1F′,I,I′) whose diameter of approximately 1 μm exceeded the normal protrusion diameter of less than 0.2 μm. Sometimes this enlarged terminal tip was the only part of the protrusion that contacted the squamous epithelium; however, in other cases, the terminal tract of the protrusion, preceding the tip, also seemed to contact the squamous epithelium (Fig. 1I,I′).
Some apical protrusions had local bulges or widenings, which may have been due to vesicles inside the protrusion (Fig. 1I,I′), as previously reported for other types of cell protrusions (Ramirez-Weber and Kornberg,1999; Rustom et al.,2004; Hsiung et al.,2005). We conclude that wing imaginal disc columnar cells display apical protrusions that contact the squamous epithelium and that have specific morphological features.
Apical Protrusions Contain Actin
Many cellular protrusions contain an actin-based core (reviewed in Rorth,2003). To test whether apical protrusions of columnar cells also contain actin, we used ap-GAL4 to express a GFP-actin fusion protein in the dorsal compartment of wing imaginal discs. Confocal micrographs through the columnar epithelium detected GFP-actin in the cortical area of dorsal columnar cells, as expected (Fig. 2B–B″). Confocal micrographs taken at the level of the squamous epithelium of live wing imaginal discs showed GFP-actin in dot-like structures in close proximity to the squamous epithelial cells (Fig. 2A–A″), similar to the CD8-GFP labeled tips of apical protrusions (Fig. 1F,F′). The 3D rendering of the collected micrographs identified GFP-actin on apical protrusions of columnar cells that extended across the lumen and that came in close contact with the apposing squamous epithelium (Fig. 2C–E, Supplementary Movie S4), resembling CD8-GFP–labeled apical protrusions. We conclude that the apical protrusions of columnar cells contain actin.
Drosophila Prominin-like GFP Appears to Be Present on Apical Protrusions of Columnar Cells
Microvilli are protrusions of the apical membrane of polarized epithelial cells, including wing imaginal disc cells, containing an actin-based core. To test whether the CD8-GFP–labeled protrusions may correspond to microvilli, we investigated whether a marker for microvilli would identify CD8-GFP–like apical protrusions. Members of the Prominin family of pentaspan transmembrane proteins are among the most specific markers of microvilli (Corbeil et al.,2001). In mammalian cells, endogenous Prominin-1 and a Prominin-2-GFP fusion protein selectively localize to microvilli (Weigmann et al.,1997; Fargeas et al.,2003). To date, two members of the prominin family, Prominin and Prominin-like, have been identified in the Drosophila melanogaster genome (Fargeas et al.,2003; Zelhof et al.,2006). To test whether Prominin-like localizes to apical protrusions, we fused Drosophila Prominin-like to GFP and expressed the fusion protein using the GAL4/UAS system in a subset of columnar wing imaginal disc cells. By analyzing live wing imaginal discs expressing the Prominin-like-GFP fusion protein, we noticed, however, that the GFP fluorescence was undetectable (data not shown), possibly because the proper folding of the GFP molecule was impaired in the context of the fusion protein, as described for other GFP-fusion proteins (Pedelacq et al.,2006; and references therein). We therefore tested whether Prominin-like-GFP localized to apical protrusions of wing imaginal disc cells by immunostaining of fixed wing imaginal discs.
To this purpose, we expressed either CD8-GFP or Prominin-like-GFP in the dorsal compartment of wing imaginal discs and compared the localization of these two transmembrane proteins by immunostaining against GFP. Zonula adherens, which localize to the apical region of the lateral plasma membrane, and the more basal region of the lateral plasma membrane were identified using antibodies against E-cadherin and Fasciclin III, respectively. Consistent with our analysis of live wing imaginal discs expressing CD8-GFP, optical cross-sections (XZ) showed that CD8-GFP localized to the apical and lateral membrane of expressing cells and that CD8-GFP was enriched apical to the zonula adherens (Fig. 3A,A′). By contrast to live wing imaginal discs, the lumen between the squamous epithelium and columnar epithelium could not be clearly resolved by confocal microscopy in the fixed tissue. We therefore analyzed XY confocal sections that, because they were slightly tilted relative to the plane of the squamous and columnar epithelia, displayed the squamous cells, the apical surface of the columnar cells, and the lumen in between the two epithelia (Fig. 3G). In these images, the larger circumference distinguishes squamous cells from columnar cells. We detected CD8-GFP in dot-like structures at the interface between the squamous epithelium and the columnar epithelium in the dorsal compartment, but not in the non–CD8-GFP-expressing ventral compartment (Fig. 3B,B′,C). These dot-like structures presumably corresponded to cross-sections of the apical protrusions that we have identified in live wing imaginal discs (Fig. 1).
To test whether Prominin-like-GFP localizes to apical cell protrusions in fixed wing imaginal discs, we compared its localization with that of CD8-GFP expressing, fixed wing imaginal discs. In XZ confocal micrographs, Prominin-like-GFP localized to the plasma membrane apical to the zonula adherens and was mainly excluded from the lateral plasma membrane (Fig. 3D,D′). A weaker but reproducible staining of Prominin-like-GFP was also detected on the basal plasma membrane (Supplementary Figure S1). In apical XY confocal micrographs, we detected Prominin-like-GFP in dot-like structures at the interface between the squamous epithelium and the columnar epithelium in the dorsal compartment (Fig. 3E,E′,F), similar to CD8-GFP in fixed wing imaginal discs (Fig. 3B,B′,C). No such dot-like structures were detected in the ventral compartment. The localization of Prominin-like-GFP is consistent with CD8-GFP and Prominin-like-GFP identifying the same apical protrusions. Because members of the Prominin family are markers for microvilli, this finding indicates that apical protrusions may correspond to microvilli.
Lateral Protrusions of the Columnar Epithelium Extend in Between Cells Within the Same Epithelium
In addition to apical protrusions, expression of CD8-GFP in clones of cells also identified previously uncharacterized protrusions that extended from the lateral plasma membrane of columnar cells (Fig. 4A–A″). These lateral protrusions extended in between neighboring cells, usually one or two cells apart from the expressing cells. The lateral protrusions had a diameter of less than 0.2 μm and appeared to be randomly distributed along the lateral side of the cell. On average, approximately 0.5 ± 0.2 (S.D.) lateral protrusions per cell (n = 77) were detected on the apical most 8 μm of the lateral plasma membrane. Of interest, we noticed that some of the lateral protrusions displayed an enlarged terminal tip (Fig. 4A″), similar to the apical protrusions. Prominin-like-GFP was not detected on the lateral membrane (Supplementary Figure S1), suggesting that Prominin-GFP is excluded from lateral protrusions. This finding indicates that lateral protrusions might have a distinct nature from the apical protrusions.
By 3D-rendering of micrographs serially acquired along the Z-axis (see the Experimental Procedures section), we observed that some lateral protrusions extended in between neighboring cells parallel to the plane of the epithelium (XY) as well as perpendicular to the plane of the epithelium (XZ), that is parallel to the lateral membrane of columnar cells (Fig. 4B–B′,D–E′; see Fig. 4C for a scheme). To test whether these lateral protrusions contained, like the apical protrusions, actin, we expressed a GFP-actin fusion protein in clones of few cells. As shown in Figure 4F,F′, GFP-actin identified lateral cell protrusions of columnar epithelial cells that closely resembled the CD8-GFP–labeled lateral protrusions, indicating that the lateral protrusions contained actin.
Tkv-GFP Identifies Apical and Basal Protrusions of Columnar Wing Imaginal Disc Cells
The growth and patterning of wing imaginal discs requires the communication between cells through the exchange of secreted signaling molecules. The location and connectivity of apical protrusions (columnar cells to squamous cells) and lateral protrusions (columnar cells to columnar cells) makes them well suited to play a role in the exchange of such signaling molecules. Dpp is a member of the transforming growth factor-β family of secreted molecules that is important for the growth and patterning of wing imaginal discs (Spencer et al.,1982; Padgett et al.,1987; Capdevila and Guerrero,1994; Zecca et al.,1995) and that has been shown to be present in the lumen between columnar cells and squamous cells (Gibson et al.,2002). To test whether apical or lateral protrusions could mediate the communication between cells by displaying a receptor for Dpp, we expressed a GFP-tagged version of the Dpp receptor Thickveins (Tkv; Hsiung et al.,2005) in the dorsal compartment of columnar wing imaginal disc cells. XY confocal sections through the lumen of live wing imaginal discs were analyzed. Most of the Tkv-GFP fluorescence was detected on the nonprotruding plasma membrane of expressing cells. In addition, Tkv-GFP was also present on dot-like structures. These dot-like structures were costained with the lipophilic dye FM4-64 and presumably represent cross-sections of apical protrusions (Fig. 5A–B″). To test this further, we rendered the tissue in three dimensions. As shown in Figure 5C, Tkv-GFP–labeled protrusions of the apical plasma membrane projected toward the squamous epithelium, indicating that Tkv-GFP was present on apical protrusions (see also Supplementary Movies S5 and S6).
To test whether Tkv-GFP was also present on lateral protrusions, we expressed Tkv-GFP in clones of a few cells. As shown in Figure 5D–D″, Tkv-GFP identified protrusions of the lateral membrane that resembled CD8-GFP–labeled lateral protrusions. Thus, we detected Tkv-GFP on both apical and lateral protrusions of wing disc columnar cells. The presence of Tkv-GFP on apical and lateral protrusions is consistent with the notion that these cell protrusions could mediate communication between cells.
GFP-Rab5 Compartments Are Detected Inside Both Apical and Lateral Protrusions
Our analysis of CD8-GFP–labeled apical protrusions of live wing imaginal disc cells revealed bulges that exceeded the normal protrusion diameter of 0.2 μm (Fig. 1I,I′), suggesting that vesicles may be transported inside these cell protrusions. To test this hypothesis, we expressed GFP-Rab5 (Wucherpfennig et al.,2003), a marker of early endosomes (Chavrier et al.,1990), in clones of a few cells and analyzed its localization in live wing imaginal discs. GFP-Rab5 compartments were detected in XY confocal micrographs corresponding to the luminal space or to the squamous epithelium overlying the expressing cells (Fig. 6A–A″). FM4-64–labeled plasma membrane was encircling GFP-Rab5 compartments (Fig. 6A–A″), suggesting that GFP-Rab5 compartments were inside apical protrusions. In addition, by analyzing XY micrographs corresponding to the lateral membrane of columnar cells, we observed the presence of distinct GFP-Rab5 compartments in lateral protrusions (Fig. 6B–F″). Thus, both apical and lateral protrusions contain GFP-Rab5 compartments.
Cell protrusions are a common attribute of most cell types and play fundamental roles in development and disease. Here, we identify to-date uncharacterized protrusions of the apical and lateral plasma membrane in the Drosophila melanogaster wing imaginal disc columnar epithelium. We provide evidence that both apical and lateral protrusions contain actin and, furthermore, that apical protrusions might be microvilli. We find a receptor for the Dpp signaling molecule, Tkv-GFP, as well as the endosomal marker Rab5-GFP, on lateral and apical protrusions, consistent with the view that these protrusions are involved in cell-to-cell communication.
Apical Protrusions Connect Columnar Cells With Squamous Cells
Microvilli are a paradigm for apical cell protrusions of epithelia (e.g., Heintzelman and Mooseker,1992). Several electron microscopic studies have described microvilli on the apical surface of columnar wing imaginal disc cells (Schlichting et al.,2006; Ursprung,1972; Poodry and Schneiderman,1970). As seen in electron microscopic preparations, microvilli are numerous, short (on average approximately 500 nm long) protrusions extending into the lumen of wing imaginal discs. Using CD8-GFP, we have identified protrusions that extend from the apical surface of columnar wing imaginal disc cells. Unlike the microvilli identified by electron microscopy, these protrusions are long (up to 10 μm), extend through the entire lumen getting in close contact to the overlaying squamous cells, and display enlarged terminal tips (Fig. 1). Despite these apparent differences, several observations suggest that apical CD8-GFP–labeled protrusions and microvilli are one and the same structure. First, both kinds of protrusions are extensions of the apical plasma membrane. Second, both kinds of protrusions contain actin. Third, the two kinds of protrusions have a comparable diameter of less than 0.2 μm. Finally, our data are consistent with the notion that apical protrusions contain the microvilli marker Prominin-like. It is therefore conceivable that the morphological differences between microvilli, as seen in electron microscopic preparations, and CD8-GFP–labeled protrusions, as seen in live preparations, are a consequence of the technique used to visualize the protrusions. If CD8-GFP–labeled protrusions were microvilli, this would indicate that microvilli of columnar cells extended through the entire lumen and got in close contact to the overlying squamous epithelium.
Analysis of the Drosophila follicular epithelium has provided evidence that also in this epithelium microvilli can contact overlying cells, in this case the oocyte (Mahowald,1972; Mahowald and Kambysellis,1980). Microvilli of follicle cells projecting toward the oocyte have also been described in amphibians and mammals (Dantzer,1985; Villecco et al.,2002; Makabe et al.,2006), suggesting that follicle cell microvilli might establish connections with oocytes in many animal species. Considering that, over the past decades, microvilli have been identified in a plethora of epithelia and organisms, we speculate that microvilli might connect apposed epithelia during different developmental stages in several organisms. This raises therefore the intriguing possibility that microvilli, in addition to increasing the surface area of cells, also serve additional functions, for example, to mediate the communication or cohesion of cells from apposed cell layers.
What is the function of apical protrusions? Cells of the squamous and columnar epithelium exchange signals that are important for the growth and patterning of wing imaginal discs (Cho et al.,2000; Gibson and Schubiger,2000; Gibson et al.,2002; Pallavi and Shashidhara,2003,2005). Cells of the squamous epithelium display long microtubule-based protrusions directed toward the columnar epithelium (Cho et al.,2000; Gibson and Schubiger,2000). It has been proposed that these protrusions could mediate the exchange of instructive signals between the two apposed epithelia that are important for proliferation and patterning of imaginal discs (Cho et al.,2000; Gibson and Schubiger,2000). However, these protrusions were detected mainly in the prospective hinge and notum regions, but not in the pouch region of the wing imaginal disc that gives rise to the adult wing (Gibson and Schubiger,2000). It is therefore uncertain whether squamous wing imaginal disc cells could also exchange signals with the underlying columnar pouch cells through protrusions. Our results now open up the possibility that the apical protrusions of columnar cells, which are present on all columnar cells, fulfill the role in transmitting signals between the two apposed epithelia. Other possible functions for these protrusions include the provision of physical links between the two epithelia, which might be important for maintaining the disc-like shape of the wing imaginal disc. Tools to specifically ablate these protrusions will be ultimately required to address the function of these apical protrusions.
Lateral Protrusions Interconnect Cells Within the Columnar Epithelium
The development of the wing imaginal disc and other epithelia relies on short-range cell interactions among columnar cells (e.g., Diaz-Benjumea and Cohen,1995; Rulifson and Blair,1995; Zecca et al.,1995; Neumann and Cohen,1996; Milan et al.,2001,2002). Short-range cell interactions could be mediated by the spread of signaling molecules through the extracellular space in between cells (reviewed in Teleman et al.,2001; Entchev and Gonzalez-Gaitan,2002; Vincent and Dubois,2002). In addition, short-range cell interactions could be mediated by specialized cellular structures. Gap junctions are present between neighboring cells of the columnar epithelium of imaginal discs (Agrell,1968; Poodry and Schneiderman,1970). However, exchange of signals by means of gap junctions is limited to molecules of small size (< 1 kDa; Loewenstein,1981). Electron microscopy studies have further identified cytoplasmic bridges in the columnar epithelium of imaginal discs (Poodry and Schneiderman,1970). Cytoplasmatic bridges are short cellular protrusions of the basolateral membrane that connect neighboring cells and establish direct membrane and cytoplasmic continuity between connected cells, indicating that local exchange of information could occur by means of these cellular structures (Poodry and Schneiderman,1970). However, cytoplasmic bridges are similar to remnants of cytokinesis (Poodry and Schneiderman,1970), suggesting that they only connect sister cells and would, therefore, not be of major importance for pattern formation in imaginal discs, which has been shown to be independent of cell lineage (Bryant and Schneiderman,1969; Garcia-Bellido and Merriam,1969). Finally, recent reports suggest that short-range communication in the wing imaginal disc can occur by means of cellular protrusions (De Joussineau et al.,2003).
We observed in live wing imaginal discs that columnar cells have protrusions of the lateral membrane that extend in between the lateral plasma membranes of neighboring cells. In common with the apical protrusions, these lateral protrusions also displayed enlarged terminal tips (Fig. 4D–D″), suggesting that this might be a common morphological feature of cell protrusions in live tissue. The localization and direction of these protrusions is consistent with a role in mediating some of the short-range signaling that has been observed in wing imaginal discs. In principle, connections established by lateral protrusions could result in a global interaction network facilitating the exchange of survival information and positional cues. Moreover, the lateral protrusions may also contribute to long-range signaling between cells, for example, by increasing the rate of transport of signaling molecules.
The apical and lateral protrusions described here differ from cytonemes and protrusions emanating from sensory organ precursor cells, two kinds of cellular protrusions that were recently described in wing imaginal discs (Ramirez-Weber and Kornberg,1999; Hsiung et al.,2005; De Joussineau et al.,2003). Apical protrusions project toward the overlying squamous epithelium, whereas cytonemes and sensory organ precursor cell protrusions project toward columnar cells. Similar to the lateral protrusions described here, sensory organ precursor cell protrusions are short protrusions interconnecting cells in the columnar epithelium. However, sensory organ precursor cell protrusions were mainly detected at the basal side of cells or apical to the zonula adherens (De Joussineau et al.,2003). In contrast, the lateral protrusions observed here mainly emanate from below the zonula adherens. The lateral protrusions also differ from cytonemes present in the wing imaginal disc pouch in that they are shorter (3–5 μm instead of 20 μm in length) and that they apparently emanate in all directions and are not, as cytonemes, directed toward the anteroposterior or dorsoventral compartment boundaries.
Apical and Lateral Protrusions and Intercellular Signaling
Several recent studies have identified signaling molecules or their receptors associated with various cellular protrusions, providing evidence that these cellular protrusions are important for mediating intercellular signaling. Molecules associated with protrusions include Sevenless, a receptor for the Boss ligand (Tomlinson et al.,1987); Delta, a Notch ligand (De Joussineau et al.,2003); Scabrous, a signaling molecule (Chou and Chien,2002); and Tkv, a receptor for Dpp (Hsiung et al.,2005) on cellular protrusions in Drosophila and members of the erbB/HER family, receptors for EGF and related ligands, on extensions in mammalian cells (Lidke et al.,2004). We extend these findings by showing that Tkv-GFP can also localize to apical and lateral cell protrusions. We note, however, that we do not find Tkv-GFP enriched on these protrusions compared with the nonprotruding plasma membrane, nor do we find Tkv-GFP in these protrusions on motile punctae, as previously described for cytonemes (Hsiung et al.,2005).
A further indication that apical and lateral protrusions might be involved in mediating signals between cells is our finding that Rab5-GFP localizes to these protrusions. One way by which cellular protrusions convey instructive cues is by endocytosis and trafficking of signaling complexes on endosomes. The early endosomal protein Rab5a localizes, for example, to axons and dendrites of neurons (de Hoop et al.,1994) and is important for coupling clathrin-dependent endocytosis to axonal retrograde transport (Deinhardt et al.,2006). Furthermore, several signaling molecules are associated with endosomes (reviewed in Miaczynska et al.,2004). It will therefore be interesting to learn whether endocytosis takes place from the plasma membrane of apical and lateral protrusions and whether endosomes in these protrusions associate with signaling complexes.
Molecular Cloning and Fly Stocks
To generate the UAS-prominin-like-GFP transgene, the enhanced GFP (EGFP) coding sequence was polymerase chain reaction (PCR) amplified with primers 5′-CGGCTCGAGATGGTGAGCAAGGGCGAGGAGCTG-3′ and 5′-CTAGTCTAGATTACTTGTACAGCTCGTCCATGCC-3′ (the underlined sequences are XhoI and XbaI restriction sites used for cloning) using pEGFP-N1 (Clontech) as template (nn 679-1398), and cloned in the pUAST vector (Brand and Perrimon,1993). Drosophila melanogaster prominin-like (CG7740) coding sequence (nucleotides 158-3199 of AF197345) and 5′-UTR (Fargeas et al.,2003) were PCR amplified from cDNA clone LD16666 (kindly provided by Denis Corbeil) with primers: 5′-CCGCTCGAGTCGATCGGATAAATTTGGAATAGAAAAGGC-3′ and 5′-CCGCTCGAGATCCTGCTCGGAGGCACCCGGATAG-3′ (the underlined sequences are XhoI restriction sites used for cloning). The PCR product was digested with XhoI and cloned in frame in the pUAST-EGFP vector. The correct nucleotide sequences of the cloned PCR products were confirmed by sequencing before injection into y w embryos to obtain transgenic flies.
Additional fly stocks used in this work include ap-GAL4 (Calleja et al.,1996), act5c>CD2>GAL4 (Pignoni and Zipursky,1997), UAS-CD8-GFP (Lee and Luo,1999), UAS-tkv-GFP (Hsiung et al.,2005), UAS-GFP-Rab5 (Wucherpfennig et al.,2003), and UAS-GFP-actin (Verkhusha et al.,1999). Marked clones composed of few cells were generated by Flp-mediated mitotic recombination (Golic and Lindquist,1989) subjecting third-instar larvae to a 34°C heat-shock for 30 min at 20 hr before dissection.
Late third-instar larvae were dissected in ice-cold Ringer's solution and fixed for 40 min at room temperature in PEM solution (0.1 M PIPES, 2 mM MgSO4, 1 mM EGTA) with 4% formaldehyde and 0.1% Triton X-100. Larval carcasses were washed in PBT (phosphate buffered saline, 0.1% bovine serum albumin, 0.1% Triton X-100) and incubated with appropriate primary antibodies for 1 hr at room temperature. Subsequently, larval carcasses were washed in PBT, blocked with PBT containing 5% heat-inactivated goat serum, and incubated with fluorophore-conjugated secondary antibodies for 1 hr at room temperature. Wing imaginal discs were then washed, dissected out of the carcasses, and mounted in 50% glycerol, 0.1 M sodium carbonate pH 9, and PPDA (p-phenylene diamine, Sigma, Taufkirchen). Stained tissues were observed with a ZEISS Laser Scanning Microscope 510 Meta (Carl Zeiss AG, Jena). Primary antibodies used in this study were mouse anti-Fasciclin III 7G10 (1:100, Developmental Studies Hybridoma Bank), rabbit anti-GFP (1:2,000, Clontech 8372-1), and rat anti–DE-Cad (DCAD1 and DCAD2, 1:100; (Oda et al.,1994). Secondary antibodies used were Alexa-488-, Alexa-594- (Molecular Probes, Eugene, OR), and Cy5- (Jackson ImmunoResearch, West Grove, PA) conjugated anti-rabbit, anti-mouse, or anti-rat IgG (1:200).
Laser Scanning Microscopy of Live Wing Imaginal Discs and 3D Rendering
Wing imaginal discs were dissected in Ringer's solution from third-instar larvae and processed as previously described (Greco et al.,2001). Briefly, live wing imaginal discs were transferred to a microscope glass-slide with a chamber delimited by double-sided adhesive tape containing either Shields and Sang's M3 culture medium (Shields and Sang,1970) or Ringer's solution with 9 μM FM4-64 (Molecular Probes). Wing imaginal discs were oriented with the squamous epithelium facing the coverslip. The wing imaginal discs were observed immediately using a ZEISS LSM 510 laser scanning confocal microscope (Carl Zeiss AG), usually using a water immersion 63× objective. Stacks of closely spaced (0.05 μm) sections were taken starting from the squamous epithelium progressing toward the basolateral side of the columnar epithelial cells. Confocal sections were then exported as TIF files and assembled in a library for processing using the 3D rendering software Volocity v2.5 (Improvision, Lexington, MA). Quicktime movies and single snapshots were then obtained from the 3D rendering of the tissue. Cytonemes were not observed in our preparations as their identification requires a slight flattening of imaginal discs (Ramirez-Weber and Kornberg,1999; Hsiung et al.,2005). All images shown here refer to the pouch region of wing imaginal discs.
We thank Konrad Basler, Marcos González-Gaitán, Suzanne Eaton, Tom Kornberg, and the Bloomington Drosophila Stock Center for fly stocks; Tadashi Uemura and the Developmental Studies Hybridoma Bank for antibodies; and Denis Corbeil for the cDNA clone LD16666. We also thank Suzanne Eaton and Andrew Oates for critical comments on the manuscript. This work was supported by the Max Planck Society and a grant from the Deutsche Forschungsgemeinschaft (C.D.).