Antibodies were raised against the Drosophila EGF receptor homolog (DER) and used for immunohistochemical analyses of Drosophila embryos. We found that DER is localized in a wide array of embryonic tissues, displaying a dynamic pattern of expression. DER appears to be expressed in all cells at the cellular blastoderm and gastrula stages. In extended-germ-band embryos, it is found predominantly in the mesoderm and the head. Finally, in retracted-germ-band embryos, DER immunoreactivity is most pronounced at sites of somatic muscle attachments and along the ventral midline of the CNS. We have thus observed that DER is expressed in the diverse tissues which are affected in the DER faint little ball (flb) embryonic lethal phenotype. The different pattern and extent of expression in each tissue suggests that the disparate aspects of the flb phenotype may result from different mechanisms of DER function. To understand the basis for the CNS phenotype of DER/flb mutants, we have closely followed the collapse of the CNS in mutant embryos. Our observations on the evolution of the final CNS phenotype, in combination with the temporo-spatial pattern of appearance of DER in the ventral neuroepithelium, suggest that this receptor participates in the second phase of neuron– glia interactions, namely in stabilization of the ladder-like CNS scaffolding formed by outgrowth of pioneer axonal processes along the glial pre-pattern.

The faint little ball (flb) locus was first recognized in a screen for embryonic lethal loci on the second chromosome of Drosophila (Nusslein-Volhard et al. 1984). Cuticles of unhatched homozygous flb embryos display characteristics indicative of several, not obviously related, defects. In cuticles of flb embryos, ventral denticle bands are missing or reduced while dorsal hairs are unaffected, suggesting preferential involvement of the ventral epidermis. Absence of the head skeleton points to defects in the cephalic region. Finally, the rounded shape of the cuticle is a manifestation of the failure of the extended germ band to retract. Further characterization of flb embryos has indicated that initial normal formation of head structures is followed, beginning at germ band extension, by secondary deterioration in the anterior region (Schejter and Shilo, 1989). The use of antibodies to visualize the central nervous system has also demonstrated that the CNS of flb embryos that have failed to retract is collapsed along the midline, displaying fused longitudinal connectives and horizontal commissures (Schejter and Shilo, 1989). Taken together, these results suggest that the flb locus encodes an essential protein, which fulfills a wide range of functions during Drosophila embryogenesis.

The flb locus encodes the Drosophila homolog of the vertebrate EGF receptor (Schejter and Shilo, 1989; Price et al. 1989), a transmembrane tyrosine kinase receptor that we have previously characterized (Livneh et al. 1985; Schejter et al. 1986). As a tyrosine kinase, the Drosophila EGF receptor, termed DER, is a member of a diverse gene family highly represented in vertebrate species (Hunter and Cooper, 1985; Hunter, 1987; Yarden and Ullrich, 1988) and in Drosophila melanogaster (reviewed in Shilo, 1987). DER displays an equal degree of similarity to the vertebrate EGF receptor (Ullrich et al. 1984) and neu (Schechter et al. 1984; Bargmann etal. 1986) proteins. The DER protein possesses tyrosine kinase activity (Schejter and Shilo, 1989; Wides et al. 1990), demonstrating its functional relatedness to the vertebrate receptors. However, neither the ligand(s) that bind to it, nor the substrate(s) upon which it acts have been identified to date.

The structure of transmembrane tyrosine kinases suggests that they function in transmission of external signals into the cell. This appears to be the case for three Drosophila tyrosine kinase receptors. The sevenless protein is crucial for the reception of environmental cues by the eye imaginal disc cells destined to become R7 (Hafen et al. 1987). torso is required for the generation of terminal structures in the embryo (Sprenger et al. 1989). Finally, the DER/flp locus was recently demonstrated to be allelic to two additional known mutations, Ellipse (Baker and Rubin, 1989) and torpedo (Price et al. 1989), whose phenotypes suggest that specific developmental information has failed to pass between cells. Ellipse (Elp), is a dominant ‘gain of function’ mutation, which gives rise to eyes with a reduced number of ommatidia. The Elp phenotype suggests that, in the larval eye imaginal disc, the wildtype receptor is involved in the transmission of an inhibitory signal between cells, which determines the proper number of cells that will initiate differentiation as photoreceptor preclusters. Homozygous torpedo (top) females are sterile due to the development of ventralized eggs and egg shells. Analysis of the top mutation has demonstrated that the gene product is required in the somatic follicle cells surrounding the germ line-derived egg, but not in the egg itself (Schup-bach, 1987). DER thus appears to be involved in communication within the egg chamber responsible for determination of dorso-ventral polarity, specifically in the reception of signals from the egg by the follicle cells.

The roles of DER at later stages of development suggest that its earlier embryonic functions may also involve intercellular communication. The different tissues that are affected in the DER/flb phenotype suggest that the DER protein will be expressed in a diverse array of embryonic cells. This notion is supported by the in situ hybridization pattern to DER RNA during embryonic development, which shows that transcripts are present in most embryonic tissues (Schejter et al. 1986; Kammermeyer and Wadsworth, 1987). The pleio-tropic nature of the embryonic fib phenotype raises several questions. Do the defects in mutants arise from aberrant differentiation of embryonic tissues, or do they reflect cell death and tissue disintegration? Is there a different basis for each aspect of the phenotype? Does DER participate in the same signal transduction pathway in the various embryonic tissues, or do the disparate manifestations of the phenotype stem from subtle differences in the pathway, such as the recognition of divergent ligands and/or substrates by DER?

This paper describes the generation of anti-DER antibodies, and their utilization to monitor the localization of the DER protein during embryonic development. We show that the protein is expressed in a wide array of tissues that coincide with those affected in flb mutants. Our results provide support for the notion that DER participates in the transmission of developmental signals crucial to cell determination (in the ventral epidermis) and to cell viability (in the head and CNS). We show that in mutant flp embryos the CNS develops normally, and the severe CNS phenotype ultimately observed is the result of secondary collapse.

Fly lines

Fly strains bearing the Df(2R)PKl, which is deficient in the DER gene, and the JE1 mutant DER allele were provided by J. O’Donnell. A second allele, 2C82, was provided by C. Nusslein-Volhard and E. Wieschaus. Strain 242, with an ‘enhancer trap element’ (Wilson et al. 1989) inserted into the third chromosome, and strain 5704 with the P [ry, ftz/lacZ] element on the third chromosome (Hiromi et al. 1985) were both the kind gifts of Y. Hiromi and C. S. Goodman. These elements were crossed into the fly strains bearing a mutant DER allele (JE1 or 2C82) to yield strains that were balanced for the mutant DER allele on the second chromosome and homozygous for the transposable element on the third. When not otherwise specified, Canton-S flies were used.

Preparation of antibodies

Polyclonal antibodies were prepared against a trpE/DER fusion protein by inserting a restriction fragment from a DER cDNA clone into the appropriate pATH vector (Dieckmann and Tzagoloff, 1985). To generate a construct encoding amino acid residues 113 –404 of the DER extracellular domain, a DER cDNA clone was grown in methylation-deficient bacteria and the BclI – BamHI fragment, including the first cysteine-rich domain, was inserted into the pATH 1 BamHI site.

Following transformation of E. coli DH1 cells with the pATH fusion vector, the fusion protein was induced and partially purified by the following procedures. Typically, 1 ml of overnight starter cultures of transformed bacteria was added to 100ml M9 medium with 0.5% casamino acids, 10 μgml-1 thiamine, 0.2% glucose and ampicillin. After 2h shaking at 33°C under well-aerated conditions, bacteria were induced by the addition of freshly prepared beta-indoleacrylic acid (Sigma) dissolved in ethanol (5 μgmr-1 final concentration) and left shaking overnight. The following day, bacteria were spun down, resuspended in 2ml of phosphate-buffered saline (PBS) and lysed by boiling in a 1:1 suspension with sample buffer (50 mM tris, pH 7.0, 25% glycerol, 2% beta-mercaptoethanol, 2 % SDS and bromophenol blue). The cleared suspension was loaded onto preparative 10% SDS – polyacrylamide gels. The induced fusion protein was located by lightly staining the gels with Coomassie blue. The portion of the gel containing the protein of interest was excised and minced. The protein was eluted by diffusion into PBS 0.1% SDS for 48 h.

Inoculation of two guinea pigs with the Bcl-Bam fusion protein was initiated by injection of about 50 μg of protein (determined by the Bio-Rad Protein Assay) with complete Freund’s adjuvant into the foot pad. Two boosts of about the same amount of protein in incomplete Freund’s were given at one month intervals. Bleedings began one month after the initial injection.

Antisera were assayed for their ability to immunoprecipitate the DER protein that was produced by in vitro transcription and translation in the following manner. A plasmid (SP-DER) containing the full DER type II cDNA minigene construct under the control of the SP6 promoter was prepared by inserting the DER minigene from pDER (Wides et al. 1990) into the EcoRI site of pSP65 (Melton et al. 1984). The DER minigene was constructed by combining a cDNA sequence encoding 80bp of 5 ’ untranslated sequence, the type II 5 ’ alternative exon, and the three common exons following it, with a genomic fragment encoding the large 3 ’ exon.

RNA transcripts were produced with the Amersham SP6 in vitro transcription system using circular SP-DER template. In vitro translation of SP-DER RNA was performed by using 6 μl of micrococcal nuclease-treated rabbit reticulocyte lysate (kindly provided by A. Zilberstein) in a 25 μl mixture containing 200ng SP-DER RNA, 8mM phosphocreatine, 20 μgml-1 creatine kinase, 0.2 mM MgC12, 0.4 mM DTT, 20 mM Hepes, pH 7.6, 70 mM potassium acetate, 0.3 mM spermidine, and 20 μCi [35S]methionine (1115 Ci mmol-1, Amersham). Reactions were run for 60 min at 30°C followed by a 1mM cold methionine chase for 5 min. 1 μl of antiserum was incubated with 2.5 μl of in vitro translation product in 200 μ1 RIPA and immunoprecipitated as described below.

Before use in immunohistochemistry, the guinea pig antiserum that gave the highest initial staining signal was affinity-purified by passage over an Affigel 10 plus Affigel 15 column (Bio-Rad) in a ratio of 3:1, linked to bacterial proteins and an unrelated trpE fusion protein. The eluate was collected and further purified by absorption against 0 – 2 h fixed embryos.

Immunohistochemistry of whole mount embryos

For immunocytochemistry involving fluorescent-conjugated secondary antibodies, dechorionated embryos were fixed for 20 min in 3.7% paraformaldehyde in PEM (0.1M Pipes, 2MM EGTA, ImM MgSO4, pH6.95) saturated with an equal volume of heptane. Vitelline membranes were removed by replacing the aqueous layer with methanol and shaking vigorously. Embryos that ‘cracked’ out of their vitelline membranes and sank to the bottom of the tube were washed 3 times with methanol and 3 times with PBT (1 ×PBS, 1% BSA, 0.1 % Triton X-100). They were then blocked for 4h at room temperature with PBT containing 10 % BSA. After an additional wash with PBT, embryos were incubated overnight at 4°C with a 1:200 dilution (in PBT) of guinea pig anti-DER amino terminal antiserum, preabsorbed as described above.

The following day the embryos were washed 5 times quickly in PBT. They were then incubated for 2 h at room temperature with a 1:300 dilution of rhodamine-conjugated affinity-purified goat anti-guinea pig antibody (Jackson Immunoresearch Laboratories), which had been pre-absorbed against fixed embryos. In the case of double labeling with anti-HRP, fluorescein-conjugated goat anti-HRP (Cappel) was included at this time at a 1:300 dilution in PBT. The embryos were then washed as above and mounted in 90% glycerol, 10% PBS. They were viewed with a Nikon Microphot-FX microscope under fluorescence and phase optics. Pictures were taken with T-Max 400 film, with timed 0.5 min exposures for immunofluorescent pictures.

For immunocytochemistry involving peroxidase-conjugated secondary antibodies, dechorionated, devitellinized and washed embryos were blocked for 4 h in normal goat serum (50 μl NGS ml-1 in PBT). They were then washed and treated with antibody diluted in PBT plus NGS. Depending on the experiment, the primary antibody was guinea pig anti-DER (at 1:200 dilution), mouse monoclonal anti-BP104 (at 1:2 dilution, from A. Bieber, N. Patel and C. S. Goodman) or mouse monoclonal anti-beta-galactosidase (at 1:1000 dilution, Promega). The following day, after 5 washes, the embryos were blocked again for 0.5 h in NGS, as above, and then incubated for 2h in peroxidase-conjugated affinity-purified secondary antibody (Jackson Immunoresearch Laboratories), preabsorbed against fixed embryos and diluted 1:300 in PBT plus NGS. After washes as above, embryos were incubated for 10 min in 200 μl of PBT plus 100 μl of DAB (1mgml-1 DAB, Sigma, in PBT). 3 μl of a 3% solution of H2O2 was added and the reaction was allowed to proceed for 15 min. The embryos were then washed twice with PBT, dehydrated in ethanol and allowed to clear overnight in methyl salicylate. They were mounted the following day and viewed by Nomarski optics. Slides were taken with Ekta-chrome tungstenl60 film.

Immunolocalization of the DER protein in embryos

The embryonic DER/flb phenotype suggests that the protein plays diverse roles in embryogenesis. It was therefore of interest to follow the localization of DER during embryonic development. The embryonic expression of the DER gene product was studied by staining embryos with guinea pig antibodies raised against the extracellular domain of DER and preabsorbed against both nonspecific bacterial proteins and 0 – 2 h Drosophila embryos. We are confident that the immunoreactivity of the purified guinea pig anti-DER antibody is solely due to the DER gene product, because PK1 embryos, homozygous for a deficiency of the DER locus and identified as such by their faint little ball phenotype (Fig. 1), were devoid of any specific signal.

Fig. 1.

Immunohistochemical staining showing the specificity of the anti-DER antibody. (A) Phase contrast. (B) Immunofluorescent image of A. The figure shows a mixed population of wild-type and flb embryos, stained with a primary guinea pig anti-DER antibody, and a secondary rhodamine-conjugated goat anti-guinea pig antibody. (Unless otherwise stated, all embryos shown in Figs 2 – 4 were also stained in the same way.) The two wild-type embryos in the upper right corner are brightly immunofluorescent. The lower right embryo displays the faint little ball (flb) phenotype and does not stain. On the lower left, a cellular blastoderm stage embryo (bl) in which the phenotype due to DER deficiency is not yet apparent, is also not immunoreactive. Bar=50 μm. bl, blastoderm; flb, faint little ball.

Fig. 1.

Immunohistochemical staining showing the specificity of the anti-DER antibody. (A) Phase contrast. (B) Immunofluorescent image of A. The figure shows a mixed population of wild-type and flb embryos, stained with a primary guinea pig anti-DER antibody, and a secondary rhodamine-conjugated goat anti-guinea pig antibody. (Unless otherwise stated, all embryos shown in Figs 2 – 4 were also stained in the same way.) The two wild-type embryos in the upper right corner are brightly immunofluorescent. The lower right embryo displays the faint little ball (flb) phenotype and does not stain. On the lower left, a cellular blastoderm stage embryo (bl) in which the phenotype due to DER deficiency is not yet apparent, is also not immunoreactive. Bar=50 μm. bl, blastoderm; flb, faint little ball.

The guinea pig anti-DER antiserum did not stain unfertilized eggs or wild-type embryos that had not yet undergone cellularization (not shown). The DER gene product is first observed by immunostaining after cell formation. It is widely distributed throughout the cellular blastoderm in a manner that suggests that it is localized at the periphery of cells in the newly formed plasma membranes (not shown). The initial emergence of the epitopes recognized by this antiserum thus coincides with the first appearance of DER transcripts previously visualized by in situ hybridization (Schejter et al. 1986), and agrees with Northern analysis indicating that there is no maternal component of DER transcripts (Lev et al. 1985).

During gastrulation, DER continues to be expressed in all ectodermal epithelial cells (Fig. 2B). When embryos are observed at the focal plane of the yolk, rather than at the surface, the density of DER appears to be highest in the basal portions of these cells, including the basal membranes (Fig. 2C). While the pattern of DER ectodermal expression remains largely uniform at this stage, hints of banding can be discerned that are perhaps associated with the deepening cephalic furrow and anterior and posterior transverse folds (Fig. 2B). Especially intense immunofluorescence is observed in association with epithelia that take on a high columnar configuration such as the epithelium invaginating to form the posterior discoid plate, which carries the pole cells towards the dorsal side of the embryo and gives rise to the posterior midgut invagination (Fig. 2B). The pole cells themselves, however, do not appear to express the DER protein. At this stage, hazy immunofluorescence is also visible in the ventral part of the embryo above the ventral furrow (Fig. 2B).

Fig. 2.

DER expression in gastrulating embryos. In this figure and Fig. 3, anterior is to the left and posterior to the right. Dorsal is up and ventral down. (A) Phase contrast, note the pole cells at the posterior surface of the embryo.(B)Surface view. Individual cell outlines can be seen. (C)View at the level of the yolk, showing bright immunofluorescence in the high epithelium of the dorsal plate (dp) which carries the pole cells dorsally. Bar=50 μm. dp, dorsal plate.

Fig. 2.

DER expression in gastrulating embryos. In this figure and Fig. 3, anterior is to the left and posterior to the right. Dorsal is up and ventral down. (A) Phase contrast, note the pole cells at the posterior surface of the embryo.(B)Surface view. Individual cell outlines can be seen. (C)View at the level of the yolk, showing bright immunofluorescence in the high epithelium of the dorsal plate (dp) which carries the pole cells dorsally. Bar=50 μm. dp, dorsal plate.

In the germ-band-extended embryo, continued DER expression in the ectoderm is overshadowed by the bright immunofluorescence visualized in an internal layer of cells, which exhibits a clear pattern of metamery (Fig. 3A). In order to identify the cell layer that expresses DER, embryos were double stained with anti-DER antibodies followed by a rhodamine-conjugated secondary antibody (Fig. 3B) and fluorescein-conjugated anti-HRP antibodies (Fig. 3C). By comparison of Fig. 3B with Fig. 3C, it can be seen that the cell layer that is immunoreactive with the anti-DER antibody is internal to the neuronal layer recognized by the anti-HRP antibody (Jan and Jan, 1982). Therefore DER appears to be localized in the newly formed mesodermal cell layer which has invaginated during gastrulation and is becoming segmented at this stage.

Fig. 3.

DER expression in germ-band-elongated embryos. (A)Surface view showing metameric staining in the trunk and widespread expression in the head, in the mandibular bud (mb), procephalic lobe (pl) and clypeolabrum (cl). (B)Internal view of a different embryo. (C) Same embryo as in B, double stained with fluorescein-conjugated anti-HRP antibody. Note that the cell layer in the trunk that is recognized by the anti-DER antibody is internal to the cells expressing neural antigens. Bar=50 μm. cl, clypeolabrum; mb, mandibular bud; pl, procephalic lobe

Fig. 3.

DER expression in germ-band-elongated embryos. (A)Surface view showing metameric staining in the trunk and widespread expression in the head, in the mandibular bud (mb), procephalic lobe (pl) and clypeolabrum (cl). (B)Internal view of a different embryo. (C) Same embryo as in B, double stained with fluorescein-conjugated anti-HRP antibody. Note that the cell layer in the trunk that is recognized by the anti-DER antibody is internal to the cells expressing neural antigens. Bar=50 μm. cl, clypeolabrum; mb, mandibular bud; pl, procephalic lobe

In addition to staining of the trunk ectoderm and mesoderm, we always observed very intense immunofluorescence in the head region (Fig. 3A). This can be seen in the mandibular bud (immediately anterior to the cephalic furrow), the procephalic lobe, and, most prominently, in the clypeolabrum. The clypeolabrum is distinguished by its high cylindrical epidermis, unlike the flat cells of the rest of the procephalon, and is also the site where cephalic mesoderm is found. Since staining of the clypeolabrum is both peripheral and internal, it is likely that the immunoreactivity we observe in the head of the germ-band-elongated embryo is at least partly associated with the cephalic mesoderm.

The final embryonic pattern of DER expression becomes defined during germ band shortening and subsequent stages (Figs 4 and 5). The epidermis at the tip of the clypeolabrum is still intensely immunofluorescent (Fig. 4A). Also intensely immunofluorescent, but to a lesser degree, is the epithelium of the terminal portion of the hindgut (Fig. 4A), which is also of ectodermal origin. DER epidermal immunoreactivity in the thorax and abdomen becomes localized primarily to the segmental grooves. The epidermal staining tapers off on either side of the grooves. This periodic nature of staining with the anti-DER antibody results in what can be seen at the surface focal plane as arch-like immunofluorescence criss-crossing the face of the embryo (Fig. 4B).

Fig. 4.

DER expression in germ-band-shortened embryos. Anterior is to the left and posterior to the right. (A) Internal view, showing immunoreactivity in the clypeolabrum (cl) and hindgut (hg), along the outline of the gut (g), in the fat body (fb) and in somatic muscle attachments (ma). (B) Ventral view, showing staining of the segmental furrows (sf) causing a criss-crossing pattern across the embryo, and bright immunofluorescence along the midline (ml) of the CNS. Bar=50 μm. CNS, central nervous system; cl, clypeolabrum; fb, fat body; g, gut; hg, hindgut; ma, muscle attachments; ml, midline; sf, segmental furrow.

Fig. 4.

DER expression in germ-band-shortened embryos. Anterior is to the left and posterior to the right. (A) Internal view, showing immunoreactivity in the clypeolabrum (cl) and hindgut (hg), along the outline of the gut (g), in the fat body (fb) and in somatic muscle attachments (ma). (B) Ventral view, showing staining of the segmental furrows (sf) causing a criss-crossing pattern across the embryo, and bright immunofluorescence along the midline (ml) of the CNS. Bar=50 μm. CNS, central nervous system; cl, clypeolabrum; fb, fat body; g, gut; hg, hindgut; ma, muscle attachments; ml, midline; sf, segmental furrow.

Fig. 5.

DER expression in muscle attachments and the CNS. Higher magnification of DER expression in embryos slightly older than in Fig. 4, stained with the primary guinea pig anti-DER antibody, and a secondary HRP-conjugated goat anti-guinea pig antibody. (A) DER is present in sites of muscle attachment (ma) to the body wall, probably in tendon cells, but not in the somatic muscles (m). (B) DER is present along the midline (ml) of the CNS. No other immunoreactivity is observed in the CNS or PNS. Bar=10 μm. CNS, central nervous system; m, muscles; ma, muscle attachments; ml, midline; PNS, peripheral nervous system.

Fig. 5.

DER expression in muscle attachments and the CNS. Higher magnification of DER expression in embryos slightly older than in Fig. 4, stained with the primary guinea pig anti-DER antibody, and a secondary HRP-conjugated goat anti-guinea pig antibody. (A) DER is present in sites of muscle attachment (ma) to the body wall, probably in tendon cells, but not in the somatic muscles (m). (B) DER is present along the midline (ml) of the CNS. No other immunoreactivity is observed in the CNS or PNS. Bar=10 μm. CNS, central nervous system; m, muscles; ma, muscle attachments; ml, midline; PNS, peripheral nervous system.

Intimately related to the segmental furrows are the muscle apodemes (Figs 4A and 5A). These are the focal attachment sites in thoracic and abdominal segments of the intersegmental muscles. The formation of these somatic muscles occurs during stages 13 – 15 (Campos-Ortega and Hartenstein, 1985), by fusion of individual somatic mesodermal cells to form syncytia. Insertion of muscles at apodemes and subsequent stretching occurs at stages 16 – 17. DER expression at these muscle apodemes is striking. Localization of DER seems to be primarily in tendon cells at the ectodermal epithelial aspect of the apodemes. The focal planes of Figs 4A and 5A were chosen to emphasize the concentration of DER at these points of attachment.

Most splanchnic mesodermal derivatives appear to continue to express DER at this stage. Thus DER can be discerned in the fat body extending from the head to the gonads between the gut and somatic musculature (Fig. 4A). In addition, DER immunoreactivity is readily apparent outlining the contours of the alimentary canal from the pharynx in the head (including the pharyngeal muscles), through the terminal segment of the hindgut (Fig. 4A). This pattern of staining is likely to reflect the expression of DER by the visceral musculature which ensheathes the internal body structures. DER expression, however, appears to have been lost from at least some somatic mesoderm derivatives, for example from the intersegmental muscle cells which insert into the muscle attachments in the body wall, as clearly demonstrated in Fig. 5A.

A final feature of DER expression which becomes dramatically manifested in the germ-band-retracted embryos is its presence along the midline of the ventral median cord (Figs 4B and 5B). DER immunoreactivity is discontinuous along the ventral aspect of the embryo in the plane of the central nervous system (Fig. 5B). Staining of the midline with anti-DER antibodies is not associated with the cell bodies of neurons or related supportive cells which lie along the CNS midline (Fig. 5B). The discontinuous pattern of expression along with the absence of a clear periodic pattern suggest that DER immunoreactivity may be associated with the cell processes of a subset of midline glial cells. Comparison of the staining pattern of anti-DER antibody with that of anti-HRP in older Drosophila embryos confirms that the nervous system staining with the anti-DER antibody is strictly limited to midline staining, and this antibody does not recognize any other structures in the central or peripheral nervous system (not shown).

Further characterization of the collapsed CNS phenotype of DER/flb embryos

The restricted expression of DER in a subset of midline cells in the ventral nerve cord is in striking contrast to the wide range of CNS deformations observed in DER/flb mutant embryos. Using FITC-conjugated anti-HRP antibodies to visualize neuronal surfaces, we have previously shown that the segmentally reiterated ladder-like scaffolding of the central nervous system in late DER/flb mutant embryos is completely disorganized (Schejter and Shilo, 1989). The commissures that span each neuromere are absent or fused, and longitudinal axon tracts are often fused. Misrouting of segmental and intersegmental nerves is also observed.

In light of the restricted expression of DER in the ventral midline of wild-type germ-band-retracted embryos, we chose to closely follow the evolution of the mutant phenotype during formation of the CNS in embryos homozygous for a severe allele of DER (the JE1 allele). To do this, we used monoclonal antibody BP104 (which recognizes a cytoplasmic determinant in the early central nervous system and in all cells of the peripheral nervous system) to stain homozygous JE1 flies. Heterozygous, balanced, JE1 flies were selfcrossed and the homozygous offspring were identified at the germ-band-extended stage by the detachment of cells in the cephalic region and the appearance of abnormally pronounced parasegmental furrows. When the developing CNS of germ-band-extended JE1 homozygous embryos is compared with that of similarly staged wild-type embryos (Fig. 6A, 6B), no significant differences can be observed. In the DER mutant embryos, as in the wild type, the horizontal commissures span the ventral midline and join the longitudinal axon tracts. However, when older, germ-band-retracted age embryos are stained, fusion of the longitudinal connectives and loss of the horizontal commissures are apparent in the JE1 homozygotes (Fig. 6D). These observations demonstrate that the severe CNS phenotype in DER/flb embryos does not arise as the CNS scaffolding is being generated. Rather, in mutant embryos, the CNS appears to form normally and then to collapse at some later stage.

Fig. 6.

Evolution of the collapsed CNS phenotype and analysis of the midline neuroepithelium in DER/flb embryos. (A – D) Embryos stained with antibody BP104 against the nervous system, followed by an HRP-conjugated secondary antibody. A and C are wild type embryos. B and D are DER/flb embryos, recognizeable by the deteriorating cells in the cephalic region (arrowheads). A and B are germ-band-extended embryos. C and D are germ-band-retracted embryos. (The peripheral nervous system is also visualized in D because the embryo is slightly older than in C.) (E) Wild type (on the right) and DER/flb (on the left) embryos bearing an ‘enhancer trap element’ which allows visualization of the midline neuroepithelium with anti-beta-galactosidase. (F) A DER/flb germ-band-extended embryo bearing a P element directing beta-galactosidase expression under the flz promoter, beta-galactosidase was visualized with anti-beta-galactosidase antibody and can be seen in the seven stripes typical of flz expression and in the paired MP2 cells (arrowhead) which also express flz- (In E and F the DER/flb mutant embryos were identified by anteriorly detaching cells and abnormally pronounced parasegmental furrows observed at a different focal plane.) Bar is the same for A – D and F, and different for E. Bar=50 μm.

Fig. 6.

Evolution of the collapsed CNS phenotype and analysis of the midline neuroepithelium in DER/flb embryos. (A – D) Embryos stained with antibody BP104 against the nervous system, followed by an HRP-conjugated secondary antibody. A and C are wild type embryos. B and D are DER/flb embryos, recognizeable by the deteriorating cells in the cephalic region (arrowheads). A and B are germ-band-extended embryos. C and D are germ-band-retracted embryos. (The peripheral nervous system is also visualized in D because the embryo is slightly older than in C.) (E) Wild type (on the right) and DER/flb (on the left) embryos bearing an ‘enhancer trap element’ which allows visualization of the midline neuroepithelium with anti-beta-galactosidase. (F) A DER/flb germ-band-extended embryo bearing a P element directing beta-galactosidase expression under the flz promoter, beta-galactosidase was visualized with anti-beta-galactosidase antibody and can be seen in the seven stripes typical of flz expression and in the paired MP2 cells (arrowhead) which also express flz- (In E and F the DER/flb mutant embryos were identified by anteriorly detaching cells and abnormally pronounced parasegmental furrows observed at a different focal plane.) Bar is the same for A – D and F, and different for E. Bar=50 μm.

In order to examine DER mutants for the presence and disposition of various mesectodermally derived midline cell types, two approaches were taken. In one approach, we used an ‘enhancer trap’ element which has inserted into the third chromosome in a position where endogenous enhancer elements direct the midline-specific transcription of the lacZ gene under a minimal promoter. In self-crosses of flies balanced for the JE1 DER allele on the second chromosome and homozygous for the introduced enhancer trap element on the third chromosome, one fourth of the progeny are homozygous DER mutants (in addition to bearing the enhancer trap element). When the pattern of betagalactosidase expression is followed by antibody staining of JE1 homozygous embryos, it is found to differ only slightly from the pattern in normal embryos (JE1 heterozygous siblings). While in normal germ-band-extended embryos beta-galactosidase staining is detected in a continuous band along the ventral midline encompassing a restricted set of identified and unidentified neuronal and non-neuronal cell types (Fig. 6E, right hand embryo), staining of DER/flb mutants reveals a similar pattern with only a number of discontinuities in the midline neuroepithelial band (Fig. 6E, left-hand embryo). These results indicate that the abnormalities are minimal in the midline cell population of mutant DER germ-band-extended embryos.

In the second approach that was taken, the fate of one specific pair of neurons situated on either side of the midline, the MP2 cells, was followed. Since ftz is expressed in these cells (Doe et al. 1988), their fate could be monitored in a manner similar to that described above by self-crossing flies heterozygous for the 2C82 DER allele and homozygous for a third chromosome element directing beta-galactosidase expression under the ftz promoter (Hiromi et al. 1985). In DER/flb mutant germ-band-extended embryos, each of the two bilaterally paired MP2 cells of a single neuromere are located at some distance from one another, on either side of the midline (Fig. 6F). These results again demonstrate that midline structures are essentially intact in the germ-band-extended DER/flb embryo and no extensive collapse which would bring together paired juxta-midline cells has occurred. We therefore suggest that the final CNS phenotype observed in DER homozygous embryos is due to secondary collapse of the CNS superstructure after initial normal generation of horizontal commissures and longitudinal connectives.

Localization of DER in embryos

We have been able to follow the distribution of the DER protein throughout Drosophila embryogenesis at a high level of resolution using an anti-DER antibody produced in guinea pig against an extracellular segment of DER. In general, our results are similar to previous in situ analyses that indicated that DER transcripts are widely distributed in the embryo. Specifically, we have shown that the initial homogeneous distribution of the DER protein at the cellular blastoderm stage gives way, as development proceeds, to subtle regional differences at ventral furrow formation, and a more pronounced localization to several discrete cell layers and tissues in older embryos.

Since the DER protein is a transmembrane protein with homology to known tyrosine kinase receptors, its first appearance after cellularization of the blastoderm, and apparent localization by immunocytochemistry at cell membranes, is not surprising. Expression of DER by all ectodermal cells persists throughout gastrulation and germ band elongation, but eventually resolves into differential expression at the tip of the clypeolabrum (where it appears to be present at high density from germ band elongation to the end of embryogenesis), at the terminal portion of the hindgut, and in the segmental grooves of the epidermis, particularly at the sites of segmental somatic muscle attachments. The dramatic localization of DER to these muscle apodemes puts it into the same category as the Drosophila PSI and PS2 integrins (Bogaert et al. 1987; Leptin et al. 1989) and the vertebrate fibronectin receptor (Neff et al. 1982), which are found in the adhesion plaques that characterize myoepidermal interactions. In its localization to newly invaginated mesoderm, DER is more similar to PS2, but its ultimate localization in the ectoderm at the epidermal aspect of muscle apodemes is highly reminiscent of the pattern of PSI expression. Since a basal lamina is present at the muscle – epidermal interface of insects, localization of DER at these sites suggests a possible interaction between DER and some extracellular matrix ligand. This notion is also supported by what appears to be the basal location of DER in ectoderm which is more obvious at earlier developmental stages (see Fig. 2C).

In addition to expression in some ectodermal derivatives, DER is found in the mesodermal cell layer of both the trunk and the head of the germ-band-extended embryo. Expression of DER does not persist in all mesodermal derivatives. It does appear to be present in the fat body and in the visceral musculature which forms a mantle around the foregut, midgut and hindgut of germ-band-retracted embryos. DER immunoreactivity around the contours of the gut is absent in twist embryos (unpublished results) in which mesodermal invagination fails to occur (Simpson, 1983; Nusslein-Volhard et al. 1984), and can thus clearly be attributed to the mesodermal, rather than endodermal, component of the gut wall. On the other hand, DER is clearly absent from the segmental somatic muscles which insert at focal attachments in the body wall.

At germ band retraction, prominent midline DER immunoreactivity is observed in the vicinity of mesectodermally derived glial cells. This is the only nervous system-associated staining that we observe with the anti-DER antibody. There is no staining in the neural cell bodies of the CNS or in the peripheral nervous system. A puzzling aspect of DER expression in the midline is the absence of any clear periodicity associated with its discontinuous pattern of expression. This feature may be related to the fact that, at high magnification, DER does not appear to be localized in cell bodies and may thus be localized exclusively in processes of a subset of the midline glial cells. Good candidates for the cells whose processes express DER are the six paired midline glial cells (two anterior, two middle and two posterior) which first serve as a blueprint for guiding outgrowth of pioneer axons along the future horizontal commissures and then themselves send out processes that ensheathe the axonal tracts and function in neuronal support (Jacobs and Goodman, 1989; Fredieu and Mahowald, 1989). This tentative observation will have to be resolved by electron immunomicroscopy.

Correlations between the localization of DER and the embryonic fib phenotype

The faint little ball embryonic phenotype of amorphic mutants in DER is complex and involves many structures. The most prominent defects can be grouped into four categories: cephalic defects, ventral cuticle abnormalities, failure of the germ band to retract and CNS disorganization. For every class of defects, it is important to ask how the DER phenotype relates to the distribution of the DER protein that we have described. In other words, does the localization of the DER protein in embryos help to understand the fib mutant phenotype? We must keep in mind, however, that definitive analysis of the anatomical domains in which DER functions during embryogenesis must await identification and localization of the ligands and substrates that interact with this receptor, since DER activity can only take place where the distributions of all three components overlap.

The cephalic defect is the earliest indication of the fib phenotype. It is manifested at the germ-band-extended stage in amorphic DER/flb homozygous embryos as slight anterior deterioration and retraction. This defect eventually develops into pronounced disintegration characterized by widespread cell detachment. The fib phenotype is first evident at a time of intense cephalic expression of DER in the anterior ventral mandibular region, in the clypeolabrum and procephalic lobe. The extent of DER expression appears to be roughly coincident with the scope of the cephalic phenotype. These observations are consistent with the idea that DER is important for maintenance of tissue integrity in most embryonic head structures.

Because of the massive anterior disintegration, head cuticular structures are absent in the cuticle that is formed by mutant DER embryos. In addition, the segmental cuticular ventral setae are absent or much reduced while the dorsal and dorsolateral hairs are normal. The late epidermal defect which results in these cuticular abnormalities does not appear to result from massive loss or death of the ventral epidermis. When antibodies that recognize different subsets of epidermal cells in the extended germ band stage were used to probe JE1 (a severe DER/flb allele) homozygous embryos, we noted that each of the specific subpopulations that express engrailed, ftz, and wingless are normally disposed in the flb embryo (Schejter and Shilo, 1989, and unpublished). By this criterion, cell viability and commitment to particular cell fates appears to be normal in the epidermis through germ band extension.

In order to understand further the basis for the ventral cuticle defects, it was of interest to follow the integrity of DER-expressing cells in the ventral region of mutant DER embryos. For this purpose we again chose the JE1 allele which was shown to give rise to a severe flb phenotype (similar to that of a null allele) in spite of the fact that it expresses a full-length DER protein which can be visualized by an in vitro kinase assay (Schejter and Shilo, 1989). Immunohistochemical analysis of JE1 homozygous embryos reveals that DER expressing cells are present in the epidermis and mesoderm at the germ-band-extended stage and later on (data not shown). The basis for the cuticular aspect of the DER phenotype is thus likely to be different from the head defects in that it does not involve large-scale loss of cells.

At the time when secretion of cuticular structures by the epidermis is initiated at stage 16, DER epidermal expression is limited largely to the segmental furrows. This is in contrast to earlier stages in which it is found throughout the epidermis. Our interpretation of these data is that a functional receptor is necessary for transduction of the signal for determination or maintenance of the determined cell fate in the ventral epidermis of the thorax and abdomen. Although the effects of DER’s influence are manifested only at the end of embryogenesis, when the determined epidermal cells execute their differentiated role, DER is likely to be acting at an earlier stage, prior to synthesis of cuticle. This allows us to attempt to define the ‘developmental window’ during which DER acts in the epidermis. This window is likely to be from late germ band extension (i.e. after cell determination to express the pair rule and segment polarity proteins we monitored) until the time that DER becomes localized to segmental furrows and is unlikely to be affecting the entire ventral epidermis.

Although DER is also expressed by the dorsal epidermis, no phenotype is observed in dorsal cuticular structures, even in embryos homozygous for null alleles. This may indicate that DER has no function in commitment to elaboration of dorsal cuticle in spite of its expression in that region, or it may be indicative of redundancy in the signals for the development of this domain.

The third aspect of the amorphic DER phenotype is the failure of the extended germ band to retract. The mechanism of germ band retraction is unknown. One hypothesis is that it may involve changes in cell shape (Campos-Ortega and Hartenstein, 1985). In light of the striking localization of DER to the newly invaginated mesoderm in the extended-germ-band embryo, it seems possible that mesodermally expressed DER may play some role in this major morphogenetic event. We have not been able to identify any other characteristics flb embryos that seem likely to stem from expression of DER in mesoderm or some of the mesodermal derivatives. However, it is possible that some late characteristics of the DER phenotype are masked by earlier defects and embryonic death.

By monitoring the development of the CNS in flb embryos with the BP104 antibody, we have shown that the initial near-normal formation of the CNS which is seen in younger flb embryos is followed by collapse of the ladder-like scaffolding along the ventral midline in embryos that have failed to shorten. As described above, DER is prominently distributed along the midline of the CNS from germ band retraction, thus closely paralleling in its localization and time of its appearance the dramatic buckling of the nervous system along the ventral axis seen in mutant embryos of this age. Both the configuration of the structures expressing DER (i.e. DER is not found in the round pattern typical of expression in cell bodies) and lack of periodicity of these structures suggest that the receptor is present on the cell processes of a subset of midline cells. These may be the three pairs of midline glial cells, which function first in generation of the commissural neuronal superstructure by guiding the outgrowth of pioneer axons and then in support of these axons which they come to ensheathe (Jacobs and Goodman, 1989; Fredieu and Mahowald, 1989). Taken together, these observations raise the interesting possibility that DER is a receptor that is involved only in the second phase of neuron-glia interactions, those interactions between neurons and the surrounding glial processes which are necessary for maintenance of proper neuronal structure and function. These interactions may be mediated by a distinct set of surface molecules from the various cell adhesion and recognition molecules which are thought to act in initial axon outgrowth. The ongoing axon – glia interactions play a role in nutritional support, electrical isolation and other aspects of neuronal physiology.

The extreme degree of disorganization of the nervous system in severe flb embryos that are of germ-band-retracted age (when DER appears in the neuroepithelium) has not allowed us to ascertain whether or not DER protein is present along the ventral midline of mutant embryos. However, we do know that at least some midline neuroepithelial cells are largely unaffected in DER mutant germ-band-extended embryos. These include the median neuroblast and its progeny, which are recognized by anti-engrailed antibodies in mutant embryos (Schejter and Shilo, 1989), the juxta-midline MP2 precursor cells in which the expression of ftz was followed with a ftz-lacZ construct (this work), and the entire midline strip of neuroepithelial cells in which we monitored the appearance of beta-galactosidase driven by an enhancer trap element (this work).

Recently, mutations in two other genes, singleminded (sim) (Crews et al. 1988) and slit (Rothberg et al. 1988), have been shown to give rise to a collapsed nervous system phenotype that is similar to what is observed in flb embryos. Both genes encode proteins that are specifically expressed along the midline of the CNS and eventually become restricted predominantly to the three pairs of midline ectodermal cells, the specialized glial cells which we also believe to be good candidates for the cells with DER-positive processes. Interestingly, the appearance of DER in these cells occurs later than that of slit, which in turn lags behind the expression of sim, suggesting that the three loci function at different times during midline development. While it is tempting to postulate direct or indirect interactions among the three loci (especially in light of the EGF-like motifs that are present in slit), there is as yet no evidence to support this notion. Indeed, our observations of the initial normal emergence of the CNS scaffolding in DER mutants and the separation of the bilaterally paired MP2 cells in an flb background (in contrast to sim embryos in which the two MP2 cells are always situated adjacent to one another) suggest that DER may be acting in a different developmental pathway. We therefore believe that DER functions later than sim and slit and participates in stabilization, rather than generation, of the CNS superstructure through supportive neuronal – glial interactions.

Comparison of the embryonic localization of DER with different aspects of the severe flb phenotype leads us to the conclusion that the pleiotropic nature of the phenotype reflects multiple functions that DER appears to carry out in different tissue types during embryogenesis. Surprisingly, one process in which DER does not appear to take a significant part is proliferation. This has already been suggested on the basis of the DER/flb phenotype which becomes apparent only after most cells of ectodermal origin have ceased to divide (Schejter and Shilo, 1989). The immunohistochemical staining described here offers additional indications for the later functions of DER in transmission of signals essential for maintenance of tissue integrity of cells actually expressing the protein (in the cephalic region), in reading of developmental cues that lead to the eventual execution of differentiated cell functions (in the ventral epidermis), and in stabilization of the CNS superstructure, probably by participation in supportive neuron – glia interactions. Whether these diverse roles of DER employ the same signal transduction pathway at the ligand and substrate ends is a crucial question, which should be addressed by further biochemical and genetic analyses.

The disparate roles of DER during embryogenesis discussed here are only a subset of its biological functions. DER transcripts were also shown to be present at later stages of development. In addition, genetic analysis has revealed that the Ellipse and torpedo loci are also allelic to DER, implicating DER’s central role in morphogenesis of the compound eye and the structure of the egg shell and embryo. The availability of antibodies recognizing the DER protein should allow us in the future to follow the distribution of the protein at later stages of development and ask how it pertains to these mutant phenotypes.

This work was supported by grants from NIH (GM35998) and the Israel Cancer Research Fund to B.S. N. B. Z. was supported by grants from the Israeli Ministry of Absorption and the Israel Cancer Association. R. J. W. holds an NIH-NC1 postdoctoral fellowship (CA08501).

Baker
,
N. E.
and
Rubin
,
G. M.
(
1989
).
Effect on eye development of dominant mutations in the Drosophila homologue of the EGF-receptor
.
Nature
340
,
150
153
.
Bargmann
,
C. I.
,
Hung
,
M.-C.
and
Weinberg
,
R. A.
(
1986
).
The neu oncogene encodes and epidermal growth factor receptor-related protein
.
Nature
319
,
226
230
.
Bogaert
,
T.
,
Brown
,
N.
and
Wilcox
,
M.
(
1987
).
The Drosophila PS2 antigen is an invertebrate integrin that, like the fibronectin receptor, becomes localized to muscle attachments
.
Cell
51
,
929
940
.
Campos-Ortega
,
J. A.
and
Hartentstein
,
V.
(
1985
).
The Embryonic Development of Drosophila melanogaster. SpringerVerlag
(
Berlin, Heidelberg, New York, Tokyo
).
Crews
,
T. C.
,
Thomas
,
J. B.
and
Goodman
,
C. S.
(
1988
).
The Drosophila single-minded gene encodes a nuclear protein with sequence similarity to the per gene product
.
Cell
52
,
143
151
.
Dieckmann
,
C. L.
and
Tzagoloff
,
A.
(
1985
).
Assembly of the mitochondrial membrane system. CBP6, a yeast nuclear gene necessary for synthesis of cytochrome b
.
J. biol. Chem
.
260
,
1513
1520
.
Doe
,
C. Q.
,
Hiromi
,
Y.
,
Gehring
,
W. J.
and
Goodman
,
C. S.
(
1988
).
Expression and function of the segmentation gene fushi tarazu during Drosophila neurogenesis
.
Science
239
,
170
175
.
Fredieu
,
J. R.
and
Mahowald
,
A. P.
(
1989
).
Glial interactions with neurons during Drosophila embryogenesis
.
Development
106
,
749
757
.
Hafen
,
E.
,
Basler
,
K.
,
Edstroem
,
J-E.
and
Rubin
,
G. M.
(
1987
).
Sevenless, a cell specific homeotic geno of Drosophila, encodes a putative transmembrane receptor with a tyrosine kinase domain
.
Science
236
,
55
63
.
Hiromi
,
Y.
,
Atsushi
,
K.
and
Gehring
,
W. J.
(
1985
).
Control elements of the Drosophila segmentation gene fushi tarazu
.
Cell
43
,
603
613
.
Hunter
,
T.
(
1987
).
A thousand and one protein kinases
.
Cell
50
,
823
829
.
Hunter
,
T.
and
Cooper
,
J. A.
(
1985
).
Protein-tyrosine kinases
.
Ann. Rev. Biochem
.
54
,
897
930
.
Jacobs
,
J. R.
and
Goodman
,
C. S.
(
1989
).
Embryonic development of axon pathways in the Drosophila CNS. I. A glial scaffold appears before the first growth cones
.
J. Neurosci
.
9
,
2402
2411
.
Jan
,
L. Y.
and
Jan
,
Y. N.
(
1982
).
Antibodies to horseradish peroxidase as specific neuronal markers in Drosophila and grasshopper embryos
.
Proc. natn. Acad. Sci. U.S.A
.
79
,
2700
2704
.
Kammermayer
,
K. L.
and
Wadsworth
,
S. C.
(
1987
).
Expression of Drosophila epidermal growth factor receptor homologue in mitotic cell populations
.
Development
100
,
201
210
.
Leptin
,
M.
,
Bogart
,
T.
,
Lehmann
,
R.
and
Wilcox
,
M.
(
1989
).
The function of PS integrins during Drosophila embryogenesis
.
Cell
56
,
401
408
.
Lev
,
Z.
,
Shilo
,
B.
and
Kimchie
,
Z.
(
1985
).
Developmental changes in expression of the Drosophila melanogaster epidermal growth factor receptor gene
.
Devi Biol
.
110
,
499
502
.
Livneh
,
E.
,
Glazer
,
L.
,
Segal
,
D.
,
Schlessinger
,
J.
and
Shilo
,
B.
(
1985
).
The Drosophila EGF receptor homolog: conservation of both hormone binding and kinase domains
.
Cell
40
,
599
607
.
Melton
,
D. A.
,
Krieg
,
P. A.
,
Rebagliati
,
M. R.
,
Maniatis
,
T.
,
Zinn
,
K.
and
Green
,
M. R.
(
1984
).
Efficient in vitro synthesis of biologically active RNA and RNA hybridization probes from plasmids containing bacteriophage SP6 promoter
.
Nuc. Acid Res
.
12
,
7035
7056
.
Neff
,
N. T.
,
Lowrey
,
C.
,
Decker
,
C.
,
Tovar
,
A.
,
Damsky
,
C.
,
Buck
,
C.
and
Horwitz
,
A. F.
(
1982
).
A monoclonal antibody detaches embryonic skeletal muscle from extracellular matrices
.
J. Cell Biol
.
95
,
654
666
.
Nusslein-Volhard
,
C.
,
Wieschaus
,
E.
and
Kluding
,
H.
(
1984
).
Mutations affecting the pattern of the larval cuticle in Drosophila melanogaster. I. Zygotic loci on the second chromosome
.
Wilhelm Roux’s Arch, devl Biol
.
193
,
267
282
.
Price
,
J. V.
,
Clifford
,
R. J.
and
Shupbach
,
T.
(
1989
).
The maternal ventralizing locus torpedo is allelic to faint little ball, an embryonic lethal, and encodes the Drosophila EGF receptor homolog
.
Cell
56
,
1085
1092
.
Rothberg
,
J. M.
,
Hartley
,
D. A.
,
Walther
,
Z.
and
Artavonis-Tsakonas
,
S.
(
1988
).
slit: an EGF-homologous locus of Drosophila melanogaster involved in the development of the central nervous system
.
Cell
55
,
1047
1059
.
Schechter
,
A. L.
,
Stern
,
D. F.
,
Vaidyanathan
,
L.
,
Decker
,
S. J.
,
Drebin
,
J. A.
,
Greene
,
M. I.
and
Weinberg
,
R. A.
(
1984
).
The neu oncogene: an er6-B-related gene encoding a 185.000-Mr tumor antigen
.
Nature
312
,
513
516
.
Schejter
,
E. D.
,
Segal
,
D.
,
Glazer
,
L.
and
Shilo
,
B.
(
1986
).
Alternative 5’ exons and tissue-specific expression of the Drosophila EGF receptor homolog transcripts
.
Cell
46
,
1091
1101
.
Scheffer
,
E. D.
and
Shilo
,
B-Z.
(
1989
).
The Drosophila EGF receptor homolog (DER) is allelic to faint little ball, a locus essential for embryonic development
.
Cell
56
,
1093
1104
.
Schupbach
,
T.
(
1987
).
Germ line and soma cooperate during oogenesis to establish the dorsoventral pattern of egg shell and embryo in Drosophila melanogaster
.
Cell
49
,
699
707
.
Shilo
,
B-Z.
(
1987
).
Proto-oncogenes in Drosophila melanogaster
.
Trends Genet
.
3
,
69
-
72
.
Simpson
,
P.
(
1983
).
Maternal-zygotic interactions during formation of the dorsoventral pattern in Drosophila embryos
.
Genetics
105
,
615
632
.
Sprenger
,
F.
,
Stevens
,
L. M.
and
Nusslein-Volhard
,
C.
(
1989
).
The Drosophila gene torso encodes a putative receptor tyrosine kinase
.
Nature
338
,
478
—483.
Ullrich
,
A.
,
Coussens
,
L.
,
Hayflick
,
J. S.
,
Dull
,
T. J.
,
Gray
,
A.
,
Tam
,
A. W.
,
Lee
,
J.
,
Yarden
,
Y.
,
Libermann
,
T. A.
,
Schlessinger
,
J.
,
Downward
,
J.
,
Mayes
,
E. L. V.
,
Whittle
,
N.
,
Waterfield
,
M. D.
and
Seeburg
,
P. H.
(
1984
).
Human epidermal growth factor receptor cDNA sequence and aberrant expression of the amplified gene in A431 epidermoid carcinoma cells
.
Nature
309
,
418
425
.
Wides
,
R.
,
Zak
,
N. B.
and
Shilo
,
B-Z.
(
1990
).
Enhancement of tyrosine kinase activity of the Drosophila EGF receptor homolog (DER) by alterations of the transmembrane domain
.
Eur. J. Biochem., in press
.
Wilson
,
C.
,
Pearson
,
R. K.
,
Bellen
,
H. J.
,
O’kane
,
C. J.
,
Grossniklaus
,
U.
and
Gehring
,
W. J.
(
1989
).
P-element mediated enhancer detection: an efficient method for isolating and characterizing developmentally regulated genes in Drosophila
.
Genes and Development
3
,
1301
1313
.
Yarden
,
Y.
and
Ullrich
,
A.
(
1988
).
Growth factor receptor kinases
.
Ann. Rev. Biochem
.
57
,
443
478
.