QUICK SEARCH:   [advanced] Author: Keyword(s):
Home     Help     Feedback     Subscriptions     Archive     Search     Table of Contents

First published online 17 March 2004
doi: 10.1242/dev.01058

This Article Summary Full Text Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Urban, S. Articles by Freeman, M. Search for Related Content PubMed PubMed Citation Articles by Urban, S. Articles by Freeman, M.

EGF receptor signalling protects smooth-cuticle cells from apoptosis during Drosophila ventral epidermis development

Sinisa Urban^*, Gemma Brown and Matthew Freeman

MRC Laboratory of Molecular Biology, Hills Road, Cambridge CB2 2QH, UK

View larger version (44K): [in a new window]   Fig. 1. Analysis of the denticle belt fusion phenotype (for example, see arrow in A) of spitz group mutants. (A) Penetrance (percentage of embryos having at least one fusion) of the denticle belt fusion phenotype of null mutations in different spitz group genes. The denticle belts of each parasegment are labelled to the right of the wild-type cuticle for reference. Expression of a dominant-negative form of the EGFR throughout the epidermis resulted in denticle belt fusions, confirming that this phenotype was indeed the result of reduced EGFR signalling. Note that single minded, while being a member of the spitz group, is not an EGFR component: consistent with this, mutants do not show denticle fusions. (B) Distribution of denticle belt fusions observed in each parasegment (analysis of a rhomboid-1 mutant is shown but other spitz group genes showed similar spectra). Denticle belt fusions were observed in all parasegments, but occurred most frequently between the T3 and A1, A4 and 5, and A5 and 6 denticle belts.

View larger version (58K): [in a new window]   Fig. 4. Expressivity of the denticle belt fusion phenotype in different spitz group mutants. Expressivity reflects the severity of the denticle belt fusion phenotype, and is measured as the number of fusions per embryo. Note that one denticle belt fusion results from the midline joining of two denticle belts; as such, a total of ten fusions are possible in a single embryo since there are a total of 11 (three thoracic and eight abdominal) denticle belts. (A) An example of a spitz null embryo displaying four denticle belt fusions. (B) The expressivity of various spitz group genes was analysed, which revealed that most mutant embryos contained between one and three fusions, but never more than five of the 10 possible fusions. (C) rhomboid-1 vein double mutant embryos are more severely affected than spitz group embryos, and only one in three could be analysed for ventral epidermal patterning. (D) A rhomboid-1 vein double mutant dramatically increased the expressivity (but not penetrance) of the fusion phenotype compared to that of rhomboid-1 alone.

View larger version (55K): [in a new window]   Fig. 5. A requirement for EGFR signalling in smooth-cuticle cells. (A) wg-gal4 and prd-gal4 drivers were used to analyse the fusion phenotype; their expression patterns were visualized using RNA in situ hybridisation. (B) DN-EGFR was expressed in smooth-cuticle cells in the regions of the embryos indicated by the red lines using arm-gal4, prd-gal4 and wg-gal4 drivers. Note that the prd-gal4 and wg-gal4 drivers are expressed at higher levels per cell than the arm-gal4 driver. Expression of DN-EGFR in alternating smooth-cuticle domains resulted in strong denticle belt fusions in essentially all paired domains, and never in non-paired parasegments. The wg-gal4 driver expressed only in two cell rows in each stripe, but incomplete fusions were evident (far right). (C) The specificity of activating and inhibiting EGFR signalling on the fusion phenotype was tested by modifying the fusion phenotype of rhomboid-3 rhomboid-1 double mutant (ru1rho17M43) embryos. Note that the analysis was performed only in paired domains (alternating smooth-cuticle stripes), and that these weak EGFR, ras and raf transgenes caused no phenotypes when expressed by themselves in wild-type embryos using prd-gal4. Transgenes were expressed at 25°C (graph on left), except rasN17 and DN-raf, which were expressed at 29°C (graph on right). Note that at 29°C fewer denticle belt fusions occurred in rhomboid-3 rhomboid-1 double mutant embryos. (D) Although the width of the medial fusion varied in spitz group embryos, expressing DN-EGFR laterally in smooth-cuticle cells along the entire circumference of the parasegment (see A above) resulted in fusions that were never wider than the middle two thirds of denticle belts.

View larger version (43K): [in a new window]   Fig. 2. Analysis of rhomboid function during embryogenesis using RNA interference. (A) High concentrations of rhomboid-1 dsRNA were important in maximizing penetrance, with 2 mg/ml dsRNA yielding most penetrant phenocopies. A-C show the percentage of injected embryos that hatched. Essentially all rhomboid-1-injected embryos that did not hatch had missing row 1 denticles, polarity reversal of row 4 denticles, and occasional belt fusions characteristic of the spitz group denticle phenotype. (B) Although dsRNA corresponding to both rhomboid-1 and sprouty caused lethality with strong phenotypic effects, none of the remaining rhomboids yielded any discernible phenotypes when inactivated in wild-type embryos. UN, uninjected embryos; buf, those injected with buffer only. (C) Injection of dsRNAs corresponding to rhomboid-2 and -4 into rhomboid-3 null embryos did not produce lethality, or discernible modification of the denticle phenotype resulting from co-injecting rhomboid-1 dsRNA. Note that co-injecting rhomboid-1 dsRNA with other dsRNAs resulted in decreased lethality because it reduced the concentration of rhomboid-1 dsRNA. Injection of rhomboid-1, or rhomboid-3 rhomboid-1 double mutant embryos resulted in 25% lethality because of the lethal phenotype of mutated rhomboid-1 (present in 25% of embryos derived from a cross between heterozygous parents), but no enhancement of the spitz group denticle phenotype was evident when rhomboid-2 and -4 dsRNA was injected (not shown). (D) Phenotypic analysis of wild-type embryos injected with rhomboid-1 dsRNA (white bars), rhomboid-3 null embryos injected with rhomboid-1 dsRNA (grey bars), and rhomboid-3 null embryos injected with only buffer (black bars). Only embryos injected with buffer displayed wild-type cuticles, while embryos injected with rhomboid-1 dsRNA displayed spitz group cuticle phenotypes. Strikingly, injection of rhomboid-3 null embryos with rhomboid-1 dsRNA produced a marked increase in denticle belt fusions compared to wild-type embryos injected with rhomboid-1 dsRNA.

View larger version (92K): [in a new window]   Fig. 3. Ventral narrowing of rhomboid mutants. (A) The ventral cuticle of thoracic parasegments from a wild-type embryo is shown for reference. The Keilin's organs (KO) and ventral black dots (vbd) are indicated with arrowheads, and the denticle belts are labelled to the left of the image. (B) The distance between KOs and vbds was measured in arbitrary units. This analysis served as a measure of ventral narrowing and thus of ventrolateral specification (Mayer and Nusslein-Volhard, 1988). Mutation of rhomboid-3 alone caused no ventral narrowing, nor did it enhance the narrowing phenotype of rhomboid-1. Note that rhomboid-1 nulls often lack KOs and as such embryos could not be scored for this measurement (n is the number of embryos analysed; SE is standard error). (C) No additional defects in abdominal denticle specification were evident in rhomboid-3, or rhomboid-3 rhomboid-1 double mutant embryos compared to wild-type and rhomboid-1 mutants, respectively.

View larger version (131K): [in a new window]   Fig. 6. Phenotypic analysis of denticle belt fusions during embryogenesis. (A) The denticle belt fusion phenotype resulted in folds around the surrounding fused areas (left panel, arrowhead), and in rare events a hole in the cuticle could be seen at the centre of the fusion (right panel, arrowhead). (B) No defects could be observed in paired domains of stage 11/12 embryos expressing DN-EGFR compared to wild-type embryos: paired domains were marked by co-expression of GFP (green) and Engrailed was used as a segmental marker (in red). In these and subsequent images anterior is to the right. (C,D) Apoptosis (detected by TUNEL labelling in green) was elevated predominantly in paired domains (in red) of ventral epidermal cells of stage 12 embryos (C), and persisted in stage 13 and 14 embryos (D) expressing DN-EGFR (right panels) compared with wild-type embryos (left panels). (E,F) By stage 15, fusions became evident as curvature of Engrailed stripes laterally (arrowheads) and missing stripes medially only in paired domains of embryos expressing DN-EGFR (F shows a merged image of E with paired domain marked by co-expression of GFP in green, and Engrailed in red). (G) The cuticle phenotype of embryos expressing DN-EGFR in paired domains compared to those expressing GFP alone: the only cuticle phenotype of the DN-EGFR-expressing embryos was strong denticle belt fusions in alternating parasegments (paired domains).

View larger version (101K): [in a new window]   Fig. 7. Epidermal cell apoptosis in spitz group mutants causes denticle belt fusions. (A-D) Apoptosis of ventral epidermal cells was greatly elevated in spitz null embryos compared to wild-type or balanced embryos carrying one null copy of spitz. The ventral surface is shown for each embryo, with anterior being up in all images. Homozygous spitz null embryos were marked by the absence of Engrailed/ß-galactosidase staining (in red), and apoptotic cells were detected by TUNEL labelling (in green). Although epidermal apoptosis was elevated as early as stage 10/11 (A), apoptosis in stage 13/14 spitz null embryos was much stronger, particularly in medial regions (B). Note that these spitz null embryos display the strongest fusion phenotype of any spitz group mutants (Fig. 4A,B). (C) Removing the three main apoptosis activators (grim, reaper, hid) using the H99 deletion (White et al., 1994) suppressed the fusion phenotype of rhomboid-3 rhomboid-1 double mutant embryos, but did not eliminate it completely. It should be noted that some apoptosis has been observed in the absence of these genes (Foley and Cooley, 1998). (D) No additional phenotypes were detected in smooth-cuticle cells of rhomboid-3 rhomboid-1 double mutant embryos partially blocked for apoptosis by virtue of the H99 deletion. Unexpectedly, the entire A4 denticle belt was frequently missing in these embryos.

View larger version (19K): [in a new window]   Fig. 8. A refined model for the roles of EGFR signalling in embryonic epidermal patterning (see Discussion). A lateral cross section of a part of the ventral epidermis is depicted with smooth-cuticle cells in white and denticle-secreting cells and denticles in black. The rhomboid-1-expressing cells that act as the source of cleaved Spitz are depicted with green nuclei, while cells that are the source of the Wingless signal have blue nuclei. Arbitrary levels of EGFR signalling versus two thresholds for the observed phenotypes are shown above the cells. Mutation of spitz group genes (red curve) results in apoptosis of smooth-cuticle cells when the levels of EGFR signalling fall below a threshold, and this is manifest as a denticle belt fusion. Reducing signalling further (for example, by removing the EGFR, black curve) results in denticle belt fusions and failure of denticle fate specification (Szüts et al., 1997).