EGF Receptor Signaling: Putting aNew Spin on Eye Development

Tanya Wolff

The coordinated polarization of cells within an epithe-lium is required for the development and function ofsome tissues. Recent work has shown that the EGFreceptor signaling pathway plays a key role in estab-lishing epithelial polarity in the compound eye ofDrosophila.

The 800-fold reiteration of facets that tile the Drosophilacompound eye is exquisite not only in its flawlessnessand aesthetic appeal, but also, from a molecular andcell biological vantage point, in its elaborate executionof multiple signaling pathways and cellular processes.For example, the epidermal growth factor (EGF) recep-tor signaling pathway plays a myriad roles in eye devel-opment, including the regulation of cell proliferation andrecruitment, ommatidial spacing, and apoptosis [1].Recent studies [2–4] have now identified a role for thispathway in a specialized type of cellular movement thatoccurs during eye development.

One of the early events in patterning the fly eye isrecruitment of the future photoreceptor cells intosymmetric ommatidial precursors [5] (Figure 1A). Thesecells then adopt unique photoreceptor fates withrespect to their proximity to the dorsoventral midline,thereby breaking the symmetry and giving rise to twochiral forms of ommatidia. These chiral forms lie onopposite sides of the dorsoventral midline (Figure 1B).Next, these clusters undergo a striking morphogeneticmovement, known as ommatidial rotation, which bothpolarizes the epithelium and organizes it to optimizevisual acuity.

Ommatidial rotation is an elegant and unique type ofcellular movement in which groups of cells rotatewithin an epithelium as individual units, independentof their undifferentiated neighbors. Ommatidial pre-cursors rotate precisely 90° counterclockwise in thedorsal half of the eye and 90° clockwise in the ventralhalf of the eye (Figure 1A). This specialized movementrequires the coordinated maintenance of cell–cellcontacts between photoreceptor cells within a clusterand the simultaneous abolition of contacts betweencells within a cluster and their undifferentiatedneighbors. Subsequently, the cytoskeleton must drivethe physical movement of rotating clusters.

Little is known about the control of rotation at amechanistic level. Only three genes in which mutatonsspecifically affect the degree of rotation, and not thedirection of rotation or chirality, have been identified:nemo, scabrous and roulette. Mutations in nemo causeommatidia to arrest rotation at 45° [6], whereas muta-tions in scabrous [7,8] cause ommatidia to rotate more

than 90° [9]; roulette mutant eyes contain a mixture ofover-rotated and under-rotated ommatidia [2–4,6].

The molecular identities of the nemo and scabrousgene products have been known for some time, butthe roulette gene product has only just been charac-terized. Three groups [2–4] have now reported evi-dence that roulette is an allele of argos — henceforthreferred to as argosrlt — which encodes a secreted,inhibitory ligand of the EGF receptor [10,11].Mapping and sequence data for argosrlt [3,4], as wellas information on the argos mutant phenotype [2,4],indicate that the argosrlt allele carries a regulatorymutation of argos.

The three groups all found that altering levels of EGFreceptor activity, either positively or negatively, resultsin defects in ommatidial rotation. The authors proposethree distinct interpretations for a mechanism by whichEGF receptor signaling regulates ommatidial rotation:that it plays a role in cementing rotated clusters in posi-tion to prevent them from getting jostled around duringlater morphogenetic events; that it is generally requiredfor the physical process of rotation; or that, as a conse-quence of extra cells being recruited into ommatidialclusters when the EGF receptor signaling level is abnor-mal, localization of rotational cues is disrupted.

Brown and Freeman [2] explored the effects ofincreasing EGF receptor signaling and found that itresulted in defects in ommatidial orientation in adulteyes. Surprisingly, ommatidial precursors rotatenormally, suggesting the EGF receptor pathway isdispensable for the process of rotation per se. Byearly pupal life, though, ommatidia are out of registerwith the 90° axis. Given that ommatidia initiate andstop rotation normally, these authors conclude thatEGF receptor signaling functions to anchor rotatedommatidial precursors in place, thereby insulatingthem from physical stresses imposed on the retinaduring subsequent morphogenetic events.

Gaengel and Mlodzik [3] also examined the effectof genetically increasing EGF receptor activity, butthey found contrasting results: in an argosrlt mutantbackground, both over-rotated and under-rotatedommatidial precursors were observed, suggestingthat normal argos expression is required for properommatidial rotation. These discrepant results mightpoint towards an additional signaling pathway, medi-ated by Argos, which controls rotation.

As reported recently in Current Biology, Strutt andStrutt [4] base their model on the discerning obser-vation that ommatidial clusters have extra photore-ceptor cells, leading them to suggest yet a differentrole for EGF receptor signaling. As the photorecep-tors play an essential role in directing rotation, Struttand Strutt [4] speculate that the presence of toomany photoreceptors might interfere with rotationalcues that are propagated by these cells. One suchcue is the subcellular localization of Frizzled, a Wing-less receptor protein known to be involved in the

Dispatch

Current Biology, Vol. 13, R813–R814, October 14, 2003, ©2003 Elsevier Science Ltd. All rights reserved. DOI 10.1016/j.cub.2003.09.053

Washington University School of Medicine, Department ofGenetics, Box 8232, St. Louis, Missouri 63110, USA.

establishment of tissue polarity. Their finding thatFrizzled localization was drastically disrupted inargosrlt mutant clusters at the time the clusters arebeginning to rotate supports their hypothesis.

Signaling pathways are used repeatedly throughoutdevelopment to direct diverse events. One means ofachieving this versatility is through the use of differenttransducers, which exert their effects at multiple levels.In Drosophila, the canonical Ras–Raf–MAP kinasepathway serves as the main effector pathway for EGFreceptor signaling (reviewed in [12]), and the transcrip-tion factor Pointed is frequently part of this cascade.The three groups [2–4] sought to determine if theRas–Raf–MAP kinase pathway is involved in thecontrol of ommatidial rotation by the EGF receptor,and also to identify additional effectors in this event.Together, they demonstrated links between the EGFreceptor pathway and transcription factors, cytoskele-tal elements and cadherins.

The link to the nucleus and transcriptional regulationwas made by showing that Ras [3,4] and Pointed [2,3]act downstream of EGF receptor signaling to mediaterotation. Given that rotation involves remodeling of thecytoskeleton and that Ras, by acting through a subsetof its effectors, can mediate such events [13,14],Gaengel and Mlodzik [3] tested several candidategenes for their ability to genetically modify the Rasrotation phenotype. This work, in addition to loss-of-function studies, provides evidence that Canoe — acomponent of adherens junctions [15] and an effectorof Ras — is the link between EGF receptor signalingand the cytoskeleton. Finally, cadherins could also be

affected by EGF receptor signaling, given their roles inommatidial rotation. The papers show that DE-Cad-herin/shotgun interacts genetically with the EGF recep-tor pathway [2,3], and that Flamingo, a tissue polarityprotein that co-localizes with Frizzled, is mislocalizedin argosrlt eye discs [3,4]. While this work has shownthat Argos, and more broadly, EGF receptor signaling,regulates ommatidial rotation, further work is neces-sary to better define the mechanism.

References1. Dominguez, M., Wasserman, J.D. and Freeman, M. (1998). Multiple

functions of the EGF receptor in Drosophila eye development. Curr.Biol. 8, 1039-1048.

2. Brown, K.E. and Freeman, M. (2003). EGF receptor signallingdefines a protective function for ommatidial orientation in theDrosophila eye. Development, in press.

3. Gaengel, K. and Mlodzik, M. (2003). EGF receptor signaling regu-lates ommatidial rotation and cell motility in the Drosophila eye viaMAPK/Pnt signaling and the Ras effector Canoe/AF6. Development,in press.

4. Strutt, H. and Strutt, D. (2003). EGF Signaling and OmmatidialRotation in the Drosophila Eye. Curr. Biol. 13, 1451-1457.

5. Wolff, T. and Ready, D.F. (1993). The Development of Drosophilamelanogaster. Plainview, N.Y.: Cold Spring Harbor LaboratoryPress. 2 v.

6. Choi, K.W., and Benzer, S. (1994). Rotation of photoreceptor clus-ters in the developing Drosophila eye requires the nemo gene. Cell78, 1251-1236.

7. Baker, N.E., Mlodzik, M., and Rubin, G.M. (1990). Spacing differen-tiation in the developing Drosophila eye: a fibrinogen-related lateralinhibitor encoded by scabrous. Science 250, 1370-1377.

8. Mlodzik, M., Baker, N.E., and Rubin, G.M. (1990). Isolation andexpression of scabrous, a gene regulating neurogenesis inDrosophila. Genes Dev. 4, 1848-1861.

9. Chou, Y.H. and Chien, C.T. (2002). Scabrous controls ommatidialrotation in the Drosophila compound eye. Dev, Cell 3, 839-850.

10. Freeman, M., Klambt, C., Goodman, C.S. and Rubin, G.M. (1992).The argos gene encodes a diffusible factor that regulates cell fatedecisions in the Drosophila eye. Cell 69, 963-975.

11. Schweitzer, R., Howes, R., Smith, R., Shilo, B.Z. and Freeman, M.(1995). Inhibition of Drosophila EGF receptor activation by thesecreted protein Argos. Nature 376, 699-702.

12. Shilo, B.Z. (2003). Signaling by the Drosophila epidermal growthfactor receptor pathway during development. Exp. Cell Res. 284,140-149.

13. Joneson, T., White, M.A., Wigler, M.H. and Bar-Sagi, D. (1996). Stim-ulation of membrane ruffling and MAP kinase activation by distincteffectors of RAS. Science 271, 810-812.

14. Rodriguez-Viciana ,P., Warne, P.H., Khwaja, A., Marte, B.M.,Pappin, D., Das, P., Waterfield, M.D., Ridley, A. and Downward, J.(1997). Role of phosphoinositide 3-OH kinase in cell transformationand control of the actin cytoskeleton by Ras. Cell 89, 457-467.

15. Matsuo, T., Takahashi, K., Suzuki, E. and Yamamoto, D. (1999). TheCanoe protein is necessary in adherens junctions for developmentof ommatidial architecture in the Drosophila compound eye. CellTissue Res. 298, 397-404.

DispatchR814

Figure 1. Ommatidial rotation in development of theDrosophila compound eye.

(A) Pattern formation in the Drosophila eye begins in the thirdlarval instar. Photoreceptors (colored cells) are recruited intoommatidial precursors. These future ommatidia then rotate 90°counterclockwise in the dorsal half of the eye and 90°clockwise in the ventral half of the eye. Upon the completion ofrotation, ommatidia in the dorsal and ventral halves of the eyeare mirror-symmetrical to one another, about the dorsoventral(D/V) midline. Arrows indicate direction of rotation. (B) Twochiral forms of ommatidia (red in dorsal half, blue in ventral half)lie on opposite sides of the dorsoventral midline.

A B

D/V Current Biology