Cell, Vol. 104, 687–697, March 9, 2001, Copyright 2001 by Cell Press

EGF Receptor and Notch Signaling Act Upstreamof Eyeless/Pax6 to Control Eye Specification

the patterning genes Hedgehog (Hh) and Decapen-taplegic (Dpp) are essential (Heberlein and Treisman,2000). Furthermore, all seven of these eye specification

Justin P. Kumar and Kevin Moses*Department of Cell BiologyEmory University School of Medicine1648 Pierce Drive genes are nuclear factors that must receive patterning

inputs from one or more signal transduction pathways.Atlanta, Georgia 30322Notch is a transmembrane receptor activated by

“DSL” class ligands that transduce signals to the nu-cleus by means of a pathway that includes the EnhancerSummaryof Split Complex genes [E(spl)C; Artavanis-Tsakonas etal., 1995]. Notch acts in eye development in setting upThe Drosophila compound eye is specified by the con-

certed action of seven nuclear factors that include the dorsal-ventral compartment boundary, in establish-ing planar polarity, in the spacing of ommatidial clusters,Eyeless/Pax6. These factors have been called “master

control” proteins because loss-of-function mutants and in cell fate specification (Cagan and Ready, 1989;Blair, 1999; Baker, 2000). The Drosophila EGF receptorlack eyes and ectopic expression can direct ectopic

eye development. However, inactivation of these genes homolog (Egfr) is a transmembrane receptor tyrosinekinase (RTK) that acts through the Ras cascadedoes not cause the presumptive eye to change identity.

Surprisingly, we find that several of these eye specifi- (Schweitzer and Shilo, 1997; Nilson and Schupbach,1999). In the developing eye, Egfr signaling has beencation genes are not coexpressed in the same embry-

onic cells—or even in the presumptive eye. We demon- shown to control cell fate specification, inhibit pro-grammed cell death, and modulate cell cycle progres-strate that the EGF Receptor and Notch signaling

pathways have homeotic functions that are genetically sion (Bergmann et al., 1998; Freeman, 1998; Kurada andWhite, 1998; Kumar and Moses, 2000).upstream of the eye specification genes, and show

that specification occurs much later than previously Here we show that the Egfr and Notch signaling cas-cades act antagonistically as homeotic determinants ofthought—not during embryonic development but in

the second larval stage. eye specification: hyperactivation of Egfr signaling ordownregulation of Notch activity within the presumptiveeye field leads to the complete transformation of theIntroductioneye into an antenna. Under these conditions, the eyespecification genes are no longer transcribed in theSeven “master control” genes have been reported to

act in the specification of the compound eye: twin of transformed tissue. We propose that both the Egfr andNotch pathways are genetically upstream of the knowneyeless (toy), eyeless (ey), eyes absent (eya), sine oculis

(so), dachshund (dac), eye gone (eyg), and optix (opt). complex of eye specification genes and direct the for-mation of the eye and antenna. To our knowledge, thisWhere mutants are known, loss of these functions re-

sults in the failure of the eye to form, while their ectopic is the first report of a homeotic function in organ specifi-cation for any receptor tyrosine kinase.expression (except for so) is sufficient to induce ectopic

eyes (Gehring and Ikeo, 1999; Heberlein and Treisman,2000; Seimiya and Gehring, 2000). The complexity of Resultsthe genetic epistatic relationships between these sevengenes and the observed in vitro interactions of their Egfr Signaling Can Direct Eye to Antennaprotein products suggests that they lie in a regulatory Fate Transformationnetwork and not in a simple linear hierarchy (Heberlein We found that the Egfr and Notch pathways function inand Treisman, 2000). The simplest hypothesis based on the specification or determination of the eye (Figure 1).the published data is that these seven proteins form a We used an ey-GAL4 driver to express target proteinsmultimeric complex that is an eye-specifying transcrip- in all of the experiments described in this section. Thistion factor. One might expect them all to be coexpressed element drives expression first in the eye and antennaat an early time in a limited number of cells and that anlagen in the embryo (by stage 11) and then in regionsthese cells are thus specified to form the eye. ahead of the furrow in just the eye imaginal disc (Figure

However, it is clear that these proteins cannot be the 2A; Gehring, 1998). For these experiments, we were un-sole determinants of eye fate, nor can they lie at the top able to use more conventional and direct means of dis-of the regulatory hierarchy that controls eye specifica- rupting these pathways; null mutations are lethal withtion. In loss-of-function mutants, the presumptive eye more global phenotypes before disc fate can be ob-disc does not adopt a different fate but degenerates served, and mosaic clones do not usually target everylate in development (Heberlein and Treisman, 2000). In founder cell of the eye.addition, the ability of these genes to direct eye develop- Egfr function was removed in this domain by express-ment in non-eye tissue is limited to subregions of the ing a dominant negative form of the receptor. Both theantennal, wing, and leg discs, in which the presence of eye and antenna were deleted from eclosed adults indi-

cating that both structures require Egfr signaling fortheir specification, determination, or survival (Figure 1A).* To whom correspondence should be addressed (e-mail: kmoses@

cellbio.emory.edu). Under these conditions, the larval discs do not form,

Cell688

Figure 1. Egfr, Notch, Wg, and Hh ControlEye and Antenna Fate

SEMs: dorsal view of head, anterior up (Aand B), side view of head, anterior right (C–I).Arrowheads show endogenous antennae,arrows show transformed antennae, aster-isks show missing appendages. (A) eyflp; act-5C.y1.GAL4/UAS-EgfrDN. (C) Wild-type. (Band D–H) All driven by ey-GAL4, proteindriven via UAS indicated top right corner ofeach panel. (D) Note: ectopic antenna hasall segments including arista and is orientedsimilarly to the endogenous antenna.

making analysis of later developmental phenotypes im- way in the eye anlagen using the same driver and wefind that hyperactivation of many elements leads to thepossible. The same phenotype was obtained with domi-

nant negative Ras indicating that this activity is Ras homeotic transformation of the eye into a morphologi-cally complete antenna. This is different from the hyper-dependent (not shown). We expressed wild-type and

activated forms of several components of the Ras path- plastic growth of the eye reported by others (Karim and

Figure 2. Egfr and Notch Regulate Expression of Antenna Specification Genes

(A) Expression pattern of the ey-GAL4 driver. (B, E, and G) Wild-type, indicated in lower left of each panel. (C) ey1 homozygote. Note: despitethe lack of photoreceptors, the eye disc still forms. (D and F) Protein driven by ey-GAL4 indicated in lower left of each panel. (E–H) Actin inred, green shows molecule indicated in top right of each panel. Anterior right, all panels to same scale.

Egfr and Notch in Early Eye Development689

Rubin, 1998) and may reflect the different promoter con- both Su(H) and many of the proteins of the E(spl) com-plex (m4, m7, m8, m8DN, ma, mb, mg, and md) but ob-texts for the ey transcriptional enhancer used in their

construct. While we have not verified their ey-Gal4 lines, served no effect on either eye or antenna disc develop-ment (not shown). We did, however, obtain homeoticwe have shown that the ey-Gal4 line used here does

direct the correct expression pattern of a UAS-lacZ re- eye to antenna transformations when we removed Mas-termind (Mam, using a dominant negative construct), aporter (not shown).

Homeotic transformation of the eye to antenna can member of the neurogenic gene group that encodes anuclear protein of unknown function (Helms et al., 1999).also be induced by the Egfr ligand Spitz (Figure 1B) but

not by two other known activators (Vein and membrane These results suggest a Su(H) and E(spl)C independentpathway for eye and antenna disc development thatbound and secreted forms of Gurken, not shown). The

membrane bound version of Spitz does not induce ho- involves Mam.The Wg and Hh pathways have been shown to regu-meotic transformations, suggesting a requirement for

paracrine signaling (not shown). We expressed wild- late the size and shape of the eye (Royet and Finkelstein,1996; Royet and Finkelstein, 1997; Treisman and He-type and constitutively active forms of the Egfr and two

other Drosophila RTKs (Breathless, Btl, and Heartless, berlein, 1998). Overexpression of a transgene containingthe full-length wild-type Cubitus Interruptus (Ci) protein,Htl; Klambt et al., 1992; Beiman et al., 1996; Gisselbrecht

et al., 1996) and only Egfr was able to induce the transfor- a downstream component of Hh signaling (Aza-Blancand Kornberg, 1999), can mediate eye to antenna ho-mation (not shown). Expression of the constitutively ac-

tive version of Egfr gave a significantly stronger pheno- meotic transformations (Figure 1I). We cannot be certainif these data show a positive or negative Hedgehogtype than the wild-type version of Egfr suggesting that

the level of Egfr signaling is important for maintaining the signal; Ci may be cleaved into the repressor form therebymimicking the effects on eye development seen in lossbalance between eye and antennal identities (compare

wild-type in Figure 1C to transformed in Figure 1D). The of Hedgehog signaling. Overexpression of Wg resultsin the near complete deletion of the eye (Figure 1G).downstream elements of the pathway that can induce

this transformation include Ras, Raf (Figure 1D), and Downstream of Wg lie two transcription factors, SloppyPaired 1 and 2 (Slp1, Slp2; Grossniklaus et al., 1992;PntP1, while neither MEK, MAPK, nor PntP2 induces

this effect in this assay. Aop, Tramtrack (Ttk), and BarH1/ Bhat et al., 2000). The effect of Wg on eye specificationappears to be mediated through Slp2 (Figure 1H) but notH2, which mediate negative feedback inhibition of Egfr

signaling (Dickson, 1998), delete the eye (Figure 1F). Slp1 (not shown). Taken together, these observationssuggest that along with Egfr and Notch, both Wg andIn all cases where a particular transgene failed in our

assay, we introduced several copies of the transgene Hh signaling function during eye and antenna disc speci-fication.(to increase the genetic dose) and conducted the experi-

ment at 298C (to maximize Gal4 function). Both of thesesteps should increase the amount of protein that is pro- Do Egfr and Notch Act Upstream of the Eyeduced. In all of the reported cases here, these steps Specification Genes?failed to produce any additional phenotypic effects. We We undertook a molecular epistasy study, examiningwere able to determine that each element is functional the expression of some of the eye and antennal specifi-by inducing embryonic lethality via ubiquitous expres- cation genes in the transforming conditions (describedsion of each transgene using a hsp70-GAL4 driver. above) during the third larval stage (before cell typesThese three tests were also used to validate each ele- differentiate, Figure 2B). In eye specification gene mu-ment that produced a negative result described here- tants (such as ey), ommatidial development is blocked,after. Thus, it may be that the failure of Mek, Mapk, and but the eye disc remains in a reduced form (Figure 2C).PntP2 to induce this transformation reflect the existence Conditions that produce eye to antenna transforma-of actual branch points in the pathway. However, it is tions, whether through hyperactivation of Egfr or down-also possible that the quantitative levels of expression regulation of Notch signaling, show a complete replace-of these three elements are not limiting for this signal at ment of the eye disc with an antenna disc (Figure 2D).this time and place, indeed their phosphorylation states Distal-less (Dll) and Spalt-Major (SalM, not shown) aremay be more relevant. normally expressed within subdomains of the antenna

disc and are required for antenna development (Figure2E; Si Dong et al., 2000). Dll and SalM are expressed inNotch Is Required for Eye Fate Determination

Notch and Egfr have been shown to often antagonize the correct locations in the transformed antenna discsuggesting that both endogenous and transformed an-each other during cell fate decisions in the fly eye (Fortini

et al., 1993; Sawamoto and Okano, 1996) and elsewhere tenna are both morphologically and molecularly equiva-lent (Figure 2F).(Price et al., 1997; zur Lage and Jarman, 1999). We re-

moved Notch function with a dominant negative form We examined the transcription of five of the sevenknown eye specification genes (toy, ey, eya, so, andand obtained results similar to the effects of Egfr signal

hyperactivation (Figure 1E). Consistent with this, when eyg). In transforming conditions, transcription levels ofall five of the seven genes are below the levels of detec-we expressed an activated form of Notch, the size of the

eye was reduced and there were severe dysmorphies (as tion (ey: Figures 2G and 2H, others not shown). This isconsistent with both Egfr and Notch signaling actingreported by Kurata et al., 2000). Expression of dominant

negative transgenes of the ligands Delta (Dl) or Serrate genetically upstream to both the eye and antennal speci-fication genes. The downregulation of ey suggests that(Ser) also results in the eye to antenna transformation

(not shown). We drove elevated expression levels of the ey-GAL4 driver may also be downregulated via an

Cell690

Figure 3. Expression of Eye and Antennal Specification Genes in the Embryo

(A–I) Horizontal optical sections of wild-type embryos. Identity of red and green molecules at the bottom left of each panel. Arrows indicatepresumptive eye imaginal discs. Asterisks indicate auto-fluorescence in the presumptive gut. Anterior left, developmental time indicated attop right of each panel.

autoregulatory mechanism. That the transformation oc- antenna transforming function of Egfr and Notch path-way signaling should be coincident with, or earlier than,curs despite this may reflect a phenocritical period for

the eye-antenna transformation; once the transforma- the time at which the eye and antenna specificationgenes are first specifically coexpressed. We set out totion has occurred the system is refractory to the loss of

Egfr signaling (see below). test all three of these predictions.

When and Where Are the Eye and AntennaWhen and Where the Eye and AntennaAre Specified: a Hypothesis Specification Genes First Expressed?

To test the first prediction above, we collected embryosThe seven known eye specification genes are thoughtto act in a genetic and biochemical complex; by pairwise (at 1 hr intervals from 1 to 16 hr after egg deposition,

AED) and analyzed them for expression of the canonicaltests, their products have been shown to either directlyregulate each other’s transcription or to interact at the eye specification gene ey (Pax6) and the antenna specifi-

cation protein Dll (Figure 3). Dll is first detected at 7 hrprotein level, or both (Heberlein and Treisman, 2000).From the few published reports of the early expression in the leg imaginal disc primordia and in several seg-

ments in the embryonic head (data not shown). ey tran-patterns of eye specification genes and from fate map-ping experiments, it has been suggested that eye versus scription in the eye imaginal disc is first detectable at

11 hr (arrows in Figure 3A) while Dll is seen in an adjacentantennal fate specification occurs during the latter stagesof embryogenesis (Heberlein and Treisman, 2000). These region (arrowheads in Figure 3A) as well as other sites.

In latter stages of embryogenesis, the eye imaginal discconcepts lead to a straightforward hypothesis: at somepoint in the developing embryo, the seven eye specifica- invaginates and assumes a more dorsal-medial position

within the embryonic head, just above the developingtion genes’ products are coexpressed in the presump-tive eye and act to specify its fate. A similar event (with embryonic brain (Younossi-Hartenstein et al., 1993). We

observe regions of ey expression that correspond todifferent genes acting) also specifies the antenna.If this hypothesis is true, then three predictions should this (arrows in Figures 3B and 3C). Furthermore, this ey

expression corresponds to domains of Escargot expres-hold: (1) At some time during embryonic development,there should be two domains of expression of the eye sion (Esg), a general imaginal disc marker (data not

shown; Hayashi et al., 1993). However, Dll expressionspecification genes that correspond to the future eyesand anterior to these should be two domains of antenna is more anterior and it is not clear if these sites corre-

spond to the presumptive antennae (arrowheads in Fig-specification gene expression marking and acting todirect antenna fate. These gene products should be ures 3B and 3C). It thus appears that ey is expressed

in both the presumptive eye and antenna by 13 hr andspecific to the future structures they mark, and shouldnot be found elsewhere. (2) The eye specification genes remains there through the last embryonic time point

observed, and that Dll is not expressed in the futureshould be coexpressed in the same cells. This is knownto be true of toy, ey, and eyg (Jones et al., 1998; Czerny antenna at any embryonic time. It is also quite clear that

ey is expressed in many sites in the embryo that willet al., 1999). (3) The phenocritical period for the eye to

Egfr and Notch in Early Eye Development691

never form eye (such as the segmental grooves). Inshort, we cannot distinguish the position of the pre-sumptive eye or antenna during embryonic developmentbased on the specific expression of their respective“master control” genes—neither ey nor Dll expressionare sufficient to specify the eye or the antenna; therefore,prediction 1 (above) does not hold true.

Are the Eye Specification Genes Coexpressedduring Embryonic Development?We examined the expression pattern of Eya and Dacproteins and so transcription at 1 hr time points (from1 to 16 hr AED) and found that none of these three eyespecification genes are coexpressed with ey within thepresumptive eye (Figures 3D–3I). The fact that thesegenes are not expressed within the same cells duringembryonic development precludes any possibility thattheir products act in a multiprotein complex critical foreye specification in the embryo and, thus, prediction 2(above) does not hold true either. However, eye specifi-cation might occur later in development.

The Eye Specification Genes Are FirstCoexpressed in the Second Larval StageIn second stage larva, the eye specification gene prod-ucts are completely segregated into the eye portion ofeye-antennal disc (Figures 4A, 4C, 4E, and 4G), but theantennal marker Dll is evenly expressed in both the eyeand antennal segments (Figure 4I). Interestingly, the ex-pression patterns of the eye specification genes are stillnot completely overlapping. For instance, toy appearsto be expressed throughout the entire eye field (Figure4A) while both eya and dac are expressed just in theposterior portions of the eye disc (Figures 4C and 4E).In the third larval stage, the eye specification genesremain within the eye portion (Figures 4B, 4D, 4F, and4H) and Dll is now segregated to just the antennal seg-ments (Figure 4J).

The Phenocritical Period for Eye to AntennaTransformation Is Also in the SecondLarval StageWe made use of the cold sensitivity of the GAL4 proteinto determine the phenocritical period. GAL4 is a yeastprotein and is fully functional at 258C but is less activeat 188C. Flies of the ey-GAL4/UAS-SerDN genotype wereraised at 188C, shifted to 258C for a consecutive seriesof 24 hr periods, and then returned to 188C until latethird instar imaginal discs could be examined (Figure5). The use of the dominant negative Ser construct inthis experiment effectively eliminates Notch pathway Figure 4. Expression of Eye Specification Genes in Second and

Third Larval Stage Imaginal Discsfunction in the developing eye. Indistinguishable trans-formations were observed in other experiments with All wild type. (A, C, E, G, and I) Second instar to same scale, bar in

(A) shows 100 mm. (B, D, F, H, and J) Third instar to same scale,constructs that either hyperactivate Egfr pathway sig-bar in (B) shows 100 mm. Molecules visualized are identified at thenaling or inactivate Notch (see above). The developingleft of each row. Anterior right.eye-antennal complex is completely normal if kept con-

tinuously at 188C (negative control, data not shown)while constant exposure to 258C temperatures resulted val stages failed to induce any effects. The eye-antennal

discs are completely normal as seen in Figure 5B. This isin the eye to antenna transformation (positive control,Figures 2D, 2F, and 2H). These controls confirm that the consistent with our expression data (above) suggesting

that the eye is not specified during embryogenesis. Acold sensitivity of GAL4 protein activity is sufficient inour hands to control the transformation. temperature shift during the first half of the second larval

stage resulted in a reduced eye field, but no transforma-Temperature shifts during the embryonic and first lar-

Cell692

Figure 5. The Phenocritical Period for Eye toAntenna Transdetermination Is during theSecond Larval Stage

(A) Time course of temperature shifts. Bluelines indicate time course of loss of Notchactivity by means of shift from 188C to 258Cfor 24 hr periods (ey-GAL4/UAS-SerDN). Let-ters at right indicate corresponding panelsbelow. Phenocritical period is denoted witha shaded box. The dark gray central strip indi-cates time period in which the eye to antennatransformation occurs. Green line indicatestime of morphogenetic furrow initiation. Timeline shows hours AED and the correspondingstages of development. (B–F) ey-GAL4/UAS-SerDN. Red is Dll and green is Elav. Anteriorright, (B)–(F) same scale.

tion to antenna (Figure 5C). Interestingly, this phenotype cells that normally distribute preferentially to the eyeare now equally allocated to both antennae.is similar to that seen in ey mutant homozygotes. The

eye to antenna transformation is fully induced in allcases when the temperature shift occurs during the lat-ter half of the second larval stage (Figure 5D). The trans- The Notch Pathway Signals Differentially

in the Eye and Antenna Primordiaformed antenna expresses Dll in an identical pattern asseen in the endogenous antenna. Subsequent tempera- in the Second Larval Stage

We have determined that loss of Notch activity duringture shifts during the earliest phase of the third larvalstage do not result in a complete transformation (Figure the second larval stage results in the transformation of

the eye into an antenna. Thus, we predict that Notch5E). Interestingly, loss of Notch during the next day ofthe third larval stage results in defects in the regulation signaling should be elevated in the presumptive eye

versus the antenna at the critical time. Cells that areof the morphogenetic furrow (Figure 5F, details to bepublished elsewhere). These results clearly indicate that actively receiving a Notch signal upregulate Notch pro-

tein expression (Artavanis-Tsakonas et al., 1999). Thusthe phenocritical period is chiefly within the latter halfof the second larval stage. This phenocritical period elevated Notch antigen expression can be used as a

reporter of elevated Notch signaling. We examineddoes not predate the expression of any of the eye speci-fication genes, but it is coincident with their first co- Notch and ey expression in imaginal discs from first,

second, and third stage larvae (Figure 6). Both Notchexpression and, thus, prediction 3 (above) holds true.Is the presumptive eye actually transformed into a and ey are expressed throughout the entire eye-antennal

disc anlagen during the first larval stage (Figures 6A andsecond antenna under these conditions, or does the eyedegenerate and get replaced by regrowth from else- 6D). By the second larval stage, Notch is differentially

upregulated within the presumptive eye (Figure 6B). In-where? We favor the former interpretation (transforma-tion) because in hundreds of transformed L2 disc com- terestingly, Notch appears especially active along the

eye margins and midline, where it is thought to regulateplexes dissected, we never observed a degeneratingeye disc, or a small (presumably regrowing) antennal retinal polarity (Blair, 1999). ey, on the other hand, ap-

pears to be exclusively within the eye field (Figure 6E).disc. In all cases, the transformed antenna is equal insize to the normal one. Indeed, both are somewhat larger In the third larval stage, Notch expression is upregulated

in the morphogenetic furrow, where it acts to controlthan normal (compare Figures 2B and 2D, shown to thesame scale). This might suggest that a fixed number of ommatidial spacing (Cagan and Ready, 1989; Baker and

Egfr and Notch in Early Eye Development693

Figure 7. Models for Eye Specification

Left panel shows genetic pathways that regulate eye and antennalspecification at the L2 stage. Our data suggest that Egfr signalinginhibits eye specification and promotes antennal specification, whileNotch signaling is antagonistic to this (red and green arrows). Simi-larly, our data (and others) suggest that Hedgehog signaling is eyeinhibiting and antenna promoting and that Wingless is eye inhibiting.These pathways act upstream of the most upstream eye specifica-tion gene (toy). Others have deduced the genetic hierarchy withinthe eye specification genes (black arrows). Right panel illustrateshow nonoverlapping patterns of eye specification gene expressionin the embryo and L1 stages become overlapping only at the L2stage. This coincidence with the Notch and Egfr signaling phenocrit-ical period suggests that this is the time at which these signals actto control the expression of the eye and antenna specification genesand thus specify these organs fates.

Figure 6. Notch Signaling Activity Is Upregulated in the PresumptiveEye but Not Antenna at the Phenocritical Period

target of Notch or Egfr signals may require direct bio-All are wild type. (A–C) Notch protein. (D–F) ey-lacZ expression.Arrows indicate presumptive eyes, arrowheads indicate presump- chemical assays.tive antennae. Larval stages are indicated on the left of each row. The homeotic effects described here represent the(B and C) Asterisk indicates highest level of N, at the midline and first report known to us of complete eye to antennamargins (B), then in the morphogenetic furrow (C). Anterior right.

transformation. Neither have there been reports of anScale bars in (A) and (C) show 100 mm, (A), (B), (D), and (E) and (C)eye to antennal transdetermination event from serialand (F) to same scale.disc transplantation experiments (Postlethwait andSchneiderman, 1971; Wieschaus, 1974; Hadorn, 1978).Furthermore, we believe that this is the first reportedBronner-Fraser, 2000) while ey remains upregulated

ahead of the furrow (Figures 6C and 6F). example of a homeotic role for not only the Egfr itselfbut for a receptor tyrosine kinase in the specification ofan organ (Ras has been implicated in the modulation ofDiscussionhomeotic transformations of the antenna to maxillarypalps; Boube et al., 1997). As this function is specificEgfr and Notch Signaling in Eye Specification

We have shown that Egfr and Notch signaling act in to Egfr (FGF receptor homologs cannot substitute forit), this homeotic function is specifically attributable tothe specification of the eye: Egfr signaling promotes an

antennal fate while Notch signaling promotes an eye the Egfr itself and not just the Ras cascade.While activating Egfr or blocking Notch signals trans-fate (Figure 7). This role for Notch is consistent with a

recent report in which removal of Notch signaling can forms the eye cleanly into an antenna, the reciprocaltransformation is not complete, suggesting that therepartially inhibit compound eye development (Kurata et

al., 2000). Furthermore, we have shown by molecular may be additional positive regulators of eye fate. Wewere unable to conduct the reciprocal transformationepistasy that several of the eye and antennal specifica-

tion genes (ey, toy, eya, so, eyg, salM, and Dll) are down- experiment (i.e., antenna to eye switch via hyperactiva-tion of Notch or downregulation of Egfr signaling solelystream of the Egfr and Notch inputs. We also have shown

that Wg and Hh pathway signaling affect this specifica- within the antennal anlagen). Unlike the ey-GAL4 driver,there is not an equivalent known driver that is expressedtion (Figure 7). The eye specification genes form a regu-

latory network and the direct control of any one of these solely with the antennal anlagen. All known antennal-determining genes are also expressed in other imaginalgenes may affect the others (Gehring, 1998; Treisman,

1999; Heberlein and Treisman, 2000). Thus, which (if discs. For instance, the Dll-GAL4 driver is expressedin several places within the embryonic head and legany) of the known eye specification genes is a direct

Cell694

imaginal disc. Expression of Egfr or Notch constructs 2000). This led us to expect, perhaps naively, that theexpression of these genes would be restricted to thewith this driver results only in embryonic lethality. It

may be that the antenna can be changed to an eye via presumptive eye and that they should be coexpressedtogether in the eye anlagen from some embryonic stage.alterations of Egfr or Notch signaling provided that the

appropriate tools for their missexpression are available. We found this not to be so; these genes are not restrictedto expression in the eye (they show many other sites ofWhy do homozygous mutants for eye specification

genes not transform the eye into an antenna? While it embryonic expression) and they are mostly not co-expressed (the expression patterns do not overlap inmay be that some alleles are not nulls (e.g., ey1), a more

interesting possibility is that there may be functional the embryo). Thus, they cannot be acting together atthis early stage to specify the eye and none of them areredundancy in some cases—particularly that of ey and

toy. Thus, only when both genetic functions are elimi- sufficient to specify the eye on their own.We suggest that these genes come under separatenated will a true null condition exist. Just such a situation

confused the phenotypic analysis of two other twin ho- regulation by different patterning signals in early devel-opment and that there are overlapping domains (Figuremeodomain proteins, engrailed and invected (Hidalgo,

1996). Unfortunately, mutations of the toy gene do not 7), much like a set of colored spotlights. Only when allof the domains coincide (during the second larval stage)yet exist.do the spotlights form white light, and the eye specifica-tion genes specify the eye (Figure 7). This seems toThe Eye and Antenna Are Specified Latebe the simplest explanation since the eye specificationWe have presented three lines of evidence that suggestgenes form a very tight genetic, biochemical, and tran-that the eye and antenna are not specified until thescriptional regulatory network suggesting that they aresecond larval stage. These are: (1) the late time of thetogether required for eye specification. It may be thatfirst coexpression of the eye specification genes, (2)the final coexpression of the eye specification genes’the phenocritical period for our transgenic, GAL4-drivenproducts (and the exclusion of the antennal specificationtransforming condition, and (3) direct detection of ele-factors) is the last step required to allow the morphoge-vated Notch signaling in the second instar (and not be-netic furrow to initiate in response to the next localfore) in the presumptive eye relative to the antenna. Thisexpression of hh and for the final specification of retinalstage is much later than the time suggested by otherscell types and pattern.based on embryonic morphology (Younossi-Hartenstein

To extend the spotlight metaphor, we suggest thatet al., 1993). Mosaic analysis has shown that a line ofNotch and Egfr signaling (as well as Wg and Hedgehog)clonal restriction arises between these two organs atare some of the “hands” that guide the spotlights. Ouran earlier stage (Postlethwait and Schneiderman, 1971;results indicate that in the specification of the eye, EgfrWieschaus, 1974). However, the existence of a boundaryand Notch signals antagonize each other. In the eye asdoes not require that the two fields separated by it adoptelsewhere, Notch and Egfr are the Shiva and Brahmadifferent fates at that time. Thus, the published experi-of developmental control.ments and our new data are consistent with a two stage

process—an early establishment of a subdivision of theimaginal disc complex into two distinct presumptive or- Experimental Proceduresgan fields, and a later allocation of their specific organidentities (eye versus antenna). Loss of Egfr signaling Drosophila Stocks

The following were used in this study: ey-GAL4 (Bonini et al., 1997),in the developing wing during the second larval instardpp-GAL4 (Staehling-Hampton et al., 1994), actin5C.y1.GAL4 (Ito(z136 hr AED) can transform the developing notum intoet al., 1997), eyFLP (Newsome et al., 2000), UAS-EgfrDN, UAS-sspi,wing tissue (Baonza et al., 2000). Similar to our results,UAS-mspi (Freeman, 1996), UAS-vn (Schnepp et al., 1996), UAS-

this transformation does not occur during the embryonic aos (Freeman, 1994), UAS-spry (Kramer et al., 1999), UAS-Egfrtsla

stage when the wing disc and the compartment bound- (a temperature-sensitive; Kumar et al., 1998), UAS-EgfrtopCO (a null;aries were thought to have been specified. Taken to- Clifford and Schupbach, 1989), UAS-Egfr type I, UAS-Egfr type II,

UAS-EgfrElp type I, UAS-EgfrElp type II (Elp forms are activated;gether with our data on eye and antennal development,Lesokhin et al., 1999), UAS-Egfrtop 4.2 (Queenan et al., 1997), UAS-it is possible that disc fate is generally not finally deter-btl, UAS-btlact (Lee et al., 1996), UAS-btlDN (Reichman-Fried et al.,mined until just prior to the third larval instar.1994), UAS-htl, UAS-htlact, UAS-htlDN (Michelson et al., 1998), UAS-Dras1V12 (activated), UAS-Dras1N17 (dominant negative; Scholz et al.,

How Do the Eye Specification Genes Function? 1997), UAS-Drafgof (Li et al., 1997b), UAS-hep (Boutros et al., 1998),UAS-rl, UAS-rlsem (sem is activated 5 MAPKact; Martin-Blanco,Published genetic epistasy and biochemical interaction1998), UAS-PntP1, UAS-PntP2 (Klaes et al., 1994), UAS-ttk69, UAS-data suggest that the seven known eye specificationttk88 (Li et al., 1997a), UAS-aop, UAS-aopact (Rebay and Rubin,genes’ products interact at the transcriptional and pro-1995), UAS-BarH1, UAS-BarH2 (Hayashi et al., 1998), UAS-NDN

tein levels to direct cells toward eye fate (Heberlein and(Brennan et al., 1997), UAS-DlDN (Huppert et al., 1997), UAS-SerDN

Treisman, 2000). This requires that they are expressed (Fleming et al., 1997), UAS-mamDN, UAS-Su(H) (Helms et al., 1999),in the same cells. Furthermore, it has been suggested UAS-m8, UAS-m8DN (Giebel and Campos-Ortega, 1997), UAS-m4,

UAS-ma (Apidianakis et al., 1999), UAS-mb, UAS-md, UAS-mg (Ali-that many, if not all, of these genes are “master regula-fragis et al., 1997), UAS-m7 (Tata and Hartley, 1995), UAS-prostors” of eye fate—that is, they are both necessary and(Manning and Doe, 1999), UAS-wg (Lawrence et al., 1995), UAS-sufficient for eye specification (Gehring, 1998). Manyarm (Boyle et al., 1997), UAS-E cad (gift of I. Davis), UAS-sggact

very compelling experiments have been described(Hazelett et al., 1998), UAS-slp1, UAS-slp2 (Riechmann et al., 1997).

showing the induction of ectopic eyes through the ec- UAS-grk, UAS-sgrk (gifts of T. Schupbach). hh-lacZ, UAS-Ci (Tabatatopic expression of these genes alone or in synergistic and Kornberg, 1994), wg-lacZ (Kassis, 1990), dpp-lacZ (Blackman

et al., 1991), ey-lacZ (Quiring et al., 1994), so-lacZ (Cheyette et al.,combinations (reviewed in Heberlein and Treisman,

Egfr and Notch in Early Eye Development695

1994). All UAS-GAL4 crosses were done at 188C, 258C, and 298C. identity in vertebrate neurogenic placodes. Development 127, 3045–3056.All figures shown in the text are from crosses done at 258C unless

specified otherwise in the text and figure legends. Note: superscripts Baker, N.E. (2000). Notch signaling in the nervous system. PiecesDN 5 dominant negative, act 5 activated, and gof 5 gain of function. still missing from the puzzle. BioEssays 22, 264–273.

Baonza, A., Roch, F., and Martin-Blanco, E. (2000). DER signalingTemperature Shift Regimes restricts the boundaries of the wing field during Drosophila develop-ey-GAL4/UAS-SerDN progeny were raised at 188C and shifted to 258C ment. Proc. Natl. Acad. Sci. USA 97, 7331–7335.for 24 hr at consecutive intervals starting at t 5 0 hr AED (after egg Beiman, M., Shilo, B.Z., and Volk, T. (1996). heartless, a Drosophiladeposition). Animals were then returned to 188C until third instar FGF receptor homolog, is essential for cell migration and establish-larval eye-antennal imaginal discs were examined. Egfrtsla/EgfrtopCO

ment of several mesodermal lineages. Genes Dev. 10, 2993–3002.progeny were raised at 188C and shifted to 288C for 12 hr at consecu-

Bergmann, A., Agapite, J., McCall, K., and Steller, H. (1998). Thetive intervals starting at t 5 0 hr AED. Animals were then returnedDrosophila gene hid is a direct molecular target of Ras-dependentto the 188C until adult compound eyes were examined. Temperaturesurvival signaling. Cell 95, 331–341.shifts that disrupted compound eye structure or caused the eye toBhat, K.M., van Beers, E.H., and Bhat, P. (2000). sloppy paired actstransdetermine into an antenna are referred to as the “phenocritical”as the downstream target of wingless in the Drosophila CNS andperiod.interaction between sloppy paired and gooseberry inhibits sloppypaired during neurogenesis. Development 127, 655–665.Immunohistochemistry In Situ Hybridization and SEMBlackman, R.K., Sanicola, M., Raferty, L.A., Gillevet, T., and Gelbart,The following antibodies were used: rat anti-Elav (O’Neill et al., 1994),W.M. (1991). An extensive 39 cis-regulatory region directs the imagi-rabbit anti-Distal-less (Si Dong et al., 2000), mouse anti-Eya (Bonininal disk expression of decapentaplegic, a member of the TGF-bet al., 1998), mouse anti-Dac (Mardon et al., 1994), mouse anti-family in Drosophila. Development 111, 657–665.Notch extracellular domain (Diederich et al., 1994), rabbit anti-Spalt

(Kuhnlein et al., 1994), rabbit anti-b-galactosidase (Cortex Biochem), Blair, S.S. (1999). Eye development: Notch lends a handedness.mouse anti-b-galactosidase (Promega). Secondary antibodies were Curr. Biol. 9, R356–R360.conjugated to FITC or TRITC (Jackson Labs). F-actin was visualized Bonini, N.M., Bui, Q.T., Gray-Board, G.L., and Warrick, J.M. (1997).with phalloidin (Molecular Probes). Immunohistochemistry on imagi- The Drosophila eyes absent gene directs ectopic eye formation innal discs was performed essentially as described in Tomlinson and a pathway conserved between flies and vertebrates. DevelopmentReady (1987). Immunohistochemistry on embryos was performed 124, 4819–4826.essentially as described in Bonini et al. (1998). In situ hybridizations

Bonini, N.M., Leiserson, W.M., and Benzer, S. (1998). Multiple roleswere performed with minor modifications as described in Maurel-

of the eyes absent gene in Drosophila. Dev. Biol. 196, 42–57.Zaffran and Treisman (2000). Probes were cDNA clones for toy, ey,

Boube, M., Benassayag, C., Seroude, L., and Cribbs, D.L. (1997).so, eya, and eyg (gifts of G. Mardon and W. Gehring). SEM on adultRas-1-mediated modulation of Drosophila homeotic function in cellflies was performed as described in Tio and Moses (1997).and segment identity. Genetics 146, 619–628.

Boutros, M., Paricio, N., Strutt, D.I., and Mlodzik, M. (1998). Dishev-Acknowledgmentselled activates JNK and discriminates between JNK pathways inplanar polarity and wingless signaling. Cell 94, 109–118.We would like to thank S. Artavanis-Tsakonas, N. Baker, P. Beachy,Boyle, M., Bonini, N., and DiNardo, S. (1997). Expression and func-N. Bonini, E. Martin-Blanco, J. Campos-Ortega, S. Campuzano, R.tion of clift in the development of somatic gonadal precursors withinCarthew, C. Delidakis, B. Dickson, C. Doe, J. Fischer, R. Fleming,the Drosophila mesoderm. Development 124, 971–982.M. Freeman, W. Gehring, I. Hariharan, U. Heberlein, Y. Hiromi, R.

Holgren, T. Kornberg, M. Krasnow, G. Mardon, M. Mlodzik, D. Mon- Brennan, K., Tateson, R., Lewis, K., and Martinez Arias, A. (1997).tell, M. Muskavitch, Z. C. Lai, M. Leptin, I. Livne-Bar, G. Panganiban, A functional analysis of Notch mutations in Drosophila. GeneticsN. Perrimon, G. Rubin, K. Saigo, R. Schuh, T. Schupbach, B. Z. 147, 177–188.Shilo, A. Simcox, J. Treisman, B. Yedvobnick, the Bloomington Cagan, R.L., and Ready, D.F. (1989). Notch is required for successiveStock Center, and the Developmental Studies Hybridoma Bank for cell decisions in the developing Drosophila retina. Genes Dev. 3,gifts of fly stocks, clones, and antibodies, R. Apkarian for SEM 1099–1112.technical support, and B. Yedvobnick, S. L’Hernault, K. Bhat, and

Cheyette, B.N.R., Green, P.J., Martin, K., Garren, H., Hartenstein, V.,R. Reifegerste for suggestions, advice, and support. This work wasand Zipursky, S.L. (1994). The Drosophila sine oculis locus encodes asupported by NIH and NSF grants (RO1 EY-12537 and IBN-9807892)homeodomain-containing protein required for the development ofto K. M., and an NIH postdoctoral fellowship (5 F32 EY06763) to J. K.the entire visual system. Neuron 12, 1–20.

Clifford, R.J., and Schupbach, T. (1989). Coordinately and differen-Received August 14, 2000; revised January 16, 2001.tially mutable activities of torpedo, the Drosophila melanogasterhomolog of the vertebrate EGF receptor gene. Genetics 123,

References 771–787.

Czerny, T., Halder, G., Kloter, U., Souabni, A., Gehring, W.J., andAlifragis, P., Poortinga, G., Parkhurst, S.M., and Delidakis, C. (1997).

Busslinger, M. (1999). Twin of eyeless, a second Pax 6 gene ofA network of interacting transcriptional regulators involved in Dro-

Drosophila, acts upstream of eyeless in the control of eye develop-sophila neural fate specification revealed by the yeast two-hybrid

ment. Mol. Cell 3, 297–307.system. Proc. Natl. Acad. Sci. USA 94, 13099–13104.

Dickson, B.J. (1998). Photoreceptor development: breaking downApidianakis, Y., Nagel, A.C., Chalkiadaki, A., Preiss, A., and Deli- the barriers. Curr. Biol. 8, R90–R92.dakis, C. (1999). Overexpression of the m4 and malpha genes of

Diederich, R.J., Matsuno, K., Hing, H., and Artavanis-Tsakonas, S.the E(spl)-complex antagonizes Notch mediated lateral inhibition.(1994). Cytosolic interaction between deltex and Notch ankyrin re-Mech. Dev. 86, 39–50.peats implicates deltex in the Notch signaling pathway. Develop-

Artavanis-Tsakonas, S., Matsuno, K., and Fortini, M.E. (1995). Notch ment 120, 473–481.signaling. Science 268, 225–232.

Fleming, R.J., Gu, Y., and Hukriede, N.A. (1997). Serrate-mediatedArtavanis-Tsakonas, S., Rand, M.D., and Lake, R.J. (1999). Notch activation of Notch is specifically blocked by the product of thesignaling: cell fate control and signal integration in development. gene fringe in the dorsal compartment of the Drosophila wing imagi-Science 284, 770–776. nal disc. Development 124, 2973–2981.Aza-Blanc, P., and Kornberg, T.B. (1999). Ci: a complex transducer Fortini, M.E., Rebay, I., Caron, L.A., and Artavanis-Tsakonas, S.of the hedgehog signal. Trends Genet. 15, 458–462. (1993). An activated Notch receptor blocks cell-fate commitment in

the developing Drosophila eye. Nature 365, 555–557.Baker, C.V., and Bronner-Fraser, M. (2000). Establishing neuronal

Cell696

Freeman, M. (1994). Misexpression of the Drosphila argos gene, a (1999). Sprouty: a common antagonist of FGF and EGF signalingpathways in Drosophila. Development 126, 2515–2525.secreted regulator of cell determination. Development 120, 2297–

2304. Kuhnlein, R.P., Frommer, G., Friedrich, M., Gonzalez-Gaitan, M.,Weber, A., Wagner-Bernholz, J.F., Gehring, W.J., Jackle, H., andFreeman, M. (1996). Reiterative use of the EGF receptor triggersSchuh, R. (1994). spalt encodes an evolutionarily conserved zincdifferentiation of all cell types in the Drosophila eye. Cell 87, 651–660.finger protein of novel structure which provides homeotic gene func-Freeman, M. (1998). Complexity of EGF receptor signaling revealedtion in the head and tail region of the Drosophila embryo. EMBO J.in Drosophila. Curr. Opin. Genet. Dev. 8, 407–411.13, 168–179.

Gehring, W.J. (1998). Master Control Genes in Development andKumar, J.P., and Moses, K. (2000). Cell fate specification in theEvolution: the Homeobox Story (New Haven, CT: Yale UniversityDrosophila retina. In Vertebrate Eye Development, E. Fini, ed. (Berlin:Press).Springer-Verlag), pp. 93–114.

Gehring, W.J., and Ikeo, K. (1999). Pax 6: mastering eye morphogen-Kumar, J.P., Tio, M., Hsiung, F., Akopyan, S., Gabay, L., Seger,

esis and eye evolution. Trends Genet. 15, 371–377.R., Shilo, B.-Z., and Moses, K. (1998). Dissecting the roles of the

Giebel, B., and Campos-Ortega, J.A. (1997). Functional dissection Drosophila EGF receptor in eye development and MAP kinase acti-of the Drosophila enhancer of split protein, a suppressor of neuro- vation. Development 125, 3875–3885.genesis. Proc. Natl. Acad. Sci. USA 94, 6250–6254. Kurada, P., and White, K. (1998). Ras promotes cell survival in Dro-Gisselbrecht, S., Skeath, J.B., Doe, C.Q., and Michelson, A.M. (1996). sophila by downregulating hid expression. Cell 95, 319–329.heartless encodes a fibroblast growth factor receptor (DFR1/DFGF- Kurata, S., Go, M.J., Artavanis-Tsakonas, S., and Gehring, W.J.R2) involved in the directional migration of early mesodermal cells (2000). Notch signaling and the determination of appendage identity.in the Drosophila embryo. Genes Dev. 10, 3003–3017. Proc. Natl. Acad. Sci. USA 97, 2117–2122.Grossniklaus, U., Pearson, R.K., and Gehring, W.J. (1992). The Dro- Lawrence, P.A., Bodmer, R., and Vincent, J.P. (1995). Segmentalsophila sloppy paired locus encodes two proteins involved in seg- patterning of heart precursors in Drosophila. Development 121,mentation that show homology to mammalian transcription factors. 4303–4308.Genes Dev. 6, 1030–1051. Lee, T., Hacohen, N., Krasnow, M., and Montell, D.J. (1996). Regu-Hadorn, E. (1978). Transdetermination. In The Genetics and Biology lated Breathless receptor tyrosine kinase activity required to patternof Drosphila vol 2c, M. Ashburner and T. Wright, eds. (New York: cell migration and branching in the Drosophila tracheal system.Academic Press), pp 555–617. Genes Dev. 10, 2912–2921.

Hayashi, S., Hirose, S., Metcalfe, T., and Shirras, A.D. (1993). Control Lesokhin, A.M., Yu, S.Y., Katz, J., and Baker, N.E. (1999). Severalof imaginal cell development by the escargot gene of Drosophila. levels of EGF receptor signaling during photoreceptor specificationDevelopment 118, 105–115. in wild-type, Ellipse, and null mutant Drosophila. Dev. Biol. 205,

129–144.Hayashi, T., Kojima, T., and Saigo, K. (1998). Specification of primarypigment cell and outer photoreceptor fates by BarH1 homeobox Li, S., Li, Y., Carthew, R.W., and Lai, Z.-C. (1997a). Photoreceptor cell

differentiation requires regulated proteolysis of the transcriptionalgene in the developing Drosophila eye. Dev. Biol. 200, 131–145.repressor Tramtrack. Cell 90, 469–478.Hazelett, D.J., Bourouis, M., Walldorf, U., and Treisman, J.E. (1998).Li, W., Skoulakis, E.M.C., Davis, R.L., and Perrimon, N. (1997b).decapentaplegic and wingless are regulated by eyes absent andThe Drosophila 14–3-3 protein Leonardo enhances Torso signalingeyegone and interact to direct the pattern of retinal differentiationthrough D-Raf in a Ras1-dependent manner. Development 124,in the eye disc. Development 125, 3741–3751.4163–4171.Heberlein, U., and Treisman, J.E. (2000). Early retinal developmentManning, L., and Doe, C.Q. (1999). Prospero distinguishes siblingin Drosophila. In Vertebrate Eye Development, M.E. Fini, ed. (Berlin:cell fate without asymmetric localization in the Drosophila adultSpringer-Verlag), pp. 37–50.external sense organ lineage. Development 126, 2063–2071.Helms, W., Lee, H., Ammerman, M., Parks, A.L., Muskavitch, M.A.T.,Mardon, G., Solomon, N.M., and Rubin, G.M. (1994). dachshundand Yedvobnick, B. (1999). Engineered truncations in the Drosophilaencodes a nuclear protein required for normal eye and leg develop-Mastermind protein disrupt Notch pathway function. Dev. Biol. 215,ment in Drosophila. Development 120, 3473–3486.358–374.Martin-Blanco, E. (1998). Regulatory control of signal transductionHidalgo, A. (1996). The roles of engrailed. Trends Genet. 12, 1–4.during morphogenesis in Drosophila. Int. J. Dev. Biol. 42, 363–368.

Huppert, S.S., Jacobsen, T.L., and Muskavitch, M.A. (1997). Feed-Maurel-Zaffran, C., and Treisman, J.E. (2000). pannier acts upstreamback regulation is central to Delta-Notch signalling required for Dro-of wingless to direct dorsal eye disc development in Drosophila.sophila wing vein morphogenesis. Development 124, 3283–3291.Development 127, 1007–1016.

Ito, K., Awano, W., Suzuki, K., Hiromi, Y., and Yamamoto, D. (1997).Michelson, A.M., Gisselbrecht, S., Buff, E., and Skeath, J.B. (1998).

The Drosophila mushroom body is a quadruple structure of clonalHeartbroken is a specific downstream mediator of FGF receptor

units each of which contains a virtually identical set of neuronessignalling in Drosophila. Development 125, 4379–4389.

and glial cells. Development 124, 761–771.Newsome, T.P., Asling, B., and Dickson, B.J. (2000). Analysis of

Jones, N.A., Kuo, Y.M., Sun, Y.H., and Beckendorf, S.K. (1998). The Drosophila photoreceptor axon guidance in eye-specific mosaics.Drosophila Pax gene eye gone is required for embryonic salivary Development 127, 851–860.duct development. Development 125, 4163–4174.

Nilson, L.A., and Schupbach, T. (1999). EGF receptor signaling inKarim, F.D., and Rubin, G.M. (1998). Ectopic expression of activated Drosophila oogenesis. Curr. Top. Dev. Biol. 44, 203–243.Ras1 induces hyperplastic growth and increased cell death in Dro-

O’Neill, E.M., Rebay, I., Tjian, R., and Rubin, G.M. (1994). The activi-sophila imaginal tissues. Development 125, 1–9.ties of two Ets-related transcription factors required for Drosophila

Kassis, J.A. (1990). Spatial and temporal control elements of the eye development are modulated by the Ras/MAPK pathway. CellDrosophila engrailed gene. Genes Dev. 4, 433–443. 78, 137–147.Klaes, A., Menne, T., Stollewerk, A., Scholz, H., and Klambt, C. Postlethwait, J.H., and Schneiderman, H.A. (1971). A clonal analysis(1994). The Ets transcription factors encoded by the Drosophila of development in Drosophila melanogaster: morphogenesis, deter-gene pointed direct glial cell differentiation in the embryonic CNS. mination and growth in the wild type antenna. Dev. Biol. 24, 477–519.Cell 78, 149–160. Price, J.V., Savenye, E.D., Lum, D., and Breitkreutz, A. (1997). Domi-Klambt, C., Glazer, L., and Shilo, B.Z. (1992). breathless, a Drosophila nant enhancers of Egfr in Drosophila melanogaster: genetic linksFGF receptor homolog, is essential for migration of tracheal and between Notch and Egfr signaling pathways. Genetics 147, 1139–specific midline glial cells. Genes Dev. 6, 1668–1678. 1153.

Queenan, A.M., Ghabrial, A., and Schupbach, T. (1997). EctopicKramer, S., Okabe, M., Hacohen, N., Krasnow, M.A., and Hiromi, Y.

Egfr and Notch in Early Eye Development697

activation of torpedo/Egfr, a Drosophila receptor tyrosine kinase,dorsalizes both the eggshell and the embryo. Development 124,3871–3880.

Quiring, R., Walldorf, U., Kloter, U., and Gehring, W.J. (1994). Homol-ogy of the eyeless gene of Drosophila to the small eye gene in miceand Aniridia in humans. Science 265, 785–789.

Rebay, I., and Rubin, G.M. (1995). Yan functions as a general inhibitorof differentiation and is negatively regulated by activation of theRas1/MAPK pathway. Cell 81, 857–866.

Reichman-Fried, M., Dickson, B., Hafen, E., and Shilo, B.-Z. (1994).Elucidation of the role of breathless, a Drosophila FGF receptorhomolog, in tracheal cell migration. Genes Dev. 8, 428–439.

Riechmann, V., Irion, U., Wilson, R., Grosskortenhaus, R., and Leptin,M. (1997). Control of cell fates and segmentation in the Drosophilamesoderm. Development 124, 2915–2922.

Royet, J., and Finkelstein, R. (1996). hedgehog, wingless and ortho-denticle specify adult head development in Drosophila. Develop-ment 122, 1849–1858.

Royet, J., and Finkelstein, R. (1997). Establishing primordia in theDrosophila eye-antennal imaginal disc: the roles of decapen-taplegic, wingless and hedgehog. Development 124, 4793–4800.

Sawamoto, K., and Okano, H. (1996). Cell-cell interactions duringneural development: multiple types of lateral inhibitions involved inDrosophila eye development. Neurosci. Res. 26, 205–214.

Schnepp, B., Grumbling, G., Donaldson, T., and Simcox, A. (1996).Vein is a novel component in the Drosophila epidermal growth factorreceptor pathway with similarity to the neuregulins. Genes Dev. 10,2302–2313.

Scholz, H., Sadlowski, E., Klaes, A., and Klambt, C. (1997). Controlof midline glia development in the embryonic Drosophila CNS. Mech.Dev. 64, 137–151.

Schweitzer, R., and Shilo, B.-Z. (1997). A thousand and one rolesfor the Drosophila EGF receptor. Trends Genet. 13, 191–196.

Seimiya, M., and Gehring, W.J. (2000). The Drosophila homeoboxgene optix is capable of inducing ectopic eyes by an eyeless-inde-pendent mechanism. Development 127, 1879–1886.

Si Dong, P.D., Chu, J., and Panganiban, G. (2000). Coexpression ofthe homeobox genes Distal-less and homothorax determines Dro-sophila antennal identity. Development 127, 209–216.

Staehling-Hampton, K., Jackson, P.D., Clark, M.J., Brand, A.H., andHoffmann, F.M. (1994). Specificity of bone morphogenetic protein-related factors: cell fate and gene expression changes in Drosophilaembryos induced by decapentaplegic but not 60A. Cell Growth Diff.5, 585–593.

Tabata, T., and Kornberg, T.B. (1994). Hedgehog is a signaling pro-tein with a key role in patterning Drosophila imaginal discs. Cell 76,89–102.

Tata, F., and Hartley, D.A. (1995). Inhibition of cell fate in Drosophilaby Enhancer of split genes. Mech. Dev. 51, 305–315.

Tio, M., and Moses, K. (1997). The Drosophila TGFa homolog Spitzacts in photoreceptor recruitment in the developing retina. Develop-ment 124, 343–351.

Tomlinson, A., and Ready, D.F. (1987). Neuronal differentiation inthe Drosophila ommatidium. Dev. Biol. 120, 366–376.

Treisman, J.E. (1999). A conserved blueprint for the eye? Bioessays21, 843–850.

Treisman, J.E., and Heberlein, U. (1998). Eye development in Dro-sophila: formation of the eye field and control of differentiation. Curr.Top. Dev. Biol. 39, 119–158.

Wieschaus, E. (1974). Clonal anlaysis of early development in Dro-sophila melanogaster. PhD Thesis, Yale University, New Haven, CT,pp 1–147.

Younossi-Hartenstein, A., Tepass, U., and Hartenstein, V. (1993).Embryonic origin of the imaginal discs of the head of Drosophilamelanogaster. Roux’s Arch. Dev. Biol. 203, 60–73.

zur Lage, P., and Jarman, A.P. (1999). Antagonism of EGFR andNotch signalling in the reiterative recruitment of Drosophila adultchordotonal sense organ precursors. Development 126, 3149–3157.