homothorax and iroquois-C genes are required for the establishment of territories within the developing eye disc

Mechanisms of Development 96 (2000) 15±25

www.elsevier.com/locate/modo

homothorax and iroquois-C genes are required for the establishment of territories within the developing eye disc Franck Pichaud a, Fernando Casares b,* a

Laboratory of Molecular Genetics, Department of Biology, New York University, 1009 Main Building, 100 Washington Square East, New York, NY 10003, USA b Department of Biochemistry and Molecular Biophysics, Columbia University College of Physicians and Surgeons, HHSC 1108, 701West 168th Street, New York, NY 10032, USA Received 1 March 2000; received in revised form 19 May 2000; accepted 22 May 2000

Abstract In Drosophila the eye-antennal disc gives rise to most adult structures of the ¯y's head. Yet the molecular basis for its regionalization during development is poorly understood. Here we show that homothorax is required early during development for normal eye development and is necessary for the formation of the ventral head capsule. In the ventral region of the disc only, homothorax and wingless are involved in a positive feedback loop necessary to restrict eye formation. homothorax is able to prevent the initiation and progression of the morphogenetic furrow without inducing wingless, which points to homothorax as a key negative regulator of eye development. In addition, we show that the iroquois-complex genes are required for dorsal head development antagonizing the function of homothorax in this region of the disc. q 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Drosophila; homothorax; Head cuticle; Eye; iroquois-complex; wingless; Dorsoventral axis; Eye-antennal disc; Development

1. Introduction The different territories of the adult Drosophila head are derived from the eye-antennal imaginal discs. The embryological origin of the eye-antennal disc is complex, as it seems to be composed from cells coming from several head segments (Younossi-Hartenstein and Hartenstein, 1993). The eye-antennal disc can be subdivided into the antennal, maxillary palp and eye-head regions. The latter is composed of the eye epithelium, surrounded by the head capsule prospective region, and the peripodial membrane. The fact that the eye and the head cuticle arise from the eye-head ®eld poses the problem of how these two different territories are allocated. In addition, further regionalization of the disc must take place in order to generate all the distinct head structures. The current view of the morphogenesis of the eye proposes that eyeless (ey) (or maybe twin of eyeless (toy) upstream) de®nes eye identity within the eye-head region (Callaerts et al., 1997; Desplan, 1997). How much of the disc is going to be assigned to be eye, and how much to be * Corresponding author. Tel.: 11-212-305-2111; fax: 11-212-305-7924. E-mail addresses: [email protected] (F. Pichaud), [email protected] (F. Casares).

head depends on the interactions between `eye-triggering' genes, such as decapentaplegic (dpp) and hedgehog (hh) (Spencer et al., 1982; Heberlein et al., 1993; Ma et al., 1993; Wiersdorff et al., 1996; Chanut and Heberlein, 1997; Dominguez and Hafen, 1997; Pignoni and Zipursky, 1997), and `eye-blocking' genes, such as wingless (wg) (Ma and Moses, 1995; Treisman and Rubin, 1995) (for a review, see Treisman and Heberlein, 1998). It has been recently shown that removal of the homeobox gene homothorax (hth) in the ventral head region causes the formation of ectopic eye tissue, while its ectopic expression, using a dppGal4 driver, leads to an eyeless phenotype (Pai et al., 1998). hth is expressed, in the late disc, in the dorsal and ventral regions of the prospective head capsule, surrounding the eye primordium (Rieckhof et al., 1997; Pai et al., 1998). These results show that the expression of hth is incompatible with eye development and suggest that hth might be involved in restricting the developing eye ®eld. However, a more detailed analysis of its expression and function during the development of the eye disc is still lacking, and the mechanisms that regulate hth expression or those by which hth prevents eye formation are unknown. Once part of the disc epithelium has been determined as prospective eye by ey, a wave of cell differentiation, called `morphogenetic furrow' (MF), sweeps across it in a poster-

0925-4773/00/$ - see front matter q 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0925-477 3(00)00372-5

16

F. Pichaud, F. Casares / Mechanisms of Development 96 (2000) 15±25

ior to anterior direction. The MF is initiated at the posterior edge of the disc by decapentaplegic (dpp) and hedgehog (hh) (Dominguez and Hafen, 1997; Borod and Heberlein, 1998). Sequential rows of cells start to express dpp, to later differentiate and express hedgehog (hh). Thus, a relay mechanism is initiated when hh induces dpp in more anterior cells. The cells that are reached by the furrow enter coordinately the differentiation pathway leading to the different eye-speci®c cell types (reviewed in Heberlein and Moses, 1995). In order to break the symmetry of the disc, a system of positional axes has to be created. The iroquois-complex (iroC) genes araucan (ara), caupolican (caup) (GoÂmez-Skarmeta et al., 1996; Cavodeassi et al., 1999) and mirror (mirr) (McNeill et al., 1997) are expressed only in the dorsal half of the prospective eye (McNeill et al., 1997; Dominguez and de Celis, 1998). iro-C genes have been shown to specify the polarity of the ommatidia in the dorsal eye as well as to participate in the formation of the equator, a midline that separates the dorsal and ventral halves of the eye, where the Notch (N) signaling pathway is activated (Cho and Choi, 1998; Dominguez and de Celis, 1998; Papayannopoulos et al., 1998; Cavodeassi et al., 1999). Nevertheless, a possible role for this gene family in specifying head capsule structures has not been described to date. In the present work we have tried to better understand the mechanisms leading both to the de®nition of the mutually exclusive eye and head capsule identities, and to the speci®cation of the distinct regions of the head capsule. For this purpose, we have analyzed the role of the hth, wg and iro-C genes in these processes by experimentally altering their expression. We show that hth is expressed early in development in all the cells of the prospective eye-head region. This early expression is required for the normal organization of the adult eye. Later, hth expression becomes restricted to the anterior part of the developing disc and to the future dorsal and ventral prospective head capsule regions. Only in the ventral region hth is required for wg maintenance and for the suppression of eye development. Furthermore, wg upregulates hth expression in both the ventral and dorsal head presumptive regions. Therefore, in the prospective head capsule, wg and hth establish a positive feedback loop. In the eye region, hth is able to stop MF progression without wg induction, suggesting that hth could be mediating some of the wg functions in the eye-head region. Since hth is expressed both dorsally and ventrally but ectopic eyes are only formed when hth is lost ventrally, we have analyzed whether the iro-C genes, that are expressed in the dorsal eye prospective region (Cavodeassi et al., 1999; DõÂez del Corral et al., 1999) have a role in setting up this dorsal-ventral difference. We show that the iro-C genes are also expressed in the dorsal head capsule, and that the loss of their function results, not only in dorsal expansion of the eye, but in the transformation of the dorsal head into ventral structures, while hth expression remains unchanged.

2. Results 2.1. hth expression and requirement are widespread in the eye-antennal disc The eye-antennal disc is a compound imaginal disc that gives rise to several different parts of the ¯y head (Haynie and Bryant, 1986): the head capsule (ventral and dorsal, including the ocellar region), that surrounds the eye, plus the antenna and maxillary palp. Because the gene homothorax (hth) has been implicated in limiting the eye ®eld (Pai et al., 1998), we examined its pattern of expression throughout the development of the eye-antennal disc and analyzed in detail its requirements in the head. Early in development, during second instar larval stage, hth is weakly expressed in all the cells of the eye-antennal disc (Fig. 1A). The same widespread expression pattern is seen for wg (Royet and Finkelstein, 1996, and Fig. 1). During the third instar larval stage, as the eye ®eld is patterned in a posterior-to-anterior direction, the expression of hth regresses anteriorly and laterally. By late third instar, hth remains strongly expressed in the prospective head capsule, antennal regions and more weakly in the maxillary palp primordium (Fig. 1B; Pai et al., 1998). In the developing eye ®eld, hth is expressed 10±15 cell diameters ahead of the morphogenetic furrow (MF), but it is switched off thereafter. In the differentiated eye hth is found in the pigment cells in the posterior region of the eye ®eld (not shown). First we examined the role of hth in the head structures using mosaic analysis. Consistent with hth expression pattern in late third instar stage, hth mutant clones induced at any time during larval life autonomously produced ectopic eyes only in the ventral head capsule (Pai et al., 1998 and Fig. 1C,D). The frequency and size of ectopic eyes is generally greater with hthB2, a weak allele, than with strong alleles such as hthC1 or hthP2 (see Section 4). The ventral head region (gena and rostral membrane) is reduced as a consequence of the production of ectopic eyes (Fig. 1D), and the maxillary palps are frequently absent or abnormal in hth 2 clones (not shown). Conversely, ectopic expression of hth in clones in the eye can cause the eye to be split by tissue resembling ventral head capsule (Fig. 1F). When hth 2M 1 clones are induced during ®rst instar larval stage, large mutant clones in the head can be recovered. These clones result in ventral overgrowths of eye tissue (Fig. 1E) and in the loss of ventral and dorsal head structures (not shown). Nevertheless, the rim of cuticle surrounding the eye (the orbital and postorbital area) is preserved, indicating that other factors are responsible for the speci®cation of these structures. hth 2 clones induced later in development cause the autonomous transformation of the dorsal part of the head capsule into mesonotum (dorsal second thoracic segment). hth 2 clones affecting dorsal head regions never induce ectopic eyes. These phenotypes are equivalent to those described for the loss of extradenticle (exd) (GonzalezCrespo and Morata, 1995), consistent with a role of hth in

F. Pichaud, F. Casares / Mechanisms of Development 96 (2000) 15±25

the nuclear localization of the Exd protein (Rieckhof et al., 1997; Pai et al., 1998). In addition, we and others have also shown that mutant clones in the antenna cause its autonomous transformation to leg tissue (Casares and Mann, 1998; Pai et al., 1998) (Fig. 1E). We analyzed if a particular cell type was absent in mutant hth 2 clones in eyes of living ¯ies (Fig. 1G±I) and in tangential section of eyes containing clones (data not shown). We occasionally observed incomplete ommatidia with missing photoreceptor and fewer bristles, most of the times at the interface between clonal and wild-type tissue in the mosaic eyes (not shown). In addition, hth mutant clones in the eye are normally pigmented, indicating that pigment cells are present, even though hth is expressed in this cell type. In hth mosaic eyes the ommatidia frequently show orthogonal

17

shape (not shown), instead of the normal hexagonal one. This shape is likely to be the result of incorrect cell stacking, probably due to disorganized ommatidia. Interestingly, in hthB2 mosaic eyes we observed the induction of ectopic equators (Fig. 1G±I) occurring parallel to the endogenous equatorial axis and accompanied by frequent inversion of ommatia polarity along the anterior± posterior (A/P) axis (Fig. 1G,H). These defects were both cell- and non-cell-autonomous. 2.2. hth is required for wg expression in the head region ventrally, but not dorsally The progression of the MF orchestrates a wave of cell differentiation that gives rise to the different cell subtypes

Fig. 1. hth is required for the proper development of both head capsule and eye tissues. (Ai±iii) Second instar eye-antennal imaginal disc stained for Hth (green, ii) and Wg (red, iii). The merged image is shown in (i). At this stage, all cells of the disc express Hth and Wg. (B) Late third instar eye-antennal imaginal disc, stained for Hth (green) and Elav (white), a marker of neuronal differentiation that de®nes the eye ®eld. The following structures express Hth at this stage: maxillary palp (mp), antenna (a), rostral membrane (rmb), orbital region (orb). The ocellar region (oc) is devoid of strong Hth stain at this stage. In the eye (e) Hth is detected only in the posterior pigment cells (not shown in this focal plane). (C±E) SEM micrographs of hth 2 mosaic heads. (C) hthC1 ectopic clonal ventral eye (arrowhead). (D) hthB2 ectopic clonal ventral eye (arrowhead). Ventral view. (E) hthP2 (Minute1) head. Lateral view. (e) marks the eye. The arrowhead points to the eye overgrowth, and the arrow points to an antenna transformed to leg. (F) Overexpression of hth in the head in clones causes splitting of the eye (e) by a cuticle with bristles resembling gena (asterisk). This clone was induced during ®rst instar stage. (G±I) hthB2 clone examined under ¯uorescence light (outlined in white, G) within the eye shows ommatidial polarity defects. (H) Close-up of the same clone showing inconsistent D/V and A/P polarity. (I) is a map of the eye shown in (G) and (H), containing hthB2 clones (white areas). The normal (green arrowhead) and ectopic equators are indicated as green lines. The gray box in (I) highlights the area magni®ed in (H). The normal dorsal and ventral chiral forms are represented as blue and yellow arrows, respectively. Both dorsal and ventral clones induce mispolarization of ommatidia.

18

F. Pichaud, F. Casares / Mechanisms of Development 96 (2000) 15±25

found in the adult eye (for a review see Heberlein and Moses, 1995). The product of the gene wingless (wg) limits the expansion of the differentiating eye, allowing head capsule development (Ma and Moses, 1995; Treisman and Rubin, 1995; Royet and Finkelstein, 1996). Since hth also limits eye development in the ventral head, we tested if hth and wg regulation is linked. In third instar eye-head region, hth is strongly expressed in the prospective regions of the dorsal and ventral head, where it overlaps with wg expression (Fig. 2A±D). Large ventral hth 2M 1 mutant clones induced in the prospective head capsule region cause the formation of ectopic eyes, visualized by the de novo expression of Elav, and the loss of wg expression (Fig. 2E±H). Ectopic eye differentiation starts at the margins and progresses inwards, in agreement with the ectopic eye tissue seen in the adult (see above). By contrast, large dorsal hth clones do not produce ectopic eyes and wg expression is not affected (Fig. 2I±L). Nonetheless, removal of hth from the dorsal head results in the loss of the ocellar region. It has been shown that this region requires the expression of wg and its downstream target orthodenticle (otd; Royet and Finkelstein, 1997). Thus, it is possible

that, in the absence of hth, wg expression is subtly reduced. Alternatively, hth could act in parallel to, or downstream of wg to specify this dorsal head region. These results suggest that, in the ventral part of the eyehead region, hth helps de®ning the territory of the disc that will become head, probably in part by maintaining wg, but not in the dorsal region. In the absence of hth, wg is lost ventrally and ectopic eyes are generated. 2.3. wg and hth are engaged in a positive feedback loop As the eye ®eld is patterned, hth expression is repressed several rows of cells ahead of the MF (Fig. 1A,B, and data not shown), and it is upregulated at the margins of the disc. This dynamic pattern of expression resembles wg expression, which is detected in all the cells of the eye-antennal disc during second instar larval, but is later restricted to the margins (Royet and Finkelstein, 1997) (see also Fig. 1A). During late third instar, hth expression straddles that of wg (Fig. 2A,B). This observation raised the possibility that hth could be also controlled by wg. To test this hypothesis, we generated clones ectopically expressing a membrane-teth-

Fig. 2. hth is required for maintaining wg ventrally, but not dorsally. wg-Z, red; Hth, green; Elav, white. wg-Z imaginal disc ((A) is a proximal confocal plane, and (B) is a distal one), showing the overlap between the wg-Z (red) and Hth (green) expression domains, seen as yellow color, in the dorsal (arrowhead) as well as ventral (arrow) head prospective regions. The wg-Z and Hth individual panels are shown in (C) and (D), respectively. The lines in (B) mark the approximate position of the morphogenetic furrow. (E±H) Ventral and (I±L) dorsal hthP2 (Minute1) clones in wg-Z animals. (E,F) and (I,J) are two focal planes of the same disc, respectively. (G) and (K) are the red channels in (E) and (I), respectively, and (H) and (L) are the green channels of the same discs, respectively. The arrow points to the ventral region, the arrowhead to the dorsal one. The asterisk in (E±H) shows a ventral eye outgrowth. Note that in (E±G), the ventral wg-Z expression is absent.

F. Pichaud, F. Casares / Mechanisms of Development 96 (2000) 15±25

ered Wg form (teth-Wg; Zecca et al., 1996), that cannot diffuse. In teth-Wg expressing clones located anterior of the furrow, Hth levels were increased (Fig. 3A±C, arrows). This was not the case when the clones were induced in regions immediately in front of or posterior to the furrow (Fig. 3A±C, arrowheads). We also tested if blocking the wg pathway could lead to modi®cations of hth expression. We addressed this issue by producing ectopic clones of a dominant negative form of dTCF, a nuclear factor required for the transduction of the wg signal (Brunner et al., 1997; Riese et al., 1997; van de Wetering et al., 1997; Cadigan et al., 1998). In these clones hth expression is strongly reduced in the presumptive head cuticle region both ventrally and dorsally (not shown). This results show that wg is necessary to maintain hth expression in the presumptive head regions of the eye-head disc. Conversely, when hth was expressed ectopically in clones, it was unable to initiate wg expression (Fig. 3D,E, arrowheads). However, ectopic hth upregulated wg in regions

19

where it (and hth) was already expressed (Fig. 3D,E, arrow). We also noticed that ectopic expression of hth can block furrow initiation without inducing wg (Fig. 3D,E, asterisk). These observations raise the possibility that hth mediates, at least partially, the eye-repressing function of wg. These experiments suggest that hth is necessary to maintain wg expression, but not suf®cient for its de novo induction. Starting during late second or early third instar larval stage wg and hth seem to be engaged in a positive regulatory feedback loop that might be important for the development of the ventral head capsule. This feedback loop would be responsible for the upregulation of hth in the ventral and dorsal head capsule while hth is required to maintain wg only in its ventral part. 2.4. hth can block furrow initiation and progression downstream of dpp hth is able to block eye differentiation (Pai et al., 1998, and this study) in a wg-independent manner. One possibility is that hth acts by repressing dpp, which is expressed in the furrow (Blackman et al., 1991). Alternatively, hth could perform its function through the repression of genes downstream of dpp. To test this possibilities we generated ectopic clones of cells over-expressing hth and analyzed their effect on both dpp expression and furrow progression. Small hth expressing clones in the furrow do not repress dpp (Fig. 4A± C, arrowhead), and do not allow photoreceptor differentiation (reported by the neuronal speci®c marker Elav; Fig.4A,D). hth expressing clones induced just anterior to the furrow delay furrow propagation: the hth 1 cells receive signals from the furrow, since they are able to turn on dpp-Z expression, and the furrow is able to advance over them, leaving some hth 1 cells behind it. Nevertheless, the furrow is retarded and the hth 1 cells posterior to it do not differentiate as photoreceptors (Fig. 4E±H). These results show that hth can block MF movement downstream of dpp. 2.5. iro-C genes repress ventral differentiation in the dorsal head capsule

Fig. 3. hth is upregulated by wg. (A) High magni®cation confocal view of a wild-type eye disc, showing the eye ®eld (Elav-positive: purple) and the front of the furrow (yellow dashed line). Hth is in green. Enclosed between the white and yellow dashed lines is the area devoid of Hth. (B,C) Clones overexpressing Wg-teth were induced at 72±96 h AEL and are marked by the Flu epitope which tags the Wg-teth protein (red). These clones are frequently associated with an upregulation of Hth (green) expression (arrows) and tissue overgrowth if they lie anterior to the furrow (between the white and yellow dashed lines), but do not upregulate Hth signi®cantly if they are within the furrow (arrowheads). (B) Merged image. (C) Hth (green) channel. (D,E) Overexpression of hth does not turn on wg expression. Larvae in which clones have been induced during late second-early third instar stages (,72 h AEL) were dissected. GFP-Hth, green; Wg, red; Elav, white. Hth-expressing clones do not express wg ectopically (arrowheads). In the case marked by an arrow, wg is upregulated close to its normal domain of expression. A clone at the posterior margin of the disc blocks the morphogenetic furrow initiation, as seen by the absence of Elav 1 differentiating photoreceptors (asterisk).

hth seems to be mainly involved in the maintenance of wg and the repression of eye development only at the ventral margin of the eye-head region. However, hth is expressed both ventrally and dorsally (Fig. 1A,B). Therefore, there must be other genes involved in setting up this asymmetry in the head. Candidates for such factors are the iroquoisComplex (iro-C) homeobox genes araucan (ara), caupolican (caup) (GoÂmez-Skarmeta et al., 1996), and mirror (mirr) (McNeill et al., 1997), because their expression is restricted to the dorsal half of the eye primordium. The iro-C genes seem to position the equator, a narrow domain where the Notch (N) signaling pathway is activated and triggers initiation of eye differentiation (McNeill et al., 1997; Dominguez and de Celis, 1998; Papayannopoulos et al., 1998; Cavodeassi et al., 1999). Also, they have been

20

F. Pichaud, F. Casares / Mechanisms of Development 96 (2000) 15±25

Fig. 4. hth blocks furrow movement downstream of dpp. Clones ectopically expressing GFP-Hth (green) were induced between 48 and 72 h AEL in a dpp-Z background. Only the eye-head portion is shown. Anterior is to the left. Hth was visualized with the GFP signal. (A±D) GFP-Hth clone (arrowhead) in the furrow does not repress dpp-Z (C). Elav is not expressed in the GFP 1 cells (D). The merged image is in (A) and the GFP-Hth channel is in (B). (E±H) GFP-Hthexpressing cells induced anterior to the furrow (green), delay furrow propagation, visualized by dpp-Z (G), even though some of the cells in the clone express dpp-Z. Elav is repressed in the clones (H). The merged image is in (E) and the GFP-Hth channel is in (F).

recently shown to be under the control of pannier (pnr), a GAGA-family transcription factor-encoding gene (Ramain et al., 1993; Maurel-Zaffran and Treisman, 2000). First, we mapped the expression of the ara/caup reporter rF209 (GoÂmez-Skarmeta et al., 1996; DõÂez del Corral et al., 1999) in the dorsal eye-head primordium relative to hth, and to pannier (pnr), which is expressed in the dorsal-most region of the head capsule (Calleja et al., 1996). In late third instar eye-antennal discs, hth is expressed in the dorsal fold that gives rise to the dorsal head capsule (Fig. 5A) and in the peripodial membrane (Pai et al., 1998 and Fig. 5B). The domain of hth expression overlaps with pnr (monitored with a Gal4 insertion line in this gene (Fig. 5A,B; Calleja et al., 1996) in the dorsalmost part of the head capsule region, and with iro-C more ventrally (Fig. 5A,B) in a thin strip of cells. At this larval stage, iro-C and pnr-Gal4 do not overlap, and wg is expressed in the iro-C domain (Fig. 5C). In order to map the hth/iro-C co-expression domain in discs to the adult head structures, we stained rF209 adult heads in Xgal solution. We observe staining in the orbital region of the dorsal head (not shown), in agreement with the fate map of the eye-antennal disc by Haynie and Bryant (Haynie and Bryant, 1986) (Fig. 6A,B). Since in hth- clones the orbital region is unaffected, the iro-C genes could be determining this dorsal head structure. To analyze the role of iro-C in dorsal head development, we removed iro-C function in clones of cells carrying a de®ciency for ara, caup and mirr (iroDFM3), or only ara and caup (iroDFM1) (DõÂez del Corral et al., 1999), and studied the phenotypic consequences in adult heads. Only results for iroDFM3 clones will be described, since iroDFM1 clones give similar results. We observe that iroC 2 clones cause a series of phenotypes, adding progres-

sively more `ventral-type' tissue in the following order: dorsal eye overgrowth or ectopic dorsal eyes, overgrowth of ventral type of cuticle (ptilinum and rostral membrane), ectopic antennal pouches, antennae and maxillary palps (Fig. 5D±G). The extra head structures are produced autonomously (labeled with the cell marker mwh), but the eyes can be composed of both mutant and wild-type ommatidia (Cavodeassi et al., 1999, and data not shown). The ectopic structures, which can duplicate the full complement of ventral structures, all grow from the orbital region of the head. The rest of the dorsal head is displaced by the overgrown tissue. As mentioned earlier, the orbital region fate maps to the domain where hth and iro-C expressions overlap. These results show that the iro-C genes are required to repress the proliferation of a group of dorsal cells, that otherwise would grow with ventral head identity, and may contribute to assign them a dorsal head (`orbital-region') identity. iro-C 2 clones in discs frequently produce overgrowths, in agreement with the structures produced dorsally. However, hth and wg expressions are not substantially altered in iroC 2 clones (not shown). Since removal of ara, caup and mirr produces the same phenotypes as removal of only ara and caup, we conclude that mirr is dispensable for suppressing ventral identity in the dorsal head. Alternatively, mirr expression could be under the control of ara and caup.

3. Discussion 3.1. A genetic network to control the eye-head development The mechanisms that control the regionalization of the

F. Pichaud, F. Casares / Mechanisms of Development 96 (2000) 15±25

21

Fig. 5. iro-C genes are expressed in the prospective dorsal head capsule and repress ventral head fates. (A,B) Eye±head region of an eye-antennal imaginal disc of an UAS-GFP; pnr-Gal4/rF209 late third instar larva. (B) is a vertical confocal section of the disc in (A) along the white line. Hth, green; b -galactosidase (rF209, iro-C-LacZ enhancer trap), red; pnr-Gal4, blue. The domain of overlap between Hth and rF209 appears yellow (arrowhead), and that between Hth and pnr-Gal4 appears pale blue. pmb, peripodial membrane. (C) rF209 imaginal disc stained for b -galactosidase (red) and for Wg (white). The arrowhead points to the area of Hth and rF209 overlap. Removal of iro-C function produces dorsal outgrowths of ventral-like identity. (D±G) SEM micrographs of heads containing iro-C- clones, showing increasingly stronger phenotypes by addition of new ectopic ventral structures. (1) Ectopic eye, (2) ventral head-like cuticle, (3) antenna and (4) maxillary palp. The antennal pocket is arrowed. a, normal antenna; e, normal eye; oc, ocelli. All the dorsal-to-ventral transformations generated in iroC 2 mosaics are cell autonomous (marked with mwh), except for the ectopic eye tissue.

22

F. Pichaud, F. Casares / Mechanisms of Development 96 (2000) 15±25

presumptive eye-head region in Drosophila are poorly understood. Here we show that hth is required for eyehead development and for wg expression in the prospective ventral head capsule. In addition, iro-C genes are responsible for repressing ventral identity in the dorsal head region. Early on in the development of the eye-antennal disc, during second larval instar stage, hth is expressed at low levels in all cells. Through the second to third instar larval stages, hth expression gets restricted to the most anterior part of the eye territory, and to the dorsal and ventral poles of the eye-head region, and to the peripodial membrane. This process seems to be a direct consequence of the dpp expression at the MF, as ectopic expression of a constitutive form of the dpp receptor thick veins (tkv) leads to an autonomous repression of hth (data not shown). In this context, hth could be involved in the control of the MF progression. In addition, hth is required for the maintenance of wg at the ventral pole of the head prospective region, while dorsally wg maintenance is independent of it. 3.2. Differential regulation of wingless

Fig. 6. Domains of gene expression in the ¯y's head and model of genetic interactions taking place during its development. (A) Lateral and (B) frontal±dorsal views of the adult head showing the different gene expression domains analyzed in the present work, as they fate map in the schematic eye-antennal disc shown in (C). (D) Vertical cross section of the eye-head region. Abbreviations as in Fig. 1. hth, green, wg, dark blue, iro-C, red. The overlap between iro-C and hth is brown. The overlap between pnr and hth is turquoise. The gray bar represents the morphogenetic furrow. The purple bar represents the equator. (E,F) Model for the regulation of hth and wg in the eye-head region of the eye-antenna imaginal disc. (E) Early third instar disc; hth (green) and wg (dark blue pattern) are expressed widespread. The initiation of the furrow causes the repression of hth and wg. Otherwise, hth would block the initiation and progression of the furrow. hth does not need wg for it. Later (F), the advancing furrow pushes wg and hth to the lateralanterior regions of the disc, which will become the head capsule. wg can act as an upregulator of hth in this regions, while hth is required for wg maintenance only ventrally. Other factors must keep wg expression on dorsally. Such a factor is most likely pnr (Maurel-Zaffran and Treisman, 2000). We do not know if pnr also activates hth, or if hth contributes to wg expression in the dorsal presumptive head (dotted arrows and question mark). hth is also repressed in front of the furrow.

In the dorsal region, wg has been recently shown to be under the control of pnr (Maurel-Zaffran and Treisman, 2000). In an analogous manner, we have shown that hth must be kept on in the ventral margin of the disc to stop eye differentiation and to maintain wg expression in this region. wg is expressed at a higher level in the dorsal region of the presumptive head cuticle than in the ventral region. Removal of wg from the dorsal edge of the disc results in ectopic furrow generation, whereas this effect is less penetrant in the ventral part of the disc (Treisman and Rubin, 1995). In that respect, the removal of hth in the ventral part of the disc `mimics' the effect of removing wg dorsally, by producing large eye overgrowths. One possibility is that both wg and hth have to be removed from the ventral edge of the disc in order to ef®ciently generate ectopic furrow, suggesting that wg might not be the only activator of hth expression in the prospective ventral head region. hth, downstream of or in parallel to wg, would then repress eye formation in this region of the disc. In our model (Fig. 6F) hth maintains wg in the ventral region of the disc, while pnr maintains wg dorsally (Maurel-Zaffran and Treisman, 2000). In turn, wg upregulates hth expression in both the dorsal and the ventral head prospective regions of the disc. hth expression in the eye ®eld may be independent of any wg input. Our analysis of hth- clones has also revealed the formation of ectopic equators and inversion of ommatidia polarity along the A/P axis (see Fig. 1G±I). These observations are consistent with the role of hth in limiting eye formation by repressing MF triggering (Chanut and Heberlein, 1995). We have shown that repression of hth is a prerequisite to allow neuronal differentiation. In clones expressing ectopic hth, dpp is still expressed. Therefore, it is likely is that hth prevents eye formation downstream of dpp. This could

F. Pichaud, F. Casares / Mechanisms of Development 96 (2000) 15±25

happen through a disruption of the hh/dpp feedback loop by a direct repression of hh, or through an inhibition of proneural genes such as atonal. 3.3. Iroquois-complex genes confer dorsal head identity by suppressing ventral fates Recent studies have pointed out the essential role of iro-C genes in triggering the N pathway in the dorsal±ventral (D/ V) boundary of the eye, located at the interface of iro1/irocells (Cho and Choi, 1998; Dominguez and de Celis, 1998; Papayannopoulos et al., 1998). Therefore, iro-C genes are involved in the D/V regionalization of the eye. Both the pattern of expression of iro-C in the dorsal presumptive head cuticle and the generation of ectopic eyes in the dorsal head in iro-C 2 clones suggested that these genes are good candidates for wg maintenance dorsally. However, we have shown that removal of the iro-C genes ara and caup does not result in a loss of wg expression in the developing dorsal head. It has been recently proposed that pnr activates wg in the dorsal region of the disc. In turn, wg activates mirr, a iroC gene (Maurel-Zaffran and Treisman, 2000; Heberlein et al., 1998). Thus pnr would be the most upstream gene in the hierarchy controlling dorsal eye-head disc development (Maurel-Zaffran and Treisman, 2000). Here we have shown that, besides the generation of ectopic eye tissue (Cavodeassi et al., 1999), the loss of iro-C function leads to duplications of ventral head structures, such as antennae and maxillary palps, in the dorsal head. Interestingly enough, the development of these appendages require hth expression (Casares and Mann, 1998; Pai et al., 1998; this work). These results, together with the overlapping expression of hth and iro-C genes, that we have mapped to the presumptive dorsal head cuticle, led us to test the possible regulatory relation between iro-C and hth and wg in that region of the disc. We have observed that neither hth nor wg expressions are signi®cantly modi®ed in the absence of iro-C function. Therefore, our results show that a major role of iro-C is to impose dorsal fate upon a ventral identity program in the head cuticle. In the absence of iro-C, the hth expressing cells of the dorsal head cuticle undergo a massive cell proliferation, generating the ventral-like duplications in this region. The results presented in this report are summarized in Fig. 6. 4. Experimental procedures 4.1. Fly stocks hthB2 is a weak allele (Rieckhof et al., 1997), hthC1 (Rieckhof et al., 1997) and hthP2 (Pai et al., 1998) are strong alleles of homothorax (hth). In hthC1 and hthP2 clones in imaginal discs we do not detect Hth antigen using our anti-Hth antibody (Casares and Mann, 1998). For the generation of the clones we have used the

23

hs¯p122 (provided by G. Struhl) and the y ey¯p (provided by B. Dickson; Newsome et al., 2000) as sources of ¯ipase. The ey-¯p transgene drives ¯ipase expression widespread in the eye-antennal disc. iroDFM1 is a de®ciency uncovering the araucan (ara) and caupolican (caup) genes; iroDFM3 is a de®ciency uncovering the whole Iroquois-Complex (iroC) (ara, caup and mirror (mirr)) (DõÂez del Corral et al., 1999). rF209 is a Lac-Z reporter of araucan and caupolican in the iro-C (GoÂmez-Skarmeta et al., 1996). dpp-Z (Blackman et al., 1991) and wg-Z (Kassis et al., 1992) are dpp and wg LacZ reporters, respectively. UAS-hth8 (Casares and Mann, 1998). UAS . CD2, y 1 . teth-wg (Zecca et al., 1996), y hs¯p122, Act . hsCD2 . Gal4 (Pignoni and Zipursky, 1997), Arm-Gal4 (AG11), UAS-dTCF-DN and UAS-GFP were provided by G. Struhl. pnr-Gal4 is a Gal4 insertion in pnr (Calleja et al., 1996). 4.2. Mosaic analysis 4.2.1. Generation of hth 2 clones (1) FRT82B hth 2/TM2 ¯ies were crossed to yw, hs¯p122; FRT82B CD2, y1/TM2 or to yw, hs¯p122; FRT82B CD2, y1, M/TM2 ¯ies (for hth 2 Minute 1clones). In the Minute technique, the mutant clones are M1/1, which gives them a growth advantage relative to the surrounding M2/1 cell population (Morata and Ripoll, 1975), therefore resulting in large clones. Twenty-four-hour embryo collections were heat-shocked at 24±48, 48±72 or 72±96 h after egg laying (AEL). (2) yw ey¯p; FRT82B P(w1) ¯ies were crossed to FRT82B hth 2 ¯ies. 4.2.2. Generation of iro-C 2 clones (1) yw ey¯p; FRT80A P(w1) were crossed to yw; mwh iro DFM3 FRT80A/TM3 or yw; mwh iro DFM1 FRT80A/TM3 ¯ies. (2) hs¯p122; FRT80A arm-Z ¯ies were crossed to yw; mwh iro DFM3 FRT80A/TM3 or yw; mwh iro DFM1 FRT80A/ TM3 ¯ies. Twenty-four-hour embryo collections were heat-shocked at 48±72 h AEL. 4.2.3. Generation of teth-Wg-expressing clones yw, hs¯p122; AG11 ¯ies were crossed to UAS . hsCD2, y1.teth-wg. Twenty-four-hour egg collections were heatshocked at 358C for 30 min at 48±72 h AEL. teth-Wg is an active membrane tethered form of Wg (Zecca et al., 1996). 4.2.4. Generation of hth-expressing clones y, hs¯p122, Act . hsCD2 . Gal4 females were crossed to UAS-GFP-hth8 males. Twenty-four-hour egg collections were heat-shocked at 358C for 30 min at 48±72 h AEL. 4.2.5. Generation of dTCF-DN-expressing clones y, hs¯p122, Act . hsCD2 . Gal4 females were crossed to UAS-dTCF-DN males. 24h egg collections were heatshocked at 35C for 30 min at 48-72h AEL.

24

F. Pichaud, F. Casares / Mechanisms of Development 96 (2000) 15±25

4.3. Immunocytochemistry Eye-antennal discs were prepared for immunocytochemistry using standard protocols. Primary antibodies used were: chicken anti-Hth, 1/500 (Casares and Mann, 1998); mouse anti-Wg, 1/30 (Brook and Cohen, 1996; from the Iowa University Hybridoma Bank); rat anti-Elav 1/20 (O'Neill et al., 1994; from the Iowa University Hybridoma Bank); rabbit anti-b -galactosidase 1/1000 (Cappel); mouse anti-CD2 1/500 (Serotec). Appropriate ¯uorescent secondary antibodies were from Jackson Labs, 1/250. 4.4. Microscopy For Nomarsky microscopy, heads were dissected in phosphate-buffered saline and mounted in Hoyer's mounting medium/acetic acid (1:1). For scanning electron microscopy (SEM), specimens were dehydrated in Ethanol, dried in a CO2 Critical-point drier and gold-coated. Imaging was performed in an Hitachi S-4700 Cold Field Emission scanning electron microscope. For ¯uorescence immunocytochemistry, samples were analyzed in a MRC1024 Biorad confocal microscope. w 2 clones in the eye (wild type and hth 2) were observed using green ¯uorescent light and a Texas red ®lter (Nikon). Ommatidia were visualized after neutralization of the cornea (Franceschini et al., 1981) using UV light. The principle of these methods will be described elsewhere (Pichaud and Desplan, in preparation). For eye sectioning, samples were prepared as described (Tomlinson and Ready, 1987) and counterstained with toluidine blue. Acknowledgements We are particularly grateful to Claude Desplan and Richard S. Mann for their support and advice. This work was performed with the support of grant NIH ROI EY 13012-01 in the Desplan Laboratory and NIH RO1 GM58575-01 in the Mann Laboratory. We thank Barry Dickson, Jessica Treisman, Ruth DõÂez del Corral, Florencia Cavodeassi and Gary Struhl for ¯y stocks, Jessica Treisman, Andrew Tomlinson, Florencia Cavodeassi and Sonsoles Campuzano for communicating results prior to publication, C. Desplan, R.S. Mann, Samantha Butler, Jessica Treisman and members of the Desplan and Mann laboratories for comments on the manuscript, and Robert Desalle, Angela Klaus and Julian Stark at the American Museum of Natural History, New York, for their help with the SEM experiments. F.P. and F.C. are HFSPO postdoctoral fellows. References Blackman, R.K., Sanicola, M., Raftery, L.A., Gillevet, T., Gelbart, W.M., 1991. An extensive 3 0 cis-regulatory region directs the imaginal disk

expression of decapentaplegic, a member of the TGF-beta family in Drosophila. Development 111, 657±666. Borod, E.R., Heberlein, U., 1998. Mutual regulation of decapentaplegic and hedgehog during the initiation of differentiation in the Drosophila retina. Dev. Biol. 197, 187±197. Brook, W.J., Cohen, S.M., 1996. Antagonistic interactions between wingless and decapentaplegic responsible for dorsal±ventral pattern in the Drosophila leg. Science 273, 1373±1377. Brunner, E., Peter, O., Schweizer, L., Basler, K., 1997. pangolin encodes a Lef-1 homologue that acts downstream of Armadillo to transduce the Wingless signal in Drosophila. Nature 385, 829±833. Cadigan, K.M., Fish, M.P., Rulifson, E.J., Nusse, R., 1998. Wingless repression of Drosophila frizzled 2 expression shapes the Wingless morphogen gradient in the wing. Cell 5, 767±777. Callaerts, P., Halder, G., Gehring, W.J., 1997. PAX-6 in development and evolution. Annu. Rev. Neurosci. 20, 483±532. Calleja, M., Moreno, E., Pelaz, S., Morata, G., 1996. Visualization of gene expression in living adult Drosophila. Science 274, 252±255. Casares, F., Mann, R.S., 1998. Control of antennal versus leg development in Drosophila. Nature 392, 723±726. Cavodeassi, F., DõÂez del Corral, R., Campuzano, S., Dominguez, M., 1999. Compartments and organising boundaries in the Drosophila eye: the role of the homeodomain Iroquois proteins. Development 126, 4933± 4942. Chanut, F., Heberlein, U., 1995. Role of the morphogenetic furrow in establishing polarity in the Drosophila eye. Development 121, 4085± 4094. Chanut, F., Heberlein, U., 1997. Retinal morphogenesis in Drosophila: hints from an eye-speci®c decapentaplegic allele. Dev. Genet. 20, 197±207. Cho, K.O., Choi, K.W., 1998. Fringe is essential for mirror symmetry and morphogenesis in the Drosophila eye. Nature 396, 272±276. Desplan, C., 1997. Eye development: governed by a dictator or a junta? (comment). Cell 91, 861±864. DõÂez del Corral, R., Aroca, P., Cavodeassi, F., Modolell, J., 1999. The Iroquois homeodomain proteins are required to specify body wall identity in Drosophila. Genes Dev. 13, 1754±1761. Dominguez, M., de Celis, J.F., 1998. A dorsal/ventral boundary established by Notch controls growth and polarity in the Drosophila eye. Nature 396, 276±278. Dominguez, M., Hafen, E., 1997. Hedgehog directly controls initiation and propagation of retinal differentiation in the Drosophila eye. Genes Dev. 11, 3254±3264. Franceschini, N., Kirschfeld, K., Minke, B., 1981. Fluorescence of photoreceptor cells observed in vivo. Science 213, 1264±1267. GoÂmez-Skarmeta, J.L., del Corral, R.D., de la Calle-Mustienes, E., FerreMarco, D., Modolell, J., 1996. Araucan and caupolican, two members of the novel iroquois complex, encode homeoproteins that control proneural and vein-forming genes. Cell 85, 95±105. Gonzalez-Crespo, S., Morata, G., 1995. Control of Drosophila adult pattern by extradenticle. Development 121, 2117±2125. Haynie, J.L., Bryant, P.J., 1986. Development of the eye-antenna imaginal disc and morphogenesis of the adult head in Drosophila melanogaster. J. Exp. Zool. 237, 293±308. Heberlein, U., Moses, K., 1995. Mechanisms of Drosophila retinal morphogenesis: the virtues of being progressive. Cell 81, 987±990. Heberlein, U., Wolff, T., Rubin, G.M., 1993. The TGF beta homolog dpp and the segment polarity gene hedgehog are required for propagation of a morphogenetic wave in the Drosophila retina. Cell 75, 913±926. Heberlein, U., Borod, E.R., Chanut, F.A., 1998. Dorsoventral patterning in the Drosophila retina by wingless. Development 125, 567±577. Kassis, J., Noll, E., VanSickle, E.P., Odenwald, W.F., Perrimon, N., 1992. Altering the insertional speci®city of a Drosophila transposable element. Proc. Natl. Acad. Sci. USA 89, 1±1923. Ma, C., Moses, K., 1995. Wingless and patched are negative regulators of the morphogenetic furrow and can affect tissue polarity in the developing Drosophila compound eye. Development 121, 2279±2289.

F. Pichaud, F. Casares / Mechanisms of Development 96 (2000) 15±25 Ma, C., Zhou, Y., Beachy, P.A., Moses, K., 1993. The segment polarity gene hedgehog is required for progression of the morphogenetic furrow in the developing Drosophila eye. Cell 75, 927±938. Maurel-Zaffran, C., Treisman, J.E., 2000. pannier acts upstream of wingless to direct dorsal eye disc development in Drosophila. Development 127, 1007±1016. McNeill, H., Yang, C.H., Brodsky, M., Ungos, J., Simon, M.A., 1997. mirror encodes a novel PBX-class homeoprotein that functions in the de®nition of the dorsal±ventral border in the Drosophila eye. Genes Dev. 11, 1073±1082. Morata, G., Ripoll, P., 1975. Minutes: mutants of drosophila autonomously affecting cell division rate. Dev. Biol. 42, 211±221. Ê sling, B., Dickson, B.J., 2000. Analysis of Drosophila Newsome, T.P., A photoreceptor axis guidance in eye-speci®c mosaics. Development 127, 851±860. O'Neill, E.M., Rebay, I., Tjian, R., Rubin, G.M., 1994. The activities of two Ets-related transcription factors required for Drosophila eye development are modulated by the Ras/MAPK pathway. Cell 78, 137±147. Pai, C.Y., Kuo, T.S., Jaw, T.J., Kurant, E., Chen, C.T., Bessarab, D.A., Salzberg, A., Sun, Y.H., 1998. The Homothorax homeoprotein activates the nuclear localization of another homeoprotein, extradenticle, and suppresses eye development in Drosophila. Genes Dev. 12, 435±446. Papayannopoulos, V., Tomlinson, A., Panin, V.M., Rauskolb, C., Irvine, K.D., 1998. Dorsal±ventral signaling in the Drosophila eye. Science 281, 2031±2034. Pignoni, F., Zipursky, S.L., 1997. Induction of Drosophila eye development by decapentaplegic. Development 124, 271±278. Ramain, P., Heitzler, P., Haenlin, M., Simpson, P., 1993. pannier, a negative regulator of achaete and scute in Drosophila, encodes a zinc ®nger protein with homology to the vertebrate transcription factor GATA-1. Development 119, 1277±1291. Rieckhof, G.E., Casares, F., Ryoo, H.D., Abu-Shaar, M., Mann, R.S., 1997. Nuclear translocation of extradenticle requires homothorax, which encodes an extradenticle-related homeodomain protein. Cell 91, 171± 183.

25

Riese, J., Yu, X., Munnerlyn, A., Eresh, S., Hsu, S.C., Grosschedl, R., Bienz, M., 1997. LEF-1, a nuclear factor coordinating signaling inputs from wingless and decapentaplegic. Cell 6, 777±787. Royet, J., Finkelstein, R., 1996. hedgehog, wingless and orthodenticle specify adult head development in Drosophila. Development 122, 1849±1858. Royet, J., Finkelstein, R., 1997. Establishing primordia in the Drosophila eye-antennal imaginal disc: the roles of decapentaplegic, wingless and hedgehog. Development 124, 4793±4800. Spencer, F.A., Hoffmann, F.M., Gelbart, W.M., 1982. Decapentaplegic: a gene complex affecting morphogenesis in Drosophila melanogaster. Cell 28, 451±461. Tomlinson, A., Ready, D.F., 1987. Cell fate in the Drosophila ommatidium. Dev. Biol. 123, 264±275. Treisman, J.E., Heberlein, U., 1998. Eye development in Drosophila: formation of the eye ®eld and control. Curr. Top. Dev. Biol. 39, 119± 158. Treisman, J.E., Rubin, G.M., 1995. wingless inhibits morphogenetic furrow movement in the Drosophila eye disc. Development 121, 3519± 3527. van de Wetering, M., Cavallo, R., Dooijes, D., van Beest, M., van Es, J., Loureiro, J., Ypma, A., Hursh, D., Jones, T., Bejsovec, A., Peifer, M., Mortin, M., Clevers, H., 1997. Armadillo coactivates transcription driven by the product of the Drosophila segment polarity gene dTCF. Cell 6, 789±799. Wiersdorff, V., Lecuit, T., Cohen, S.M., Mlodzik, M., 1996. Mad acts downstream of Dpp receptors, revealing a differential requirement for dpp signaling in initiation and propagation of morphogenesis in the Drosophila eye. Development 122, 2153±2162. Younossi-Hartenstein, A., Hartenstein, V., 1993. The role of the tracheae and musculature during path®nding of Drosophila embryonic sensory axons. Dev. Biol. 158, 430±447. Zecca, M., Basler, K., Struhl, G., 1996. Direct and long-range action of a wingless morphogen gradient. Cell 87, 833±844.