The gene vestigial has been proposed to act as a master gene because of its supposed capacity to initiate and drive wing development. We show that the ectopic expression of vestigial only induces ectopic outgrowths with wing cuticular differentiation and wing blade gene expression patterns in specific developmental and genetic contexts. In the process of transformation, wingless seems to be an essential but insufficient co-factor of vestigial. vestigial ectopic expression alone orvestigial plus wingless co-expression in clones differentiate `mixed' cuticular patterns (they contain wing blade trichomes and chaetae characteristic of the endogenous surrounding tissue) and express wing blade genes only in patches of cells within the clones. In addition, we have found that these clones, in the wing imaginal disc, may cause autonomous as well as non-autonomous cuticular transformations and wing blade gene expression patterns. These non-autonomous effects in surrounding cells result from recruitment or `inductive assimilation' of vestigial orwingless-vestigial overexpressing cells.
INTRODUCTION
Morphogenesis is normally associated with cell proliferation and genetic specification of territories or tissues. Several types of regulatory genes define these genetic specifications. Thus, in Drosophila, `selector'genes such as Ultrabithorax (Ubx) or engrailed(en) (Garcia-Bellido,1975), in combination with other selector genes, confer identity to segments or compartments. These genes are expressed in clonally restricted territories, are cell autonomous in genetic mosaics and transform the tissue to an archetypal specification. Other types of genes, such as pannier(pnr) (Calleja et al.,2000; Calleja et al.,1996), iroquois (iro)(Diez del Corral et al., 1999;Gomez-Skarmeta et al., 1996;Grillenzoni et al., 1998) orDistal-less (Dll)(Abu-Shaar and Mann, 1998;Campbell and Tomlinson, 1998;Gonzalez-Crespo et al., 1998;Wu and Cohen, 1999), may give territorial identity by themselves or in combination with others, but they specify territories that are rarely clonally delimited. `Differentiation genes', such as achaete-scute (ac-sc), are involved in the terminal development of cells or tissues(Jimenez and Campos-Ortega,1990). Finally, a new category has been introduced, the `master'genes (Halder et al., 1995;Kim et al., 1996), which includes genes such as eyeless (ey) and vestigial(vg). These genes are considered to be able to individually initiate and drive specific developmental pathways, thus changing the fate of the tissues in which they are ectopically expressed. The concept of master gene has been applied to `selector of tissue' genes later by other authors(Bray, 1999;Affolter and Mann, 2001;Halder and Carroll, 2001). In this context, the expression of ey or vg would be sufficient to give identity to eyes and wings, respectively(Halder et al., 1995;Kim et al., 1996). We show here that the ectopic expression of vg is the subject of temporal and genetic constraints in the promotion and driving of the wing developmental program.
vg encodes a nuclear protein of 453 amino acids with poor homologies to other known proteins. It is expressed at low levels in the primordial wing and haltere imaginal discs(Williams et al., 1991). Later in the development of both discs, the dorsoventral border of compartment is defined by differential expression of apterous (ap)(Diaz-Benjumea and Cohen,1993), and subsequent restricted activation of Notch(N) (Irvine and Vogt,1997) and downstream genes such as wingless (wg)(Kim et al., 1995). The activity of wg and N in the wing margin leads to the expression of the vg through the activation of the vgboundary enhancer (vg BE-lacZ)(Kim et al., 1996;Williams et al., 1994). Subsequent to vg BE-lacZ activation, the expression of vg in more proximal parts of the wing blade is regulated by thedecapentaplegic (dpp) pathway and by vg itself,acting on the vg quadrant enhancer (vg QE-lacZ)(Kim et al., 1996). Reflecting the activity of the vg enhancers, Vg is expressed in a gradient with maximal concentrations in the wing margin and minimal in more proximal territories of the wing (Williams et al.,1991). Vg interacts with the product of the genescalloped (sd), a protein with DNA recognition motifs(Campbell et al., 1992),forming a transcriptional activation complex(Halder and Carroll, 2001;Halder et al., 1998). The Vg-Sd complex is known to regulate the expression of downstream genes involved in wing development (Halder and Carroll,2001; Halder et al.,1998; Kim et al.,1996; Klein and Martínez-Arias, 1998). The absence of Vg, Sd or both causes lack of cell proliferation in the wing blade region where vg is expressed (Williams et al.,1991; Williams et al.,1993). By contrast, the ectopic expression of vg may cause the appearance of territories with cuticular and genetic expression patterns that are characteristic of distal wing blade(Halder et al., 1998;Kim et al., 1996). It is important to note that the ectopic expression of Sd alone does not induce tissue transformations. Thus, Sd is necessary to bind the complex Vg-Sd to DNA, but does not confer tissue specificity by itself(Halder and Carroll,2001).
In order to analyse the capacity and the genetic requirements of the ectopic expression of vg to initiate and drive the transformation of tissue towards wing blade identity, we studied the autonomous and non autonomous cuticular and gene expression patterns that appear after the ectopic expression of vg during development. We drove vgectopic expression using different territorial Gal4 lines (G4/UAS system)(Brand and Perrimon, 1993) or by Flip-out (FLP/FRT system) recombination in clones(de Celis and Bray, 1997;Ito et al., 1997). The results show that the morphogenetic effects of vg ectopic expression depend on developmental timing and the genetic specification of a disc territory where the overexpression takes place.
MATERIALS AND METHODS
Fly stocks
The experiments were carried out using the following G4 lines: dpp-G4 A.3(Staehling-Hampton et al.,1994); vg-G4 (Simmonds et al.,1995); Dll-G4 (Calleja et al.,1996); c253-G4 (provided by Juan Modolell); pnr-G4(Calleja et al., 1996); and patch-G4 (ptch-G4) (Speicher et al.,1994).
We used the following UAS lines: UAS-vgK(Kim et al., 1996);UAS-vgZ (Paumard-Rigal et al., 1998); UAS-wg(Lawrence et al., 1995);UAS-wg dominant negative (UAS-wgDN)(Klein and Martínez-Arias,1999); UAS-Dll(Gorfinkiel et al., 1997);UAS-homothorax-GFP (UAS-hth-GFP) (provided by Fernando Casares);UAS-nubbin (UAS-nub)(Neumann and Cohen, 1998);UAS-sd (Campbell et al.,1992); UAS-Delta (UAS-Dl)(Huppert et al., 1997);UAS-Serrate (UAS-Ser) and UAS-Notch constitutively active (UAS-Nintra12.1)(de Celis and Bray, 1997);UAS-thickvein constitutively active (UAS-tkvQ25)(Lecuit et al., 1996) or dominant negative tkv (UAS-tkvDN)(Haerry et al., 1998);UAS-Ras constitutively active (UAS-RasV12)(Karim and Rubin, 1998);UAS-Ras dominant negative (UAS-Raf3.1DN)(Kim et al., 1995); UAS-P35(provided by C. Lehner); UAS-ap(Fernandez-Funez et al.,1998); UAS-fringe (UAS-fng)(Kim et al., 1995); and UAS-GFP.
We used the following lacZ lines: vg QE-lacZ(Kim et al., 1996); vg BE-lacZ (Williams et al.,1994); wg-lacZ(Kassis et al., 1992);aprk560 (ap-lacZ)(Diaz-Benjumea and Cohen,1993); and sdETX4 (sd-lacZ)(Anand et al., 1990).
The lines used for the induction of overexpression mosaics by Flip-out werey f36a FLP122; abx/Ubx FRT f+ FRT G4 UAS lacZ (de Celis and Bray,1997); and y FLP122; Act FRT y+ FRT G4 UAS-GFP: MKRS/ SM6A-TM6B (Ito et al.,1997).
Ectopic expression using territorial G4 lines and generation of overexpression clones
The ectopic expression with different lines G4 was induced at 17, 25 and 29°C. Genetic of Flip-out to induce clones of overexpression of vg,wg or both wg-vg: larvae were transferred from 25°C to 37°C for 7 minutes at different ages 36±12, 48±12 or 60±12 hours after laying egg AEL for vg clones, and 36±12 and 60±12 hours AEL for the wg-vg or wgclones.
Inmunohistochemistry
Imaginal disc were dissected in 1×PBS and fixed in 4% PFA at 4°C for 40 minutes, followed by 3×20 minute washes in 0.3% PBTriton and 3×20 minute washes in PBT-BSA. Incubation with primary antibody was carried out overnight at 4°C. After repeating washes with PBT and PBT-BSA the imaginal discs were incubated 2 hours at room temperature with the secondary antibody. Imaginal discs were mounted in Vectashield.
We used the following primary antibodies: rabbit anti-Dll and anti-Vg(provided by Sean Carroll); mouse anti-Wg (Hybridoma Bank); guinea pig anti-Hth (gift of Fernando Casares); mouse anti-Cut (Hybridoma Bank); mouse anti-Bs (provided by M. Affolter); mouse anti-Nub (provided by S. Cohen);mouse anti-En (Hybridoma Bank); rat anti-Ser (Hybridoma Bank); rat anti-CD2(Hybridoma Bank); and mouse or rabbit anti β-gal (Amersham). We used rabbit and mouse Alexa 488, 546, Cy5 and guinea pig-Cy5 as secondary antibodies.
Microscopy and image treatment
For the processing of images in clear field and confocal microscopy we used the programs Metaview (Meta Imaging Corporation Plus) and Photoshop 6.0 (Adobe Corporation).
RESULTS
The ectopic expression of vg in G4 territories shows constraints on the promotion of the wing developmental program
We drove the ectopic expression of vg with different G4 lines, in order to observe the response of territories of cells ectopically expressingvg from early development onwards. Here, we monitor patterns of gene expression and cuticular differentiation.
It has been shown that the overexpression of vg driven by different G4 lines, such as ptch-G4(Paumard-Rigal et al., 1998)(Simmonds et al., 1998),dpp-G4 (Kim et al., 1996;Klein and Martínez-Arias,1998; Klein and Martínez-Arias, 1999) and Dll-G4(Fig. 1A,C)(Halder et al., 1998;Weatherbee et al., 1998),ectopically induce histotypes and gene expression patterns characteristic of the wing blade. We explored the maximal transformation phenotypes using the driver Dll-G4. The ectopic expression of vg in the distal territories of the appendages (proboscis, first and second pairs of legs and genitalia),where Dll is expressed, leads to the transformation into tissues with characteristic cuticular patterns and differentiation of the wing blade(Fig. 1A,C). In the first and second pairs of legs, these transformations include typical trichomes,anteroposterior and dorsoventral wing margin chaetae(Fig. 1A2-A4, C1-C3), veins and campaniform sensillae (25/77) (Fig. 1A1,A3). In the third pair of legs, the transformations are to haltere histotypes (Fig. 1A5)(Halder et al., 1998;Weatherbee et al., 1998).
Cuticular transformations caused by the ectopic expression of vg(A,C) and co-overexpression of wgDN-vg (B,D) under the control of Dll-G4. Effects of vg ectopic overexpression on gene expression in the second pair of the legs driven by Dll-G4 (E-F). Lines indicate optical sections along the z-axis. Scale bars: 0.1 mm.(A) Ventral view of a transformed thorax with distal segments of first and second legs differentiating as wing blade (blue asterisk) and third legs as haltere (red asterisk). The broken red line separates anterior (a) from posterior (p) territories in transformed legs. (A1,A2) Transformed wing blade with dorsal (D) and ventral (V) regions. Arrowheads indicate veins. (A3) The arrowhead indicates a campaniform sensillae. (A4) Higher magnification of the wing margin in anteroposterior transition. (A5) Observe the histotypic differences between wing blade (blue asterisk) and haltere territories (red asterisk). (B) Ectopic co-expression of vg and a dominant-negative form of wg reduces the transformation phenotypes obtained with the ectopic expression of vg alone. (B1) Distal (D) (red arrowhead) and proximal (P) tarsal parts in the third pair of legs (red asterisk) at higher magnification. Distal segments (arrowhead indicates bracts) are shown in B2 and proximal segments (without bracts) in B3. (C) Ectopic expression ofvg leads to transformation of antenna with anterior (a) and posterior(p) wing margin elements. The insets show the differentiation of wing margin elements with anterior (a; C1,C2) and posterior (p; C1,C2) specifications. (D)Transformation in the antenna is reduced by the co-expression of vgand a dominant-negative form of wg. (E) Ectopic expression of vg QE-lacZ (green) in the transformed legs. Notice repression of vg QE-lacZ report (green) in territories with high levels of wgexpression (red). (F) Dorsal transformed territories in adult legs are correlated with ap (green) expression. Notice that the annular expression of ap-lacZ in the leg is modified: it is expanded in ventral territories but reduced (white arrowhead) in dorsal ones. The expression of wg (red) is activated at high levels in the border(wing margin) of ap expression. (G) Anterior and posterior specification in transformations denoted by the wild-type expression ofen (green). The expression of wg-lacZ (red) corresponding to the wing margin appears in anterior as well as in posterior territories.
Cuticular transformations caused by the ectopic expression of vg(A,C) and co-overexpression of wgDN-vg (B,D) under the control of Dll-G4. Effects of vg ectopic overexpression on gene expression in the second pair of the legs driven by Dll-G4 (E-F). Lines indicate optical sections along the z-axis. Scale bars: 0.1 mm.(A) Ventral view of a transformed thorax with distal segments of first and second legs differentiating as wing blade (blue asterisk) and third legs as haltere (red asterisk). The broken red line separates anterior (a) from posterior (p) territories in transformed legs. (A1,A2) Transformed wing blade with dorsal (D) and ventral (V) regions. Arrowheads indicate veins. (A3) The arrowhead indicates a campaniform sensillae. (A4) Higher magnification of the wing margin in anteroposterior transition. (A5) Observe the histotypic differences between wing blade (blue asterisk) and haltere territories (red asterisk). (B) Ectopic co-expression of vg and a dominant-negative form of wg reduces the transformation phenotypes obtained with the ectopic expression of vg alone. (B1) Distal (D) (red arrowhead) and proximal (P) tarsal parts in the third pair of legs (red asterisk) at higher magnification. Distal segments (arrowhead indicates bracts) are shown in B2 and proximal segments (without bracts) in B3. (C) Ectopic expression ofvg leads to transformation of antenna with anterior (a) and posterior(p) wing margin elements. The insets show the differentiation of wing margin elements with anterior (a; C1,C2) and posterior (p; C1,C2) specifications. (D)Transformation in the antenna is reduced by the co-expression of vgand a dominant-negative form of wg. (E) Ectopic expression of vg QE-lacZ (green) in the transformed legs. Notice repression of vg QE-lacZ report (green) in territories with high levels of wgexpression (red). (F) Dorsal transformed territories in adult legs are correlated with ap (green) expression. Notice that the annular expression of ap-lacZ in the leg is modified: it is expanded in ventral territories but reduced (white arrowhead) in dorsal ones. The expression of wg (red) is activated at high levels in the border(wing margin) of ap expression. (G) Anterior and posterior specification in transformations denoted by the wild-type expression ofen (green). The expression of wg-lacZ (red) corresponding to the wing margin appears in anterior as well as in posterior territories.
The adult transformations are correlated in imaginal discs with autonomous expression of genes characteristic of wing blade. These gene expressions only appear within GFP-expressing cells, as mobilised by Dll-G4. The ectopic wing margin differentiated in the outgrowths correlates with the expression ofcut (ct) (not shown) and wg in the presumptive wing margin in the disc (Fig. 1E-G). Thus, as in the wild-type wing margin, wg represses the expression ofvg QE-lacZ (Fig. 1E). The transformations have large ventral wing territories and small dorsal territories encircled by the new wing margin(Fig. 1A1-4,E,F). The dorsoventral and anteroposterior transformations are correlated with the differential expression of en and ap(Fig. 1A,E,F,G), in the mature imaginal discs. The expression of en is maintained in the wild-type topology but the expression of ap is modified(Fig. 1F,G). Thus, the ring ofap in the leg is repressed in dorsal leg territories and expanded in ventral ones (Fig. 1F). In the first and second leg-transformed territories, the genes characteristic of wing blade territories, such as spalt (sal), blistered(bs) (Halder et al.,1998; Weatherbee et al.,1998) and vg QE-lacZ(Fig. 1G) are expressed. In the third leg, where transformations are to haltere, some markers of wing transformation (such as sal or vg QE-lacZ) are not expressed(not shown), probably because of Ubx activity(Halder et al., 1998;Shashidhara et al., 1999;Weatherbee et al., 1998).Ubx expression is never modified by overexpression of vg orwg-vg driven by different G4 lines or in clones (not shown, see below). All described transformations and expression of wing blade genes are autonomously restricted to the Dll expression domain, visualised by GFP.
In contrast to these G4 lines, which induce transformation phenotypes, the ectopic expression of vg driven by G4 lines is not associated with histotypic transformations, and only causes tissue-specific malformations. Thus, with pnr-G4, we observe defects in thorax closure, and with c253-G4 duplications of chaetae in the notum (not shown). We also failed to obtain transformations when the ectopic expression of vg alone in the eye was driven by vg-G4.
These results indicate that expression of vg is necessary but not sufficient for the initiation of the wing blade developmental pathway.
Cooperative effects of vg and wg in the ectopic transformations
We analyzed why some G4 lines may lead transformations while others fail to do it driving UAS-vg. We have found that the overexpression ofvg only induces transformations when the G4 line shares expression domains with high levels of wg in early stages of larval development. Thus, we have found that the overexpression of vg only induces transformations when the G4 line shares expression domains with high levels ofwg in early stages of larval development. Furthermore, thatwg is necessary for the augment the transformation is confirmed by experiments in the leg in which a dominant-negative form of wg(wgDN) is ectopically co-expressed with vg, using the driver Dll-G4. In these legs, the transformation is strongly reduced(Fig. 1B,D). Without transformation, these legs maintain distal tarsal structures (territories of the legs with chaetae and without bracts), suggesting that the lack of transformation is not simply a consequence of the low levels of Wg activity(Fig. 1B). To test the hypothesis of collaboration of wg and vg in wing blade transformation, we co-expressed them in the expression domain ofvg-G4 in the eye (Fig. 2B). Whereas the ectopic expression of vg or wgalone does not show cuticular transformations (not shown), the co-expression of both causes wing outgrowths with histotypical characteristics of wing blade(Fig. 2A). The transformation is autonomously associated with the expression of wing blade genetic markers as nub, vg (Fig. 2C,D)and Dll (not shown). Surprisingly, in the transformed territories the expression of wg is lower than we would expect of an overexpression using the G4/UAS system (Fig. 2D) (see Discussion). These results demonstrate that thatwg and vg collaborate to initiate wing development in imaginal discs other than the wing.
Ectopic wg-vg co-expression driven by vg-G4 causes the transformation of eye territories into wing blade with cuticular structures(A) and gene expression (C,D) characteristic of wing blade. The expression ofvg-G4 is detected in the eye imaginal disc by the simultaneous mobilisation of UAS-GFP (GFP) (B-D). Scale bars: 0,1 mm. (A) Observe chaetae(black arrowhead) and wing blade trichomes which appear when we ectopically express vg in the eye. (B) In the wild-type eye imaginal disc, vg-G4(GFP) and wg (red) are expressed in different territories. High levels of wg expression are detected in the poles of the eye, whereas vg-G4 is expressed in the equatorial line in the eye. (C) Notice the autonomous ectopic expression of nub (red; white arrowhead) in transformed eye cells caused by wg-vg co-expression (GFP). (D)Ectopic expression of vg (red; white arrowhead) in eye transformations caused by the wg-vg co-expression. Notice that the level of wg expression is very low in the transformed territories(white arrowhead).
Ectopic wg-vg co-expression driven by vg-G4 causes the transformation of eye territories into wing blade with cuticular structures(A) and gene expression (C,D) characteristic of wing blade. The expression ofvg-G4 is detected in the eye imaginal disc by the simultaneous mobilisation of UAS-GFP (GFP) (B-D). Scale bars: 0,1 mm. (A) Observe chaetae(black arrowhead) and wing blade trichomes which appear when we ectopically express vg in the eye. (B) In the wild-type eye imaginal disc, vg-G4(GFP) and wg (red) are expressed in different territories. High levels of wg expression are detected in the poles of the eye, whereas vg-G4 is expressed in the equatorial line in the eye. (C) Notice the autonomous ectopic expression of nub (red; white arrowhead) in transformed eye cells caused by wg-vg co-expression (GFP). (D)Ectopic expression of vg (red; white arrowhead) in eye transformations caused by the wg-vg co-expression. Notice that the level of wg expression is very low in the transformed territories(white arrowhead).
Other experimental or genetic conditions that affect ectopicvg transformations
We have explored other experimental conditions that might allow vgto induce wing transformations.
We searched for other genes, in addition to wg, that might cooperate with vg, using vg-G4 expression in the eye. First, we studied: (1) the effects of one or several UAS-vg doses and G4 induction temperatures (not shown); (2) variations in the stoichiometry of Vg with Sd (Paumard-Rigal et al.,1998; Simmonds et al.,1998) co-expressing several doses of the corresponding UAS; (3)co-expression of vg with genes involved in the specification of the proximodistal axes of the wing, such as nub, Dll and hth(Abu-Shaar and Mann, 1998;Azpiazu and Morata, 2000;Cifuentes and Garcia-Bellido,1997); (4) co-expression with genes involved in dorsoventral wing margin specification, such as ap(Diaz-Benjumea and Cohen,1993) and fng (Irvine and Wieschaus, 1994); (5) co-expression of vg with UAS p35 to rescue possible cell death (Hay et al., 1994); and (6) co-expression of vg with several constructs of genes that are involved in signalling pathways during development, such as UAS-Dl, UAS-Ser,UAS-Nintra12.1, UAS-tkvQ25,UAS-RasV12, UAS-tkvDN,UAS-Raf3.1DN.
Secondly, we tested whether the size of the territory of ectopic expression was critical in the process of transformation, overexpressing vg with other G4 lines such as pnr-G4 and c253-G4.
We failed, in all instances, to enhance the histotypic transformation caused by overexpressing vg alone, and therefore conclude that neither the extension of the territory ectopically expressing vg nor the amount of overexpression is significantly relevant to the extent of transformation to wing. Thus, only wg in collaboration withvg seems to be specifically relevant in the promotion of wing development.
Phenotypes induced by the ectopic expression of vg in clones: temporal and genetic limitations to the initiation of a wing developmental program
Gene expression driven by a given G4 line occurs simultaneously in all the cells of the territory at a given developmental stage, allowing collaborative effects between cells. By contrast, the ectopic expression in clones provides temporal and positional limits to the transformation. In clonal mosaics,individual mutant cells are confronted with wild-type cells, allowing the study of autonomous and non-autonomous effects in cell proliferation,cuticular patterning and gene expression. We have monitored the cuticular and genetic effects of vg ectopic expression in Flip-out clones [labelled with forked (f) or GFP], in the wing, haltere, leg and eye-antenna imaginal discs.
In the wing blade and wing hinge, vg clones induce tubular,perpendicular outgrowths to the wing surface(Fig. 3A). The outgrowths include vg-expressing cells and surrounding wild-type cells(Fig. 3A). vg clones are frequently located at the tip of the outgrowth but they can also grow along the lateral zones (Fig. 3A). In those clones that appear in central parts of the wing, all the cells of the outgrowth and the clone always show a differentiation corresponding to wing blade trichomes (Fig. 3A). Clones near the wing margin may differentiate marginal chaetae (not shown). The size of the clones and the non-autonomous part of the outgrowth depend on their distance to the wing margin(Liu et al., 2000).
Adult phenotypes (A) and imaginal disc gene expression (B-F) caused by the overexpression of vg in clones initiated in the wing blade and wing hinge. Adult clones are labelled with f and delimited by a broken red line. In the imaginal discs, vg clones are associated with green fluorescent protein (GFP). Scale bars: 0.1 mm. (A) Tubular outgrowth including a clone of vg (red arrowhead). The age of clone initiation is 36±12 hours AEL. In the inset a detail of the same clone is shown. (B)In the wing hinge, vg overexpression in clones (green) displaces the rings of wg expression (red) by several cellular diameters. The age of clone initiation is 60±12 hours AEL. (C) In the wing hinge, thevg overexpression in clones (green) non-autonomously activates the expression of Dll (red) and autonomously represses Hth expression (blue). The age of clone initiation is 60±12 hours AEL. (D) In the wing hinge,vg overexpression in clones (green) (white arrowhead) autonomously activates the vgQE-lacZ reporter (red). The age of clone initiation is 60±12 hours AEL. (E) vg overexpression in clones(green; white arrowhead) never activates the vg BE-lacZ reporter(red) reporter in the wing imaginal disc (or in other tissues). The age of clone initiation is 60±12 hours AEL. (F) In all studied tissues,vg overexpression in clones (green; white arrowhead) autonomously activates the sd-lacZ reporter (red). The age of clone initiation is 60±12 hours AEL.
Adult phenotypes (A) and imaginal disc gene expression (B-F) caused by the overexpression of vg in clones initiated in the wing blade and wing hinge. Adult clones are labelled with f and delimited by a broken red line. In the imaginal discs, vg clones are associated with green fluorescent protein (GFP). Scale bars: 0.1 mm. (A) Tubular outgrowth including a clone of vg (red arrowhead). The age of clone initiation is 36±12 hours AEL. In the inset a detail of the same clone is shown. (B)In the wing hinge, vg overexpression in clones (green) displaces the rings of wg expression (red) by several cellular diameters. The age of clone initiation is 60±12 hours AEL. (C) In the wing hinge, thevg overexpression in clones (green) non-autonomously activates the expression of Dll (red) and autonomously represses Hth expression (blue). The age of clone initiation is 60±12 hours AEL. (D) In the wing hinge,vg overexpression in clones (green) (white arrowhead) autonomously activates the vgQE-lacZ reporter (red). The age of clone initiation is 60±12 hours AEL. (E) vg overexpression in clones(green; white arrowhead) never activates the vg BE-lacZ reporter(red) reporter in the wing imaginal disc (or in other tissues). The age of clone initiation is 60±12 hours AEL. (F) In all studied tissues,vg overexpression in clones (green; white arrowhead) autonomously activates the sd-lacZ reporter (red). The age of clone initiation is 60±12 hours AEL.
vg clones in the presumptive wing blade do not modify the wild-type expression of genes expressed, such as Dll, bs andnub (not shown). However in the wing hinge, clones of vgoverexpression autonomously repress proximal genes such as hth(Fig. 3C)(Azpiazu and Morata, 2000;Casares and Mann, 2000;Liu et al., 2000) and activate antagonist distal genes such as Dll(Azpiazu and Morata, 2000;Liu et al., 2000), vgQE-lacZ (Fig. 3D),nub (Liu et al.,2000) and bs (Fig. 3C; Table 1)(Liu et al., 2000). In some cases, the ectopic expression of distal genes may appear non-autonomously outside the clone, up to a distance of several cell diameters(Fig. 3B,C;Table 1)(Liu et al., 2000). Vg, or thevg enhancer lacZ reporters, are never detected non-autonomously in vg clones located outside of the wing blade. Whereas early vg clones can show co-expression with wg (not shown), later in development, wg expression is displaced outside of the clone several cell diameters (Fig. 3A) (Liu et al.,2000). The absence of wing margin cuticular elements in adultvg clones and ct or vg BE-lacZ(Fig. 3E;Table 1) expression in the discs suggests that the clones are specified as wing blade, not including wing margin territories. vg clones in the wing imaginal disc show a correlated and autonomous expression of sd-lacZ within the clone(Fig. 3F). In tissues other than the wing, the ectopic expression of vg may drive the expression of its transcriptional partner (not shown)(Halder and Carroll, 2001;Halder et al., 1998).
In territories other than the wing, vg clones show transformations toward wing histotype only when they are initiated in specific positions and stages of development. Thus, ectopic expression of vg in clones is only associated with wing outgrowth phenotypes when it is initiated in territories that normally express high levels of wg(Fig. 4C,Fig. 5C,H). For example,vg clones in the notum, only show wing histotype when they are initiated very early in development (36±12 hours AEL), whereas in the eye and leg imaginal discs, transformations may appear later (48±12 hours AEL). Clones initiated later in development or in territories with low levels of wg expression cause cuticular abnormalities(Fig. 5B) but not transformations towards wing histotype.
Adult `mixed' tissue phenotypes (A,B) and imaginal disc expression (C,D)caused by the ectopic clonal expression of vg in the notum. Adult clones are labelled with f and delimited by a broken red line. In the imaginal discs, vg clones are associated with GFP. The age of clone initiation is 36±12 hours AEL. Scale bars: 0.1 mm. (A) Clone ofvg ectopic expression in the notum initiated at 36±12 hours AEL; high magnification is shown in B. The clone territory contains trichomes of wing blade (red arrowhead) and notum chaetae in a `salt and pepper'distribution. Notice the differences between trichomes with wing blade characteristics (red arrowhead) contained in the clone with notum trichomes(black arrowhead). (C) Clones of vg ectopic expression (green) in the notum that simultaneously contain cells expressing [dorsal territories (d)]and non-expressing [ventral territories (v)] the ap-lacZ reporter. Notice that endogenous wg expression in the notum appears displaced(yellow arrowhead), whereas wg expression is enhanced within the clone (white arrowhead) at the confrontation between cells expressing and non-expressing ap-lacZ, as occurs in the wild-type wing margin. (D)Adult mixed tissues are correlated, in imaginal discs, with the autonomous expression of wing blade genes such as bs (red). Notice that the expression of bs is non-autonomously induced outside the clone(green; white arrowhead), as well as partially induced within the clone.
Adult `mixed' tissue phenotypes (A,B) and imaginal disc expression (C,D)caused by the ectopic clonal expression of vg in the notum. Adult clones are labelled with f and delimited by a broken red line. In the imaginal discs, vg clones are associated with GFP. The age of clone initiation is 36±12 hours AEL. Scale bars: 0.1 mm. (A) Clone ofvg ectopic expression in the notum initiated at 36±12 hours AEL; high magnification is shown in B. The clone territory contains trichomes of wing blade (red arrowhead) and notum chaetae in a `salt and pepper'distribution. Notice the differences between trichomes with wing blade characteristics (red arrowhead) contained in the clone with notum trichomes(black arrowhead). (C) Clones of vg ectopic expression (green) in the notum that simultaneously contain cells expressing [dorsal territories (d)]and non-expressing [ventral territories (v)] the ap-lacZ reporter. Notice that endogenous wg expression in the notum appears displaced(yellow arrowhead), whereas wg expression is enhanced within the clone (white arrowhead) at the confrontation between cells expressing and non-expressing ap-lacZ, as occurs in the wild-type wing margin. (D)Adult mixed tissues are correlated, in imaginal discs, with the autonomous expression of wing blade genes such as bs (red). Notice that the expression of bs is non-autonomously induced outside the clone(green; white arrowhead), as well as partially induced within the clone.
Adult phenotypes (A,B,G) and gene expression patterns caused by the ectopic expression of vg in clones located in the leg (C-F) and eye (H,I)imaginal discs. Adult clones are labelled with f and delimited by a broken red line. In the imaginal discs, vg clones are associated with GFP. Scale bars: 0.1 mm. (A) Clones of vg ectopic expression include`mixed' cuticular differentiation patterns in the leg, with leg chaetae mixed in a `salt and pepper' distribution with wing blade trichomes. In the inset,compare trichomes of the leg (black arrowhead) with wing blade trichomes (red arrowhead). The age of clone initiation is 36±12 hours AEL. (B)Proximalisation phenotypes of the leg without cuticular transformation induced by vg ectopic expression in clones. In distal tarsal segments of the leg, typical chaetes are associated with bracts (black arrowhead in the inset), whereas in vg clones chaeates are not associated with bracts(red arrowhead in the inset). Notice the non-autonomous size reduction in distal tarsal segments of the leg. The age of clone initiation is 36±12 hours AEL. (C) Clones of vg ectopic expression (green) autonomously reduce endogenous wg expression (red; white arrowhead). The age of clone initiation is 60±12 hours AEL. (D) Clones of vg ectopic expression (green) may activate bs (red) expression (white arrowhead)in a subset of cells within the clone. Notice that only those clones situated in ventral leg territories induce bs expression. The age of clone initiation is 60±12 hours AEL. (E,F) Clones of vg ectopic expression (green) in ventral territories of the leg show autonomous expression of the ap-lacZ reporter (red) (E), whereas it is repressed in dorsal territories (F). Notice that ap expression is not detected in all cells of the clone, similar to D. The age of clone initiation is 60±12 hours AEL. (G) Outgrowth with wing blade trichomes (indicated by red arrowhead) induced by vg clone. In the inset, trichomes with wing blade characteristics (red arrowhead) are compared with surrounded ommatidial differentiation (black arrowhead). The age of clone initiation is 36±12 hours AEL. (H) Clones of ectopic vg expression (green) autonomously reduce the expression of wg (red; white arrowhead). The age of clone initiation is 60±12 hours AEL. (I) In the eye, clones of vgectopic expression (green) autonomously induce the expression of nub(white arrowhead). Only those clones induced in territories with high levels of Wg express nub. The age of clone initiation is 60±12 hours AEL.
Adult phenotypes (A,B,G) and gene expression patterns caused by the ectopic expression of vg in clones located in the leg (C-F) and eye (H,I)imaginal discs. Adult clones are labelled with f and delimited by a broken red line. In the imaginal discs, vg clones are associated with GFP. Scale bars: 0.1 mm. (A) Clones of vg ectopic expression include`mixed' cuticular differentiation patterns in the leg, with leg chaetae mixed in a `salt and pepper' distribution with wing blade trichomes. In the inset,compare trichomes of the leg (black arrowhead) with wing blade trichomes (red arrowhead). The age of clone initiation is 36±12 hours AEL. (B)Proximalisation phenotypes of the leg without cuticular transformation induced by vg ectopic expression in clones. In distal tarsal segments of the leg, typical chaetes are associated with bracts (black arrowhead in the inset), whereas in vg clones chaeates are not associated with bracts(red arrowhead in the inset). Notice the non-autonomous size reduction in distal tarsal segments of the leg. The age of clone initiation is 36±12 hours AEL. (C) Clones of vg ectopic expression (green) autonomously reduce endogenous wg expression (red; white arrowhead). The age of clone initiation is 60±12 hours AEL. (D) Clones of vg ectopic expression (green) may activate bs (red) expression (white arrowhead)in a subset of cells within the clone. Notice that only those clones situated in ventral leg territories induce bs expression. The age of clone initiation is 60±12 hours AEL. (E,F) Clones of vg ectopic expression (green) in ventral territories of the leg show autonomous expression of the ap-lacZ reporter (red) (E), whereas it is repressed in dorsal territories (F). Notice that ap expression is not detected in all cells of the clone, similar to D. The age of clone initiation is 60±12 hours AEL. (G) Outgrowth with wing blade trichomes (indicated by red arrowhead) induced by vg clone. In the inset, trichomes with wing blade characteristics (red arrowhead) are compared with surrounded ommatidial differentiation (black arrowhead). The age of clone initiation is 36±12 hours AEL. (H) Clones of ectopic vg expression (green) autonomously reduce the expression of wg (red; white arrowhead). The age of clone initiation is 60±12 hours AEL. (I) In the eye, clones of vgectopic expression (green) autonomously induce the expression of nub(white arrowhead). Only those clones induced in territories with high levels of Wg express nub. The age of clone initiation is 60±12 hours AEL.
In tissues other than the wing, clones of vg ectopic expression associated with transformation differentiate only wing blade trichomes(Fig. 4A,B,Fig. 5A,G), in contrast to the overexpression of vg with G4 lines in the same territories. The wing blade trichomes in some tissues such as notum or legs may appear `mixed', with tissue-specific chaetae in a `salt and pepper' distribution(Fig. 4B,Fig. 5A). The adult `mixed'cuticular patterns are correlated for each examined tissue with the specific expression of some wing blade genes (Table 1). Thus, we never detected the expression of nub in the notum, whereas it is induced in the leg or eye imaginal discs(Fig. 5I); and bs is never detected in the eye, whereas it is induced in the leg(Fig. 5D). Paradoxically, in contrast to `salt and pepper' distribution of the adult cuticular structures detected in the transformations, the ectopic expression of specific wing blade genes only occurs in subsets of cells within the clones(Fig. 4D,Fig. 5D,E). Moreover, the cells expressing wing blade genes are compacted and located anywhere within the clones of vg (Fig. 4D,Fig. 5D,E). The discrepancy between cuticular pattern and gene expression suggests that vg can not displace all endogenous identity signals or, alternatively, that there are non-autonomous influences of surrounding cells on of the clone expressing wing blade genes.
In tissues other than the wing, the histotype transformations and expression of wing blade gene are cell autonomous. But in the notum,vg clones that straddle the DV boundary and are initiated in territories with high levels of wg(Fig. 4C) and occasionally cause non-autonomous expression of wing gene markers such as bs(Fig. 4D). This phenomenon of non-autonomous induction of tissue to express wing blade genes, we called`inductive assimilation' (see Discussion). In these clones, the wild-type expression of wg is displaced, but where ap expressing and non-expressing cells are confronted in the clone, wg expression is autonomously enhanced, as in the wing margin(Fig. 4C). This reflects the possibility that vg clones may recruit the expression of apbefore the wild-type specification of DV wing margin, generating ectopic DV wing margins.
As in the G4 experiments, the expression of en is not modified invg clones, and the disc therefore retains the embryonic AP compartment specification. However, ap expression is altered in clones expressing ectopically vg in a tissue-specific way. Thus,ap expression is not modified in the wing imaginal disc or haltere,whereas in leg discs, ap expression is induced in ventral territories and repressed in dorsal ones (Fig. 5E,F).
Similar to en expression, Ubx is not modified in clones expressing ectopically vg or vg and wgsimultaneously (wg-vg), and therefore, the segmental identity dependent on Ubx expression is maintained. In the haltere,vg or wg-vg clones have otherwise the same autonomous and non-autonomous effects than in the wing (not shown).
The activity of wg pathway together with vg is insufficient to promote a wing developmental program in clones
In order to test the interaction of the wg pathway withvg in clones, we have studied the cuticular and gene expression patterns shown by cells expressing either vg or wg alone,and co-expressing wg and vg (wg-vg) simultaneously. The phenotypes of wg or wg-vg clones have been monitored in the wing, haltere, leg and eye-antenna imaginal discs.
In the wing, the overexpression of wg in clones does not cause outgrowths and all cells of the clone differentiate into wing margin sensory elements (not shown). These results confirm the proposition of Klein and co-workers (Klein et al.,1997; Klein and Martínez-Arias, 1998) that the overexpression ofwg induces the cells to acquire characteristics of wing margin. However, these cells fail to express vg BE-lacZ or other wing margin genes such as ct (not shown). In these clones, some wing blade markers such as Dll (not shown)(Zecca et al., 1996),vg (Fig. 6B)(Zecca et al., 1996) ornub (Fig. 6C) are autonomous and non-autonomously expressed, whereas other wing blade genes such as bs are autonomous and non-autonomously repressed (not shown).
Cuticular phenotypes (A) and imaginal disc expression pattern (B-D) caused by clonal overexpression of wg in the wing imaginal disc. Adult clones are labelled with f and delimited by a broken red line. In the imaginal discs, wg clones are associated with GFP. Scale bars: 0.1 mm. (A) Clones of wg overexpression in the notum show the typical differentiation of this territory. In the inset notice that trichomes are equal inside (red arrowhead) and outside the clone (black arrowhead). The age of clone initiation is 36±12 hours AEL. (B) Clones of wgoverexpression (green) in the wing blade autonomously and non-autonomously induce the expression of vg (red), whereas those clones outside the wing blade do not express vg (white arrowhead). The age of clone initiation is 60±12 hours AEL. (C) Clones of wg overexpression(green) in the wing imaginal disc autonomous and non-autonomously induce the expression of nub (red; white arrowhead). Notice that non-autonomous expression of nub is detected only in the closest cells surrounding the clone. The age of clone initiation is 60±12 hours AEL. (D) Clones of wg overexpression (green) show homogenous and high levels of Wg(red). The age of clone initiation is 60±12 hours AEL.
Cuticular phenotypes (A) and imaginal disc expression pattern (B-D) caused by clonal overexpression of wg in the wing imaginal disc. Adult clones are labelled with f and delimited by a broken red line. In the imaginal discs, wg clones are associated with GFP. Scale bars: 0.1 mm. (A) Clones of wg overexpression in the notum show the typical differentiation of this territory. In the inset notice that trichomes are equal inside (red arrowhead) and outside the clone (black arrowhead). The age of clone initiation is 36±12 hours AEL. (B) Clones of wgoverexpression (green) in the wing blade autonomously and non-autonomously induce the expression of vg (red), whereas those clones outside the wing blade do not express vg (white arrowhead). The age of clone initiation is 60±12 hours AEL. (C) Clones of wg overexpression(green) in the wing imaginal disc autonomous and non-autonomously induce the expression of nub (red; white arrowhead). Notice that non-autonomous expression of nub is detected only in the closest cells surrounding the clone. The age of clone initiation is 60±12 hours AEL. (D) Clones of wg overexpression (green) show homogenous and high levels of Wg(red). The age of clone initiation is 60±12 hours AEL.
In the notum, overexpression of wg in clones is not associated with histotype transformations or `mixed' tissues(Fig. 6A), but may autonomously and non-autonomously express wing blade genes such as Dll (not shown)and nub (Fig. 6C)(Table 1). In wgclones, the non-autonomous gene expression is restricted to the nearest surrounding cells of the clone (see below and Discussion). The absence of transformation detected in these clones overexpressing wg is correlated with the absence of vg expression(Fig. 6B).
In the eye or leg imaginal discs, overexpression of wg in clones usually does not activate the ectopic expression of vg or show cuticular transformations (Table 1), but causes specific cuticular perturbations and gene expression alterations as shown elsewhere(Lee and Treisman, 2001;Royet and Finkelstein, 1997;Struhl and Basler, 1993;Theisen et al., 1996).
In the wing blade, the co-expression wg-vg in clones leads to the formation of tubular and perpendicular outgrowths. The size of the outgrowths is larger than in vg clones and is dependent on the distance from the wing margin (Fig. 7A,B). All the cells of the wg-vg clones are differentiated into wign margin sensory elements, as occurs in wg clones, whereas non autonomous territories of the outgrowth are differentiated into wing blade trichomes(Fig. 7A,B). In contrast to clones of wg, in which we detect homogenously high levels of Wg(Fig. 6D), clones ofwg-vg show low levels of Wg in some cases(Fig. 7E;Table 1), which are still sufficient to promote the autonomous differentiation of wing margin sensory elements. wg-vg clones do not express vg BE-lacZ orct (not shown) (Table 1). In the wing blade, the activation or repression of wing blade genes is equal to that observed in the wg overexpression clones.
Adult phenotypes (A-C) and imaginal disc expression patterns (D-G) caused by the co-expression of wg and vg in clones in the wing imaginal disc. Adult clones are labelled with f and delimited by a broken red line. In the imaginal discs, wg-vg clones are associated with GFP. Scale bars: 0.1 mm. (A) wg-vg clones in the wing blade may induce the appearance of tubular outgrowths. Notice the large non-autonomous growth induced by the clone. The age of clone initiation is 60±12 hours AEL. (B) Clones situated near the wing margin do not induce outgrowths. The age of clone initiation is 60±12 hours AEL. (C) In the notum, all the cells of wg-vg clones (delimited by a broken red line) differentiate into wing margin sensory elements and non-autonomously induce cuticular transformation to wing blade (delimited by a broken black line). The age of clone initiation is 36±12 hours AEL. (D) wg-vg clones (GFP)induced anywhere within the imaginal disc activate autonomously and non-autonomously the expression of vg (red; white arrowhead). The age of clone initiation is 60±12 hours AEL. (E) wg-vg (GFP) clones show heterogeneous or low levels of wg expression (red; white arrowheads). The age of clone initiation is 60±12 hours AEL. (F)wg-vg clones (GFP) autonomous and non-autonomously activate the expression of nub (red) anywhere in the imaginal disc. Notice that the non-autonomous expression of nub (red) is detected in cells surrounding the clone at long distances. The age of clone initiation is 60±12 hours AEL. (G) wg-vg clones (GFP) in the wing blade autonomously and non-autonomously repress bs (red; white arrowhead)expression, but in the wing hinge or in the notum, bs is autonomously repressed and non-autonomously induced. The age of clone initiation is 60±12 hours AEL.
Adult phenotypes (A-C) and imaginal disc expression patterns (D-G) caused by the co-expression of wg and vg in clones in the wing imaginal disc. Adult clones are labelled with f and delimited by a broken red line. In the imaginal discs, wg-vg clones are associated with GFP. Scale bars: 0.1 mm. (A) wg-vg clones in the wing blade may induce the appearance of tubular outgrowths. Notice the large non-autonomous growth induced by the clone. The age of clone initiation is 60±12 hours AEL. (B) Clones situated near the wing margin do not induce outgrowths. The age of clone initiation is 60±12 hours AEL. (C) In the notum, all the cells of wg-vg clones (delimited by a broken red line) differentiate into wing margin sensory elements and non-autonomously induce cuticular transformation to wing blade (delimited by a broken black line). The age of clone initiation is 36±12 hours AEL. (D) wg-vg clones (GFP)induced anywhere within the imaginal disc activate autonomously and non-autonomously the expression of vg (red; white arrowhead). The age of clone initiation is 60±12 hours AEL. (E) wg-vg (GFP) clones show heterogeneous or low levels of wg expression (red; white arrowheads). The age of clone initiation is 60±12 hours AEL. (F)wg-vg clones (GFP) autonomous and non-autonomously activate the expression of nub (red) anywhere in the imaginal disc. Notice that the non-autonomous expression of nub (red) is detected in cells surrounding the clone at long distances. The age of clone initiation is 60±12 hours AEL. (G) wg-vg clones (GFP) in the wing blade autonomously and non-autonomously repress bs (red; white arrowhead)expression, but in the wing hinge or in the notum, bs is autonomously repressed and non-autonomously induced. The age of clone initiation is 60±12 hours AEL.
In contrast to clones of either vg or wg alone,wg-vg clones in the wing hinge and notum cause transformation phenotypes everywhere. All the cells autonomously differentiate into wing margin and non-autonomously differentiate into wing blade trichomes(Fig. 7C). These transformations are correlated with the autonomous and non-autonomous expression of the wing blade genes studied(Fig. 7D,F,G;Table 1). In wg-vgclones, the expression of vg and other wing blade genes is autonomous and non-autonomous, but, in wg-vg clones the non-autonomous expression is extended to larger cell distances surrounding the clone than inwg clones (compare Fig. 6C with Fig. 7F andTable 1). In contrast tovg clones, the expression of bs in wg-vg clones is reduced, possibly because of their genetic specification as similar to cells of the wing margin region (Fig. 7G; Table 1).
In imaginal discs other than the wing, the co-expression of wg-vgin clones shows `mixed' phenotypes (Fig. 8A,B) and gene expression specificities similar to clones expressing vg ectopically (Fig. 8C,D; Table 1),again revealing regional restrictions to the induction of transformations and specific limitations of tissue to activation of wing blade gene expression. In contrast to clones of wg alone, wg-vg clones show transformation phenotypes, probably because of the presence of vgexpression. As in vg overexpression clones, wg-vg clones modify neither ap nor en expression(Table 1). These results suggest that the co-expression of wg-vg remains insufficient to promote a wing developmental program outside the wing imaginal disc.
Adult phenotypes (A,B) and imaginal disc expression patterns (C,D) caused by the ectopic clonal co-expression of wg and vg in the leg(A,C) and eye (B,E) imaginal discs. Adult clones are labelled with fand delimited by a broken red line. In the imaginal discs, clones are associated with GFP. Age of clone initiation is 60±12 hours AEL. Scale bars: 0.1 mm. (A) wg-vg clones induce a `mixed' cuticular differentiation pattern in the leg. Notice that in the outgrowth chaetae characteristic of the leg appear mixed with wing blade trichomes in a `salt and pepper' distribution. The inset shows the differences between wing blade trichomes of the clone (red arrowhead) and leg trichomes outside the clone(black arrowhead). (B) In the eye, all cells of the wg-vg clones autonomously differentiate into sensory elements of the wing margin. (C) In the leg, wg-vg clones activate the expression of wing blade genes such as bs (red), only in a subset of cells within the clones(arrowhead). Notice that the expression of bs is not detected in all the clones. (D) In the eye, wg-vg clones (GFP) autonomously activate the expression of nub (red) in all positions.
Adult phenotypes (A,B) and imaginal disc expression patterns (C,D) caused by the ectopic clonal co-expression of wg and vg in the leg(A,C) and eye (B,E) imaginal discs. Adult clones are labelled with fand delimited by a broken red line. In the imaginal discs, clones are associated with GFP. Age of clone initiation is 60±12 hours AEL. Scale bars: 0.1 mm. (A) wg-vg clones induce a `mixed' cuticular differentiation pattern in the leg. Notice that in the outgrowth chaetae characteristic of the leg appear mixed with wing blade trichomes in a `salt and pepper' distribution. The inset shows the differences between wing blade trichomes of the clone (red arrowhead) and leg trichomes outside the clone(black arrowhead). (B) In the eye, all cells of the wg-vg clones autonomously differentiate into sensory elements of the wing margin. (C) In the leg, wg-vg clones activate the expression of wing blade genes such as bs (red), only in a subset of cells within the clones(arrowhead). Notice that the expression of bs is not detected in all the clones. (D) In the eye, wg-vg clones (GFP) autonomously activate the expression of nub (red) in all positions.
DISCUSSION
The notion of `master gene', as applied to the gene ey(Halder et al., 1995),corresponds to a gene that by itself would trigger a developmental program that is independent of the tissue where it is expressed. Although this definition has been applied to vg(Kim et al., 1996), the present results indicate otherwise. The ectopic expression of vgelicits certain characteristics of `wing blade' development but is not sufficient for a complete transformation. The effect of vg depends on the time and genetic context of the tissue where it is overexpressed. Our results, according to other authors (Klein and Martínez-Arias, 1999), reveal a strong dependence ofvg on wg to initiate a wing blade developmental pathway. Wg by itself does not lead to tissue transformations. This cooperative effect between wg and vg remains insufficient in all tissues analysed, suggesting the existence of additional genes necessary to initiate and drive wing development. We do not know the molecular mechanisms that underlie the interaction between wg pathway and vg. However,the co-expression of vg with a construct of armadillo(arm) (transcriptional effector of wg pathway) using vg-G4 fails to promote the transformation of eye tissue (preliminary data, not shown). This result suggests that the interaction of wg andvg takes place upstream of arm and, therefore, outside of the cell nucleus. Whereas vg requires high levels of wgexpression to initiate wing development, the clones of vgoverexpression contain in later stages, low or null levels of wgexpression. Moreover, wg-vg co-overexpression clones can also show low levels of wg, even when wg is also mobilised in G4 territories or in Flip-out clones. These results suggest that vg may indirectly reduce wg expression once wing development is already initiated, and may explain why the transformed tissue in vg clones does not contain wing margin cuticular elements. The late repression ofwg seems to be important to specify territories of the wing blade depending on vg expression outside of the wing margin; if high levels of Wg are maintained all cells differentiate into wing margin chaetae. We conclude that wg and vg activities together specify wing margin territories, but vg alone specifies the remaining part of the wing blade.
The ectopic expression of vg or wg-vg in clones may cause outgrowths with wing histotypic characteristics or patterning perturbations in the notum, leg or eyes. The transformed tissues show `mixed' phenotypes or`mosaic' territories where, in a `salt and pepper' distribution, wing blade trichomes co-exist with notum or leg chaetae. Adult cuticular `mixed'phenotypes are correlated with the ectopic expression of wing blade genes in particular combinations (Table 1). However, expression of wing blade genes is detected only in some compact groups of cells within the clones. These results indicate that either vg or wg-vg are insufficient by themselves to displace all endogenous signals of identity, or that reciprocal non-autonomous influences between clonal cells and surrounding cells exist, reducing the expression of wing blade genes to groups of cells within clones. The change of wing blade genes expression in compact groups of cells in the disc and `mixed'(salt and pepper) cuticular phenotypes in the adult could result from cell interactions during patterning and cell rearrangements in pupal stages.
Transformations induced by overexpression of vg or wg-vgin clones and G4 territories are, as a rule, cell autonomous, except in the wing hinge, notum and corresponding tissues in the haltere. In the wing hinge the cells of the outgrowths outside the vg clones differentiate into wing blade territories and show gene expression patterns characteristic of the wing blade cells located between the proximal vg expression and the internal ring of wg in the wild-type disc. This suggests that the non-autonomous effects in vg clones could reproduce the wild-type intercalary growth induced by the confrontation of cells expressing proximal genes with distal genes. In the notum, vg clones located simultaneously in territories expressing and not expressing ap, and initiated in the wg expression domain, may non-autonomously recruit surrounding cells to express characteristic wing blade genes at long cell distances, as wg-vg clones do. Thus, vg together withwg expression is necessary to induce and extend the transformation over long distances outside the clones. In contrast to vg orwg-vg clones, wg clones do not show non-autonomous transformation phenotypes and expression of wing blade genes at long distances. The issue of whether the recruitment process is caused by Wg diffusion, or whether it results from intercalary growth induced by the confrontation between cells expressing proximal genes (genes of the notum) and cells expressing distal genes (wing blade genes), remains unresolved.
The expression of selector genes like Ubx and en is not modified by overexpression of vg or wg-vg, but is inherited and maintained. However, the expression of the selector gene ap can be modified or inherited in some tissues, such as the legs, to give DV identity.
The comparative analysis of vg with other morphogenetic genes suggests that vg acts as Dll, pnr or iro, rather than as a `master' or `selector of tissue' gene: vg is simply a component of the genetic combination that is necessary to initiate and drive wing blade development where vg is normally expressed. Interestingly,the function of vg, in addition to conferring territorial identity,may also non-autonomously recruit surrounding cells (`inductive assimilation'), changing their specific cuticular and gene expression patterns. This is related to its function as a local organiser of growth when it is expressed among cells with different positional or regional fates. Later in development, vg, in combination with other genes, activates an inventory of downstream wing genes that specify more discrete territories within the wing blade such as veins, interveins and sensory elements.
Acknowledgements
We thank S. Cohen, F. Casares, M. Affolter, A. Martínez-Arias, S. Carroll, M. Calleja, J. F. de Celis and the Developmental Studies Hybridoma Bank for providing flies and antibodies. We also thank C. Extavour, J. Resino,R. Barrio and other members of the laboratory for reading and discussing this work. R. Hernández and A. Hernando contributed with skillful technical assistance. This work was supported by grants from the Dirección General de Investigación Científica y Técnica and an institutional grant from the F. Ramón Areces to the Centro de Biología Molecular Severo Ochoa. L. A. B. is a fellow of the Consejo Superior de Investigaciones Científicas (CSIC) in collaboration with PACISA-GIRALT (I3P-BPD2001-1).