Surgically fragmented Drosophila appendage primordia (imaginal discs) engage in wound healing and pattern regulation during short periods of in vivo culture. Prothoracic leg disc fragments possess exceptional regulative capacity, highlighted by the ability of anterior cells to convert to posterior identity and establish a novel posterior compartment. This anterior/posterior conversion violates developmental lineage restrictions essential for normal growth and patterning of the disc, and thus provides an ideal model for understanding how cells change fate during epimorphic pattern regulation. Here we present evidence that the secreted signal encoded by hedgehog directs anterior/posterior conversion by activating the posterior-specific transcription factor engrailed in regulating anterior cells. In the absence of hedgehog activity, prothoracic leg disc fragments fail to undergo anterior/posterior conversion, but can still regenerate missing anterior pattern elements. We suggest that hedgehog-independent regeneration within the anterior compartment (termed integration) is mediated by the positional cues encoded by wingless and decapentaplegic. Taken together, our results provide a novel mechanistic interpretation of imaginal disc pattern regulation and permit speculation that similar mechanisms could govern appendage regeneration in other organisms.

Epimorphic limb regeneration requires that cells change fate in order to replace lost or damaged structures (Morgan, 1901). In most cases, very little is known about how molecular cues stimulate regenerating limb cells to change their identity, although it can be assumed that signals employed in development are reused during regeneration. We are investigating cell fate respecification in cut fragments of Drosophila imaginal discs as a means of gaining insight into pattern formation during normal development and as a general model for understanding fate changes during appendage regeneration in other organisms.

Drosophila thoracic imaginal discs originate as distinct polyclones of anterior (A) and posterior (P) embryonic disc precursor cells (Crick and Lawrence, 1975). In larvae, these cells heritably maintain their A/P identity through three successive instars of exponential proliferation, resulting in mature discs comprising distinct A and P lineage compartments (Garcia-Bellido et al., 1973; Garcia-Bellido, 1975; reviewed in Cohen, 1993). Interactions across the A/P compartment boundary serve as the basis for pattern formation and axis specification during development (Fig. 1A; reviewed in Blair, 1995; Lawrence and Struhl, 1996).

In the leg disc, cells assume position-specific identities in response to spatial cues provided by localized transcription of secreted signals encoded by wingless (wg; Couso et al., 1993), decapentaplegic (dpp; Masucci et al., 1990) and hedgehog (hh; Lee et al., 1992; Basler and Struhl, 1994). The requirement for localized transcription of these genes can be demonstrated by misexpressing them outside their endogenous expression patterns, either in random cell clones (flp-out; Struhl and Basler, 1993) or targeted expression domains (Gal4/UAS; Brand and Perrimon, 1993). In general, misexpression results in overproliferation, duplication of the limb proximodistal axis, loss of selected pattern elements with concomitant multiplication of others and occasional changes in appendage identity (Struhl and Basler, 1993; Maves and Schubiger, 1995, 1998a; Johnston and Schubiger, 1996). Surprisingly, these pattern abnormalities closely resemble the effects of cutting discs into fragments and allowing them to heal and re-grow during short periods of in vivo culture.

Disc growth and patterning normally culminate at the end of the third larval instar. At this stage, leg disc cells achieve rigidly determined states (Schubiger, 1968) and assume significantly reduced rates of cell division (Graves and Schubiger, 1982) as they await metamorphic cues that trigger differentiation. Relative stasis is maintained even if discs are afforded the opportunity to undergo continued proliferation and patterning during short periods of in vivo culture (Kiehle and Schubiger, 1985). Under identical in vivo conditions, however, cut and cultured third instar disc fragments engage in an aggressive program of proliferation and fate respecification called pattern regulation (Bryant, 1978). Despite considerable theoretical attention (French et al., 1976; Meinhardt, 1983; Kondo, 1992; Campbell and Tomlinson, 1995), the functional basis for this phenomenon is essentially unknown.

Pattern regulation initiates in vivo. Disc fragment cut edges heal together, establishing a region of localized cell proliferation and fate respecification termed a blastema (French et al., 1976). Fate changes in the blastema can be mapped by forcing cultured fragments to differentiate and then scoring for adult cuticle structures relative to uncultured control fragments. Only cells within or adjacent to the blastema participate in pattern regulation (Abbott et al., 1981), which results in either replacement of missing (regeneration) or reiteration of existing (duplication) disc pattern elements (Schubiger, 1971; Bryant, 1971). Generally, one fragment of a bisected disc regenerates while its complement duplicates. In addition, cells within a cultured fragment occasionally undergo changes in overall appendage identity termed transdetermination (Hadorn, 1965; reviewed in Hadorn, 1978; and Maves and Schubiger, 1998b).

Pattern regulation is described in leg, eye-antenna, wing, genital and labial disc fragments (reviewed in Bryant, 1978; Maves and Schubiger, 1998b). Among these, fragments of the prothoracic leg (L1) disc possess unique and remarkable regulative potential. In accordance with the maxim that one disc fragment regenerates while its complement duplicates, L1 anterior 1/4 (A1/4) fragments almost always regenerate, while complementing posterior 3/4 (P3/4) fragments almost always duplicate (Fig. 1B, Schubiger, 1971). In both cases, A cells convert directly to P identity (Abbott et al., 1981), in direct violation of lineage restrictions imposed during normal development (Garcia-Bellido et al., 1973, Garcia-Bellido, 1975). Curiously, mesothoracic leg (L2) discs lack this ability (M. C. G. and G. S., unpublished data), suggesting that the capacity for intercompartmental conversion derives from a unique property of the L1 disc.

Here we investigate the molecular basis for pattern regulation in cultured fragments of the L1 imaginal disc. Specifically, we ask why L1 A1/4 fragments regenerate while P3/4 fragments duplicate and how regulating A cells convert to P identity. We show that: (1) different regulative behaviors in the two fragments do not reflect a binary decision to regenerate or duplicate, but rather a uniform response to fragmentation; (2) HH signaling induces A/P conversion in both fragments by activating engrailed (en) in regulating anterior cells, and (3) a distinct population of squamous en/hh-expressing peripodial cells (specific to L1) fuse to the disc cut edge and act as a transient ectopic source of HH in cultured disc fragments. Based on these and other results, we suggest that fragmentation and wound-healing result in distinct A/P conversion and/or intracompartmental integration modes of pattern regulation. In addition, we discuss implications of our results for understanding cell fate changes during limb regeneration in other systems.

In vivo culture

Leg discs were dissected at 96 hours after egg deposition (AED), fragmented in insect Ringer’s solution (as described in Schubiger, 1971) and injected into the abdomens of freshly eclosed, etherized female hosts. Precise staging of disc donors was crucial because overaged pupally committed late third instar discs evert during culture, interfering with and/or obfuscating pattern regulation. The use of freshly eclosed hosts facilitated recovery of fragments and ensured a consistent, larva-like cell cycle period in regulating blastema cells (Kiehle and Schubiger, 1985). Host animals were maintained in the presence of males on standard food supplemented with yeast paste. Fragments were recovered from host abdomens in a drop of Ringer’s solution on a siliconized slide and either fixed for immunocytochemistry or injected into late-wandering third instar larvae for cuticular analysis after the general method of Ephrussi and Beadle (1936) as described in Maves and Schubiger (1998a). Adult leg cuticle and differentiated implants were mounted in Faure’s water-based medium and analyzed after the leg disc fate map of Schubiger (1971). In all experiments, discs were fragmented with tungsten needles and transferred in the tip of a P-20 Pipetteman lubricated by repetitive transfer of larval fat body.

Fly stocks and temperature shifts

We used wg-lacZ (Struhl and Basler, 1993) and P30 hh-lacZ (Lee et al., 1992) to identify gene expression domains and hhts2 (Ma et al., 1993) to block hh activity during in vivo culture. For these experiments, hhts2 larvae were raised to the late third instar at 15°C (permissive) and shifted to 25°C (restrictive) 6 hours prior to fragmentation and injection into hosts. Host animals were raised at 25°C for 4 days for experimental groups or shifted back to 15°C for 10-14 days for controls. We used 25°C rather than higher temperatures (28°C, Basler and Struhl, 1994) to avoid potentially confounding effects of temperature shock on wound healing and regulative processes. At 29°C, hhts2 A1/4 fragments were not recoverable from hosts, presumably due to lack of growth or degeneration (data not shown). For clone induction (described below) we used hs-flp; Act5c>CD2>Gal4, UAS-GFP (Pignoni and Zipursky, 1997; Neufeld et al., 1998). Ectopic wg was induced using the flp-out technique of Struhl and Basler (1993) after the method of Maves and Schubiger (1995).

Clone induction

Random Green Fluorescent Protein (GFP)-labeled clones were induced with a 30 minute-long 37°C heat shock to hs-flp; Act5c>CD2>Gal4, UAS-GFP larvae at 72 hours AED. At 96 hours AED, experimental discs were fragmented and cultured in vivo for an additional 72 hours. After culture, we used confocal microscopy to analyze GFP+ clone position relative to the nascent A/P boundary (as defined by EN) in explanted duplicated disc fragments. Girton and Russell (1980) showed that as few as 20 cells participate in pattern duplication. Accordingly, we induced a relatively large number of clones in the disc in order to observe clones in the duplicate. In uncut control discs, our protocol produced a large number of small (4-to 8-cell) clones at the time of fragmentation. In experimental material, many clones were observed in non-regulating parts of the disc fragment, while an average of just 1.3 clones was observed in the duplicated area. 1-hour heat shocks resulted in too many clones for meaningful analysis (data not shown).

Immunocytochemistry and imaging

Whole discs and cultured disc fragments were fixed, immunostained and imaged after the methods of Maves and Schubiger (1998a) except that 4% paraformaldehyde fixation was for 1.5 hours on ice and disc material was preblocked in Normal Goat Serum for 1 hour prior to staining. Primary antibodies were used at the following dilutions: mouse anti-EN 4D9 (gift of Tom Kornberg) 1:1000; rabbit anti-phosphohistone H3 mitosis marker (Upstate Biotech) 1:2000; rabbit (Cappel) or mouse (Boehringer Mannheim) anti-β-galactosidase 1:500; mouse anti-WG (gift of Roel Nusse) 1:5, and rabbit anti-DLL (gift of Grace Panganiban) 1:200. All rinses were carried out in concave glass culture dishes to reduce loss of material. Images were collected with a Bio-Rad MRC 600 confocal microscope system and assembled into figures using Adobe Photoshop 5.0 and Canvas 5. Leg discs and disc fragments were oriented using the stalk (anterior-dorsal) and nerve (posterior-ventral). The position of the nerve allowed us to discern between endogenous and duplicated gene expression domains in cultured P3/4 disc fragments.

EN expression during pattern regulation

The observation that A cells switch to P identity during pattern regulation in L1 disc fragments (Abbott et al., 1981) is a curious exception to the rule of compartmental lineage restriction during Drosophila appendage development. This prompted us to ask how A/P conversion is achieved on the molecular level. Since EN is normally expressed in all P cells (Kornberg et al., 1985; Fig. 2A), we tested whether A/P conversion correlated with activation of EN in cultured disc fragments. Third instar L1 discs were cut into A1/4 and P3/4 fragments and cultured in vivo. Approximately 50% of A1/4 (11/20) and P3/4 (17/34) fragments produced novel EN expression domains (Fig. 2B,C) within 96 hours in vivo. The size and shape of nascent posterior compartments varied from small clusters of EN-expressing cells to large domains almost indistinguishable from the endogenous P compartment – suggesting that stochastic variations (subtle differences in cut sites, disc morphology and precise age of the disc donor) might influence the timing and extent of pattern regulation. In a significant number of cases, however, A1/4 disc fragments ‘regenerated’ a new P compartment, which restored the proportions of a whole disc (Fig. 2B), and P3/4 fragments ‘duplicated’ a new EN-expression domain in mirror-symmetric opposition to the endogenous P compartment (Fig. 2C).

EN is activated in anterior cells

Previously, clonal analysis has been used to show that antennal cells convert directly to wing identity (Gehring, 1967) and that A leg cells can convert directly to P identity during pattern regulation (Abbott et al., 1981; Girton and Russell, 1981). In these studies, conclusions about clone behavior were inferred from analysis of genetically marked cuticle structures. To directly observe the cellular origin of nascent EN domains in regulating disc fragments, we induced random GFP-labeled clones by heat-shocking hs-flp; Act5c>CD2>GAL4, UAS-GFP larvae 1 day prior to fragmentation and injection into host animals. Fig. 3A shows an uncut control disc, note the relatively uniform distribution of small GFP+ clones in both compartments. P3/4 disc fragments containing GFP+ clones were cultured in vivo for an additional 72 hours. Duplicated fragments were then scored for the presence of marked clones, that contained both A (EN) and P (EN+) cells (Fig. 3B,C).

17/34 clone-bearing P3/4 fragments produced a duplicated EN domain; 12 of these had a well-defined novel compartment boundary suitable for detailed analysis. Of these 12 duplicates, two contained no GFP+ clones, six had one clone, two had two clones and two had three clones in the duplicated area. Four duplicates contained a contiguous GFP+ clone comprising both A (EN) and P (EN+) cells (Fig. 3C). Because our protocol generated an average of 1.3 clones/duplicated area (16 clones/12 duplicates), we infer that these four cases represent single large clones that violate the A/P lineage restriction rather than two or more smaller clones fused exactly at the nascent A/P boundary. These observations confirm the results of Abbott et al. (1981), demonstrating that A cells directly convert to P identity during pattern regulation in L1 disc fragments.

The size and distribution of GFP+ clones in duplicated discs provided additional insights into the dynamics of proliferation in the blastema. Nascent P compartments were never entirely composed of GFP+ cells (0/12), indicating that A/P conversion is a polyclonal event in duplicating disc fragments. We also observed that single GFP+ clones often occupied up to 10-20% of the whole duplicated area, suggesting that as few as 5-10 founder cells participated in duplicative growth. This agrees with a previous clonal analysis, which suggests that similar cell numbers generate disc duplications during cell-lethality-mediated pattern regulation (Girton and Russell, 1980).

HH activates en in regulating disc fragments

Endogenous en expression in the L1 disc is activated during embryogenesis by the transcription factor Fushi tarazu (FTZ; Ingham et al., 1988). However, ftz-lacZ expression was not detected in cultured disc fragments (data not shown), suggesting an alternate mechanism for activation of en during pattern regulation. Ectopic hh activity induces en in A cells of the abdominal tergites (Struhl et al., 1997), wing (de Celis and Ruiz Gomez, 1995; Guillen et al., 1995) and leg imaginal discs (Laura Johnston, personal communication), indicating that hh might be required to activate en in cultured disc fragments. To test this hypothesis, L1 discs from hhts2 larvae raised at permissive temperature (15°C) were fragmented and cultured under restrictive (25°C) conditions (Fig. 2). In the absence of HH, cultured A1/4 fragments were small and extremely difficult to recover from hosts. Of those recovered, only 13% (3/23) regenerated new EN domains, compared with 37% (7/19) at the permissive temperature. The effect of hh loss was more pronounced in P3/4 fragments; none (0/27) duplicated at restrictive temperature while 39% (15/38) duplicated under permissive conditions. Loss of hh clearly blocked en-activation in both fragments (Fig. 2D-G). At non-permissive temperature, hhts2 disc fragments also failed to produce regenerated/ duplicated posterior leg structures upon forced differentiation in host larvae (Fig. 4). Taken together, these experiments demonstrate that hh is necessary to activate en and establish a new P compartment in cultured L1 disc fragments. Because loss of hh did not affect endogenous EN expression in cultured P3/4 fragments (Fig. 2G), we can rule out the possibility that HH is simply required to maintain EN during culture.

Without HH, P3/4 fragments switch from a duplicative to a regenerative program

P3/4 fragments normally duplicate existing pattern elements, but only rarely regenerate missing structures. In the absence of hh, however, P3/4 fragments gained the ability to regenerate all missing anterior leg structures (Fig. 4). This observation suggests a paradox: if hh is required for normal development, how do P3/4 fragments regenerate in its absence? During normal leg disc development, HH secreted from P cells acts primarily through induction of wg and dpp in A cells along the compartment boundary (Basler and Struhl, 1994). Secreted WG and DPP subsequently pattern both compartments through a variety of mechanisms (reviewed in Fig. 1), including activation of Distalless (Dll) at the center of the disc (Lecuit and Cohen, 1997). Once established, late-third instar wg and dpp expression domains may be sufficient to direct anterior pattern regulation even if HH levels are severely reduced. In intact discs, ectopic WG interacts with DPP to cause overgrowth and anterior pattern duplications (Johnston and Schubiger,1996), suggesting that wg/dpp interactions might be sufficient to direct anterior compartment regeneration without direct input from hh. This assertion is supported by the fact that hh is not required for maintenance of dpp in third instar wing discs (Capdevila et al., 1994), and predicts that WG and DLL domains are maintained in the absence of HH in cultured disc fragments.

To test this hypothesis, P3/4 fragments cultured in the absence of hh were immunostained with antibodies directed against WG and DLL (Fig. 5). Fig. 5C-F shows that hhts2 P3/4 fragments cultured at restrictive temperature maintained WG and DLL at reduced levels (relative to wild-type) in their endogenous domains. These fragments did not produce duplicated WG or DLL domains. At permissive temperature WG and DLL levels were also significantly reduced (possibly due to weak hypomorphic effects of hhts2), but both domains were duplicated or expanded in 9/11 fragments analyzed (Fig. 5E,F). These data show that wild-type levels of HH are required for activation, but not maintenance, of WG and DLL domains in cultured disc fragments. We conclude that P3/4 fragments regenerate missing anterior pattern elements in an hh-independent fashion and suggest that this occurs by a process of integration between established wg/dpp domains juxtaposed through wound healing.

To confirm that hh is not directly required for leg pattern homeostasis after the third instar, hhts2 larvae were shifted to the restrictive temperature late in development (90 hours AED, the same time used in hhts2 fragmentation experiments). These larvae pupariated but categorically failed to eclose. Analysis of pharate legs revealed largely normal cuticle patterns, with the exception of loss of the anterior claw at the distal tip of L1, L2 and L3 (data not shown). Although claw specification has previously been attributed to dpp function (Held et al., 1994), our observations are consistent with a reported late and direct requirement for hh in specification of intervein distances and sensory organ precursors along the A/P boundary in wing discs (Mullor et al., 1997), and thus a late role for HH in direct specification of some structures at the leg disc compartmental boundary.

Peripodial membrane cells act as a source of HH during pattern regulation

Peripodial cells contribute to a squamous epithelium that covers the columnar epithelium of the disc proper and participates in disc eversion and metamorphosis (Milner et al., 1984; reviewed in Fristrom and Fristrom, 1993). In wing disc fragments, peripodial cells form a transient heterotypic contact with regulating columnar cells at the site of wound healing (Reinhardt et al., 1977). We find substantial evidence that a specific population of dorsolateral peripodial membrane cells act as the source of HH in cultured L1 disc fragments (Fig. 6). In both A1/4 and P3/4 disc fragments, peripodial cells express hh-lacZ and EN prior to and during in vivo culture (Fig. 6A,B), and appear to fuse with anterior cells in the regenerating disc epithelium during wound healing (Fig. 6B,C). In A1/4 disc fragments, hh-lacZ is not expressed in nascent P cells until after a new EN-domain is clearly visible (Fig. 6B). This rules out the possibility that loss of A/P conversion in hhts2 mutants (Fig. 2) resulted from hh-dependent growth effects within the nascent P compartment and makes peripodial cells the sole potential source of HH in regenerating A1/4 fragments (Fig. 6A,B,E). These observations indicate that HH from peripodial cells activates en during a transient fusion between peripodial and columnar cells at the wound site. As confirmation, EN-expressing peripodial cells still fuse to the wounded epithelium in the absence of HH, but do not induce EN in surrounding columnar cells (Fig. 6D).

L2 discs lack EN/HH-expressing peripodial cells and cannot regenerate posterior leg structures

Ectopic HH induces duplicated en domains in intact L1 and L2 discs (Laura Johnston, personal communication), and since pattern formation in L1 and L2 discs is nearly indistinguishable, one might expect L1 and L2 disc fragments to regulate identically. However, L2 A1/4 fragments do not regenerate (data not shown), and L2 P3/4 fragments rarely duplicate but rather regenerate missing anterior structures during culture (Fig. 4). These regulative behaviors are nearly identical to those of hhts2 L1 fragments cultured at restrictive temperatures, suggesting that reduction of hh converts L1 fragments to an L2 mode of pattern regulation. Do L2 discs lack the hh/EN-expressing peripodial cells proposed to activate en in L1 disc fragments? Fig. 6E and F show that L2 discs have many fewer hh/EN-expressing peripodial cells (about 0-5 cells in L2 versus 20-30 in L1), explaining their diminished capacity for A/P conversion (Fig. 4). Because L1 and L2 discs are essentially identical in all other regards, these results are consistent with a unique role for hh-expressing peripodial cells in induction of A/P conversion in cultured L1 disc fragments.

A new interpretation of pattern regulation in L1 disc fragments

Analysis of leg disc fragments cultured at reduced levels of HH has allowed us to distinguish two distinct modes of pattern regulation. During integration, missing pattern is regenerated in a spatially appropriate manner within the A compartment. This process does not require wild-type HH signaling, and may proceed by interaction between established wg and dpp domains brought into contact through wound healing. During conversion, A cells activate EN and assume P identity. This process is strictly hh-dependent and occurs only in L1 disc fragments. Cultured fragments of L2 discs do not undergo A/P conversion. This observation led us to consider why L1 disc fragments possess such unique regulative potential.

In cultured L1 disc fragments, hh-expressing peripodial cells fuse with columnar cells at the wound site and provide a transient ectopic source of HH, which stimulates A/P conversion. Following this idea, L2 disc fragments lack hh-expressing peripodial cells (Fig. 6E,F) and hence fail to undergo A/P conversion during culture. The presence of the hh-expressing peripodial cells may imply an L1-specific developmental function. For example, left and right L1 primordia are distinguished from L2 and L3 in that they fuse together sometime during larval development. hh/EN-expressing peripodial cells localize to the precise dorsolateral position of L1 disc fusion (Fig. 6G), and are not detected in L2 or L3 discs. We infer that peripodial HH signaling could mediate right/left fusion in developing L1 discs and propose that other metamorphic disc fusion events could similarly depend on (this or other) intercellular signaling mechanisms.

A role for EN in anterior cells during development?

In wing and leg discs, EN expression is restricted to P cells for most of development, but becomes detectable in A cells along the A/P boundary late in the third larval instar (Blair, 1992; Maschat et al., 1998; Fig. 7A). Consistent with its role in A/P conversion, HH signaling is also thought to mediate en activation in these anterior boundary cells (de Celis and Ruiz-Gomez, 1995; Guillen et al., 1995). The developmental function of EN in A cells is not known, although recent evidence suggests a potential role in inhibition of dpp along the compartmental boundary (Mullor et al., 1997; Maschat et al., 1998). Levels of dpp are critical for normal growth and patterning; loss of function leads to truncated and defective appendages (Gelbart, 1989; Held et al., 1994) while ectopic expression elicits overgrowth and pattern respecification (Nellen et al., 1996). Since EN is known to inhibit dpp (Raftery et al., 1991), and dpp is not coexpressed with en in boundary cells, late-activated EN could inhibit dpp (Mullor et al., 1997) as a mechanism to restrict dpp-dependent growth and/or patterning at the end of larval life.

Why doesn’t hh-mediated en-activation result in A/P conversion and pattern regulation at the compartmental boundary? Maschat et al. (1998) show that A boundary cells, which express EN, activate polyhomeotic (ph), which in turn represses hh. The authors argue that the en/ph circuit prevents hh-dependent induction of patched (ptc), Cubitis interruptus (Ci) and dpp in adjacent posterior cells – proscribing ‘invasion’ of the posterior compartment by anterior pattern markers. Similarly, results from de Celis and Ruiz-Gomez (1995) show that hh can activate en in A wing cells and that groucho inhibits expression of hh in A cells along the compartment boundary. We propose that during pattern regulation, A/P conversion occurs in cells distant from the A/P boundary which initially lack (or somehow downregulate) expression of these or other potential inhibitors of hh (such as Ci; Domínguez et al., 1996). In this scenario, EN activation in the blastema could lead to production and secretion of HH, subsequent wg and dpp activation in neighboring A cells, respecification of the proximodistal axis via Dll, and thus a partial recapitulation of leg disc development.

Parallels with ectopic gene expression

Consistent with the view that pattern regulation is driven by spatiotemporally inappropriate gene interactions, ectopic gene expression can clearly mimic the effects of disc fragmentation. For example, ectopic hh induces A/P conversion and pattern duplication in developing abdominal tergites (Struhl et al., 1997) and wing and leg imaginal discs (Guillen et al., 1995; de Celis and Ruiz-Gomez, 1995; Laura Johnston, personal communication). Previously, hh-mediated anterior leg pattern duplications have been attributed to positionally inappropriate activation of wg and dpp (Basler and Struhl, 1994; Lepage et al., 1995). Using a wg-lacZ transgene, we found that wg transcription is also activated at the precise location where hh-expressing peripodial cells fuse to the wound site in cultured L1 disc fragments (Fig. 7A,B). This ‘ectopic’ activation of wg at the wound site extends the analogy between hh-misexpression and fragmentation experiments, and explains why cultured P3/4 fragments differentiate duplicated anterior-ventral cuticular patterns akin to those elicited by clones of constitutive wg activity in developing discs (Struhl and Basler, 1993; see Fig. 7C).

Activation of wg at the wound site also provides a compelling explanation for leg-to-wing transdeterminative phenomena observed in cultured L1 disc fragments (Schubiger, 1971). In unfragmented leg discs, ectopic wg interacts with dpp to stimulate localized cell proliferation (Johnston and Schubiger, 1996) and leg-to-wing transdetermination via activation of the wing-specific transcription factor Vestigal (Maves and Schubiger, 1995, 1998; Johnston and Schubiger, 1996). A similar pathway could be executed during culture of L1 disc fragments, since wg is activated in close proximity to dpp-expressing cells. It is interesting to note that L2 disc fragments, which lack a peripodial source of HH, fail to transdetermine during culture (G.S., unpublished observations), suggesting that HH is needed to activate wg at the wound and wg is required for transdetermination to wing.

Parallels with vertebrate limb regeneration?

Imaginal disc pattern regulation is an experimental phenomenon, and cannot be directly compared to the highly specialized and adaptively-significant process of amphibian limb regeneration. However, the striking similarities in limb pattern regulation between flies and vertebrates (French et al., 1976) may suggest a common mechanistic basis. Saunders and Gasseling (1968), studying chick limb buds, demonstrated that grafting posterior cells into anterior limb sites induces mirror-symmetric distal limb duplications. It has since been shown that vertebrate Sonic hedgehog (Shh) is the molecular agent of this phenomenon (Riddle et al., 1993). The effects of Shh misexpression are inescapably analogous to those caused by HH in cultured P3/4 disc fragments, particularly respecification of A cells to P identity and duplication of distal limb primordia (see Fig. 7D). Previous authors have cited the parallel effects of HH misexpression in flies and vertebrates as evidence that hh-family members play an evolutionarily conserved role in appendage patterning (Fietz et al., 1994). Our results are consistent with this view and permit speculation that in vertebrates, as in flies, the ability to regenerate particular limb structures may be predicated upon the availability of SHH (or the ability to express or respond to SHH) at the site of limb transection. The recently reported expression of Shh in regenerating newt limb blastemas supports this notion (Imokawa et al., 1997).

We thank Konrad Basler, Laura Johnston, Tom Kornberg and Margrit Schubiger for critical readings of the manuscript, Roel Nusse and Tom Kornberg for antibodies, and Laura Johnston and Lisa Maves for helpful insights during the course of the work. This research was supported by NIH grant GM58282 and NSF grant 96-00662 to (G. S.). M. C. G. is supported by an ARCS fellowship and Molecular and Cellular Biology Training Grant number PHS NRSA T32 GM07270 from NIGMS.

Abbott
,
L. C.
,
Karpen
,
G. H.
and
Schubiger
,
G.
(
1981
).
Compartmental restrictions and blastema formation during pattern regulation in Drosophila imaginal leg discs
.
Dev. Biol.
87
,
64
75
.
Basler
,
K.
and
Struhl
,
G.
(
1994
).
Compartment boundaries and the control of Drosophila limb pattern by hedgehog protein
.
Nature
368
,
208
214
.
Blair
,
S. S.
(
1992
).
Engrailed expression in the anterior lineage compartment of the developing wing blade of Drosophila
.
Development
115
,
21
33
.
Blair
,
S. S.
(
1995
).
Compartments and appendage development in Drosophila
.
BioEssays
17
,
299
309
.
Brand
,
A. H.
and
Perrimon
,
N.
(
1993
).
Targeted gene expression as a means of altering cell fates and generating dominant phenotypes
.
Development
118
,
401
415
.
Brook
,
W. J.
and
Cohen
,
S. M.
(
1996
).
Antagonistic interactions between wingless and decapentaplegic responsible for dorsal-ventral pattern in the Drosophila leg
.
Science
273
,
1373
1377
.
Bryant
,
P. J.
(
1971
).
Regeneration and duplication following operations in situ on the imaginal discs of Drosophila melanogaster
.
Dev. Biol.
26
,
637
651
.
Bryant
,
P. J.
(
1978
).
Pattern formation in imaginal discs
. In
Biology of Drosophila
(eds.
M.
Ashburner
and
T. R. F
Wright
), Vol.
2c
, pp.
229
335
.
New York
:
Academic Press
.
Campbell
,
G.
and
Tomlinson
,
A.
(
1995
).
Initiation of the proximodistal axis in insect legs
.
Development
121
,
619
628
.
Capdevila
,
J.
,
Estrada
,
M. P.
,
Sánchez-Herrero
E.
and
Guerrero
,
I.
(
1994
).
The Drosophila segment polarity gene patched interacts with decapentaplegic in wing development
.
EMBO J.
13
,
71
82
.
Cohen
,
S. M.
(
1993
).
Imaginal disc development
. In
The Development of Drosophila melanogaster
Vol.
II
(eds.
M.
Bate
and
A. Martinez
Arias
) pp.
747
841
.
Cold Spring Harbor Laboratory Press
.
Couso
,
J. P
,
Bate
,
M.
and
Martinez-Arias
,
A.
(
1993
).
A wingless-dependent polar coordinate system in Drosophila imaginal discs
.
Science
259
,
484
489
.
Crick
,
F. H. C.
and
Lawrence
,
P. A.
(
1975
).
Compartments and polyclones in insect development
.
Science
189
,
340
347
.
de Celis
,
J. F.
and
Ruiz Gómez
,
M.
(
1995
).
groucho and hedgehog regulate engrailed expression in the anterior compartment of the Drosophila wing
.
Development
121
,
3467
3476
.
Diaz-Benjumea
,
F. J.
,
Cohen
,
B.
and
Cohen
,
S. M.
(
1994
).
Cell interaction between compartments establishes the proximal-distal axis of Drosophila legs
.
Nature
372
,
175
179
.
Domínguez
,
M.
,
Brunner
,
M.
,
Hafen
,
E.
and
Basler
,
K.
(
1996
).
Sending and receiving the hedgehog signal: control by the Drosophila Gli protein Cubitus interruptus
.
Science
272
,
1621
1625
.
Ephrussi
,
B.
and
Beadle
,
G. W.
(
1936
).
A technique of transplantation for Drosophila. (1936)
.
Am. Nat
.
79
,
218
225
.
Fietz
,
M. J.
,
Concordet
,
J. P.
,
Barbosa
,
R.
,
Johnson
,
R.
,
Krauss
,
S.
,
McMahon
,
A. P.
,
Tabin
,
C.
and
Ingham
,
P. W.
(
1994
).
The hedgehog gene family in Drosophila and vertebrate development
.
Development
1994
Supplement,
43
51
.
French
,
V.
Bryant
,
P. J.
and
Bryant
,
S. V.
(
1976
).
Pattern regulation in epimorphic fields
.
Science
193
,
969
981
.
Fristrom
,
D.
and
Fristrom
,
J. W.
(
1993
).
The metamorphic development of the adult epidermis
. In
The Development of Drosophila melanogaster
Vol.
II
(eds.
M.
Bate
and
A. Martinez
Arias
) pp.
747
841
.
Cold Spring Harbor Laboratory Press
.
Garcia-Bellido
,
A.
,
Ripoll
,
R.
and
Morata
,
G.
(
1973
).
Developmental compartmentalization of the wing disc of Drosophila
.
Nature New Biol.
245
,
251
253
.
Garcia-Bellido
,
A.
(
1975
).
Genetic control of wing disc development in Drosophila. Ciba Foundation Symposium
,
Cell Patterning
pp.
161
-
182
.
Associated Scientific Publishers
,
London
.
Gehring
,
W. J.
(
1967
).
Clonal analysis of determination dynamics in cultures of imaginal discs in Drosophila melanogaster
.
Dev. Biol.
16
,
438
456
.
Gelbart
,
W. M.
(
1989
).
The decapentaplegic gene: a TGF-beta homologue controlling pattern formation in Drosophila
.
Development
107
Supplement,
65
74
.
Girton
,
J. R.
and
Russell
,
M. A.
(
1980
).
A clonal analysis of pattern duplication in a temperature-sensitive cell-lethal mutant of Drosophila melanogaster
.
Dev. Biol.
77
,
1
21
.
Girton
,
J. R.
and
Russell
,
M. A.
(
1981
).
An analysis of compartmentalization in pattern duplications Induced by a cell-lethal mutation in Drosophila
.
Dev. Biol.
85
,
55
64
.
González
,
F.
Swales
,
L.
Bejsovec
,
A.
Skaer
,
H.
and
Martinez-Arias
,
A.
(
1991
).
Secretion and movement of wingless protein in the epidermis of the Drosophila embryo
.
Mech. Dev.
35
,
43
54
.
Graves
,
B. J.
and
Schubiger
,
G.
(
1982
).
Cell cycle changes during growth and differentiation of imaginal leg discs in Drosophila melanogaster
.
Dev. Biol
.
93
,
104
110
.
Guillen
,
I.
,
Mullor
,
J. L.
,
Capdevila
,
J.
,
Sanchez-Herrero
,
E.
,
Morata
,
G.
and
Guerrero
,
I.
(
1995
).
The function of engrailed and the specification of Drosophila wing pattern
.
Development
121
,
3447
3456
.
Hadorn
,
E.
(
1965
).
Problems of determination and transdetermination
.
Brookhaven Symp. Biol.
18
,
148
161
.
Hadorn
,
E.
(
1978
).
Imaginal discs: transdetermination
.
The Genetics and Biology of Drosophila
(eds.
M.
Ashburner
and
T. R. F.
Wright
), Vol.
2c
, pp.
556
617
.
New York
:
Academic Press
.
Held
,
L. I.
,
Heup
,
M. A.
,
Sappington
,
J. M.
and
Peters
,
S. D.
(
1994
)
Interactions of decapentaplegic, wingless and Distal-less in the Drosophila leg
.
Roux’s Arch. Dev. Biol.
203
,
310
319
.
Imokawa
,
Y.
and
Yoshizato
,
K
. (
1997
).
Expression of Sonic hedgehog gene in regenerating newt limb blastemas recapitulates that in developing limb buds
.
Proc. Natl. Acad. Sci. USA
94
,
9159
9164
.
Ingham
,
P. W.
,
Baker
,
N. E.
and
Martinez-Arias
,
A.
(
1988
).
Regulation of segment polarity genes in the Drosophila blastoderm by fushi tarazu and even skipped
.
Nature
331
,
73
75
.
Jiang
,
J.
and
Struhl
,
G.
(
1996
).
Complementary and mutually exclusive activities of decapentaplegic and wingless organize axial patterning during Drosophila leg development
.
Cell
86
,
401
409
.
Johnston
,
L. A.
and
Schubiger
,
G.
(
1996
).
Ectopic expression of wingless in imaginal discs interferes with decapentaplegic expression and alters cell determination
.
Development
122
,
3519
3529
.
Kiehle
,
C. P.
and
Schubiger
,
G.
(
1985
)
Cell proliferation changes during pattern regulation in imaginal leg discs of
Drosophila melanogaster. Dev. Biol.
109
,
336
346
.
Kondo
,
S.
(
1992
).
A mechanistic model for morphogenesis and regeneration of limbs and imaginal discs
.
Mech. Dev
.
39
,
161
170
.
Kornberg
T.
,
Siden
,
I.
,
O’Farrell
,
P.
and
Simon
,
M.
(
1985
).
The engrailed locus of Drosophila: in situ localization of transcripts reveals compartment-specific expression
.
Cell
40
,
45
53
.
Lawrence
,
P. A.
and
Stuhl
,
G.
(
1996
).
TMorphogens, compartments and pattern: lessons from Drosophila?
Cell
85
,
951
961
.
Lecuit
,
T.
and
Cohen
,
S. M.
(
1997
).
Proximal-distal axis formation in the Drosophila leg
.
Nature
388
,
139
145
.
Lee
,
J. J.
,
von Kessler
,
D. P.
,
Parks
,
S.
and
Beachy
,
P. A.
(
1992
).
Secretion and localized transcription suggest a role in positional signaling for products of the segmentation gene hedgehog
.
Cell
71
,
33
50
.
Lepage
,
T.
,
Cohen
,
S. M.
,
Diaz-Benjumea
,
F. J.
and
Parkhurst
,
S. M.
(
1995
).
Signal transduction by cAMP-dependent protein kinase A in Drosophila limb patterning
.
Nature
373
,
711
715
.
Ma
,
C.
,
Zhou
,
Y.
,
Beachy
,
P. A.
and
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
.
Masucci
,
J. D.
Miltenberger
,
R. J.
and
Hoffmann
,
F. M.
(
1990
).
Pattern-specific expression of the Drosophila decapentaplegic gene in imaginal disks is regulated by 3’ cis-regulatory elements
.
Genes Dev
.
4
,
2011
2023
.
Maschat
,
F.
,
Serrano
,
N.
,
Randsholt
,
N.
and
Geraud
,
G.
(
1998
).
engrailed and polyhomeotic interactions are required to maintain the A/P boundary of the Drosophila developing wing
.
Development
125
,
2771
2780
.
Maves
,
L.
and
Schubiger
,
G.
(
1995
).
Wingless induces transdetermination in developing Drosophila imaginal discs
.
Development
121
,
1263
1272
.
Maves
,
L.
and
Schubiger
,
G.
(
1998a
).
A molecular basis for transdetermination in Drosophila imaginal discs: interactions between wingless and decapentaplegic signaling
.
Development
125
,
115
24
.
Maves
,
L.
and
Schubiger
,
G.
(
1998b
).
Cell Determination and Transdetermination in Drosophila Imaginal Discs
.
Current Topics in Developmental Biology
43
,
115
151
.
Meinhardt
,
H.
(
1983
).
Cell determination boundaries as organizing regions for secondary embryonic fields
.
Dev. Biol.
96
,
375
385
.
Milner
,
M. J.
,
Bleasby
,
A. J.
and
Kelly
,
S. L.
(
1984
).
The role of the peripodila membrane of leg and wing imaginal discs of Drosophila melanogaster during evagination and differentiation in vitro
.
Roux’s Arch. Dev. Biol.
193
,
180
186
.
Morgan
,
T. H.
(
1901
).
Regeneration.
New York
:
The MacMillan Company
.
Mullor
,
J. L.
,
Calleja
,
M.
,
Capdevila
,
J.
and
Guerrero
,
I.
(
1997
).
Hedgehog activity, independent of decapentaplegic, participates in wing disc patterning
.
Development
124
,
1227
1237
.
Nellen
,
D.
,
Burke
,
R.
,
Struhl
,
G.
and
Basler
,
K.
(
1996
).
Direct and long-range action of a DPP morphogen gradient
.
Cell
85
,
357
368
.
Neufeld
,
T. P.
,
de la Cruz
,
A. F.
,
Johnston
,
L. A.
and
Edgar
,
B. A.
(
1998
).
Coordination of growth and cell division in the Drosophila wing
.
Cell
93
,
1183
1193
.
Penton
,
A.
and
Hoffmann
,
F. M.
(
1996
).
Decapentaplegic restricts the domain of wingless during Drosophila limb patterning
.
Nature
382
,
162
164
.
Pignoni
,
F.
and
Zipursky
,
S.
(
1997
).
Induction of Drosophila eye development by decapentaplegic
.
Development
124
,
271
278
.
Raftery
,
L. A.
,
Sanicola
,
M.
,
Blackman
,
R. K.
and
Gelbart
,
W. M.
(
1991
).
The relationship of decapentaplegic and engrailed expression in Drosophila imaginal disks: do these genes mark the anterior-posterior compartment boundary?
Development
113
,
27
33
.
Reinhardt
,
C. A.
,
Hodgkin
,
N. M.
and
Bryant
,
P. J.
(
1977
).
Wound healing in the imaginal discs of Drosophila. I. Scanning electron microscopy of normal and healing wing discs
.
Dev. Biol.
60
,
238
257
.
Riddle
,
R. D.
,
Johnson
,
R. L.
,
Laufer
,
E.
and
Tabin
,
C.
(
1993
).
Sonic hedgehog mediates the polarizing activity of the ZPA
.
Cell
75
,
1401
1416
.
Saunders
,
J.
and
Gasseling
(
1968
).
Ectodermal-mesenchymal interaction in the origin of limb symmetry
. In
Epithelial-Mesenchymal Interaction
(eds.
R.
Fleischmayer
and
R. E.
Billingham
) pp.
78
97
.
Baltimore
:
Williams and Wilkins
.
Schubiger
,
G.
(
1968
).
TAnlageplan, Determinationszustand und Transdeterminationsleistungen der männlichen Vorderbeinscheibe von Drosophila melanogaster
.
Wilhelm Roux’ Arch. EntwMech. Org.
160
,
9
40
.
Schubiger
,
G.
(
1971
).
TRegeneration, Duplication and Transdetermination in fragments of the leg disc of Drosophila melanogaster
.
Dev. Biol.
26
,
277
295
.
Struhl
,
G. Barbash
, D. A. and
Lawrence
,
P. A.
(
1997
).
Hedgehog organises the pattern and polarity of epidermal cells in the Drosophila abdomen
.
Development
124
,
2143
54
.
Struhl
,
G.
and
Basler
,
K.
(
1993
).
Organizing activity of Wingless protein in Drosophila
.
Cell
72
,
527
540
.
Tabata
,
T.
,
Eaton
,
S.
and
Kornberg
,
T. B.
(
1992
).
The Drosophila hedgehog gene is expressed in specifically in posterior compartment cells and is a target of engrailed regulation
.
Genes Dev.
6
,
2635
2645
.
Theisen
,
H.
,
Haerry
,
T. E.
,
O’Connor
,
M. B.
and
Marsh
,
J. L.
(
1996
).
Developmental territories created by mutual antagonism between Wingless and Decapentaplegic
.
Development
122
,
3939
3948
.
van den Heuvel
,
M.
,
Nusse
,
R.
,
Johnston
,
P.
and
Lawrence
,
P. A.
(
1989
).
Distribution of the wingless gene product in Drosophila embryos: a protein involved in cell-cell communication
.
Cell
59
,
739
749
.
Zecca
,
M.
,
Basler
,
K.
and
Struhl
,
G.
(
1996
).
Direct and long-range action of a Wingless morphogen gradient
.
Cell
87
,
833
844
.