• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Dev Biol. Author manuscript; available in PMC Nov 15, 2011.
Published in final edited form as:
PMCID: PMC2976676



Imaginal discs of Drosophila have the remarkable ability to regenerate. After fragmentation wound healing occurs, ectopic wg is induced and a blastema is formed. In some, but not all fragments, the blastema will replace missing structures and a few cells can become more plastic and transdetermine to structures of other discs. A series of systematic cuts through the first leg disc revealed that a cut must transect the dorsal-proximal disc area and that the fragment must also include wg-competent cells. Fragments that fail to both transdetermine and regenerate missing structures will do both when provided with exogenous Wg, demonstrating the necessity of Wg in regenerative processes. In intact leg discs ubiquitously expressed low levels of Wg also leads to blastema formation, regeneration and transdetermination. Two days after exogenous wg induction the endogenous gene is activated, leading to elevated levels of Wg in the dorsal aspect of the leg disc. We identified a wg enhancer that regulates ectopic wg expression. Deletion of this enhancer increases transdetermination, but lowers the amount of ectopic Wg. We speculate that this lessens repression of dpp dorsally, and thus creates a permissive condition under which the balance of ectopic Wg and Dpp is favorable for transdetermination.


In all organisms the developmental potential of cells becomes more restricted as they become determined or specified to differentiate specific structures. Despite these developmental restrictions, some cells retain developmental plasticity in order to maintain tissue integrity. But the ability to repair damaged or missing parts varies among animals. The potential to regenerate is also dependent on the developmental stage and even on the site of damage. After limb amputation in lower vertebrates, for example, a new appendage is formed, but removing tissue on the side of the limb only leads to wound closure (Kragl et al., 2008). Thus it is of general interest to better understand what enables or restricts regeneration.

Here we use imaginal discs of Drosophila to study regeneration. Imaginal disc cells, the larval precursors of adult fly structures such as leg and wing, arise from 10 to 30 founder cells in the embryo, in the region where wingless (wg), decapentaplegic (dpp) and engrailed (en) expression overlap (Simcox and Sang, 1983; Cohen, 1993). During larval life, cells divide and become progressively more determined (Schubiger, 1968). Haynie and Bryant (1977) used x-rays to kill 50% of imaginal disc cells in larvae and found that despite the extensive damage, normal flies developed, demonstrating a robust capacity to regenerate. While lost appendages cannot be regenerated in Drosophila, hemimetabolous insects such as Leucophaea, Caraussius and Rhodnius exhibit strong regenerative potential after amputation of nymphal legs (for example, Bohn, 1974). Holometabolous insects with larval legs that contain the cells of the adult primordium can regenerate both the larval and subsequently the adult legs after amputating the appendages in young caterpillars (Bodenstein, 1933 and 1935; Labour, 1966; Lender and Grobocopateli, 1967). Therefore it is not surprising that evolutionarily the internal imaginal discs of Drosophila have maintained regenerative potential.

The epimorphic regeneration of imaginal discs is initiated first by wound healing (Bergantiños et al., 2010; Bosch et al., 2005; Mattila et al., 2005; Reinhardt and Bryant, 1981), followed by blastema formation, blastema growth and patterning (Bryant and Fraser, 1988; Sustar and Schubiger, 2005), in the same manner as observed in zebrafish fin and appendage regeneration in lower amphibian (Brockes and Kumar, 2005; Poss et al., 2003). In contrast to the amphibian limbs, cells in the imaginal disc are not differentiated, though they are rigidly determined, and some cells express tissue specific genes, such as proneural genes, at the time of fragmentation. Thus imaginal disc cells cannot be considered ‘embryonic’.

Our previous genome wide screen (Klebes et al., 2005) identified genes that function and are only expressed during blastema formation in leg discs. Two of them, augmenter of liver regeneration (alr) and regeneration (reg), have homologues that are involved in mammalian liver regeneration (McClure et al., 2008).

Upon fragmentation, regeneration proceeds from the regeneration blastema (Abbott et al., 1981) to restore the missing pattern. However a few disc cells within the blastema become even more plastic, and form structures normally made from different discs, in a process known as transdetermination (TD). TD has been observed with all discs (reviewed in Hadorn, 1978), and occurs at specific sites within the regenerate (Wildermuth, 1968). TD is also known as homeotic regeneration, naturally occurring after loss (amputation) in arthropods. The similarities of regenerative events between Drosophila and vertebrates have been further substantiated by recent reports showing that TD also occurs in mice in embryonic lung tissue (Okubo and Hogan, 2004) and hepatic progenitor cells (Yechoor et al., 2010).

While such sites of TD in imaginal discs have been identified almost by chance, here a series of fragmentations were used to define the properties in the disc that are required for TD. We show that a cut through the dorsal area of the disc is required but not sufficient to induce TD, and that fragments must also contain sufficient cells from the anterior compartment, supporting the notion that sustained wg and dpp expression in the dorsal area of the leg disc induces TD to wing structures (Maves and Schubiger, 1995; Johnston and Schubiger, 1996).

Expressing wg ubiquitously and at low levels in intact discs (Struhl and Basler, 1993) induces regenerative growth and TD (Maves and Schubiger, 1995; Johnston and Schubiger, 1996). We show that low levels of ubiquitous Wg activate the endogenous wg gene in the dorsal disc region, and have identified a regulatory region of the wg gene that responds to the low ubiquitous Wg level. This enhancer is also activated in the blastema after fragmentation. A functional test demonstrates its role in TD and regeneration.


Drosophila stocks

We used the following stocks: wild-type flies were from the Sevelen line. To express ubiquitous wg we crossed y; +; act>y+>wg to y, w, hs-flp flies (Struhl and Basler, 1993) and heat shocked the larvae for one hour in a 37° C water bath. With this regimen wg is activated in over 90% of the cells. We also combined the wg-LacZ reporter construct to the act flip-out system. en-Gal4; UAS-wgts larvae were grown at 25° C (restrictive temperature) to 96 hours after egg deposition (AED). Disc fragments were then injected into adults and shifted to 15° C (permissive temperature) for 18 days (corresponding to 5.1 days at 25° C). After in vivo culture, fragments were isolated and injected into larvae and kept at 25° C through metamorphosis. BRV118 is a 3’ wg enhancer reporter we obtained from S. Carroll. For testing BRV118 function we crossed the act flip-out system into wg1/wg1 animals. Since wg1/wg1 flies, even after selection, show about 45% of flies with no wings, 45% with one wing and 15% with 2 wings, we selected animals from the parental lines y, w, hs-flp; wg1; + and y; wg1; act>y+>wg with either one or no wings for the cross. We received a collection of 20 fly lines, including R25B09, R25C07 and R24G05 from G. Rubin that contain cis-regulatory regions downstream of the wg coding sequence and were constructed by the methods described in Pfeiffer et al. (2008).

Transplantation and in vivo culture

First leg discs from early wandering larvae were fragmented and transplanted into adult hosts as described in Gibson and Schubiger (1999). After specific times of in vivo culture, disc fragments were retrieved, fixed, stained and analyzed, or transplanted into larvae where they metamorphosed with the host. To clear adult cuticle we boiled host flies with implants in 5M KOH for two to five minutes. Isolated implants are composed of multiple vesicles of cuticle. To make a pattern analysis possible, vesicles were separated and spread before mounting in Faure’s water mounting medium (Ashburner, 1989). Leg and wing structures were identified based on earlier descriptions (Schubiger, 1968; Bryant, 1970).

Immunohistochemistry and imaging

Disc material was fixed for 25 minutes in 4% formaldehyde, rinsed with PBS+ 0.1% Tx-100. Before adding primary antibodies, specimens were blocked for at least 15 minutes in 5% normal goat serum (Sigma). We used mouse-anti-Wg (1:200), mouse-anti-En (1:50), and mouse-anti-β-Gal (1:50) (all from the Developmental Studies Hybridoma Bank), as well as mouse-anti-BrdU (1:100) (Becton Dickinson), rabbit-antiβ-Gal (1:1000) (Cappel), rat-anti-Ci (1:2000) (a gift from T. Kornberg), and rabbit-anti-Vg (1:100) (a gift from S. Carroll), and incubated the material overnight at 4° C. After multiple rinses secondary goat antibodies (1:500) (Alexa) were added for 2–3 hours at room temperature. The specimens were then rinsed in PBS and mounted on polylysine-coated coverslips in Fluoromount. For BrdU immunohistochemistry the disc material was labeled with 15 μg/ml BrdU in Ringer’s for 30 minutes, rinsed and fixed for 45 minutes before hydrolysis in 2 N HCl for 45 minutes. The specimens were rinsed multiple times and then processed as above. Confocal images were collected on an MRC600 or BIO-RAD Radiance2000 (except for the images in Fig. 3 which were collected on a Nikon A1R, and in Fig. 5d–d” which were obtained with a Zeiss SP2) and processed with NIH Image-J and Adobe Photoshop. Brightfield images were collected on a Zeiss Axoplan 2 compound microscope with an Axiocam digital camera.

Fig. 3
Regeneration of posterior cells from a 1/4UM fragment after two days in vivo culture
Fig. 5
The wg enhancer BRV118 responds to regeneration signals

Sequencing the wg1 deletion

The wg1 allele was amplified and sequenced using PCR products that spanned the reported approximate deletion (van den Heuvel et al., 1993). Primers used were 5′-GTCGTTTGTCGGTTTTGGTT AND 5′-AATACACAACATCGGCGACA.


Identifying fragments that transdetermine

Experiments using different fragments from various types of discs showed that TD is extremely rare, and occurs only when fragments are serially transplanted into adult hosts to prolong the proliferation period (Hadorn, 1963). However in fragments where a specific region of a first leg disc (Fig. 1) was transected, TD was induced in up to 50% of the cases after a few cell divisions (Schubiger, 1971). These sites of transection were identified by chance and little attention was given to the observation that only one of the two fragments resulting from such a cut was able to transdetermine. Therefore cutting through a specific region is not in itself sufficient for TD. This led us to systematically identify fragments capable of TD.

Fig. 1
First leg disc fragments (grey areas) analyzed after in vivo culture and differentiation to define the areas required for TD

We focus here on TD from first leg to wing because fate maps (Schubiger, 1968; Bryant, 1975), expression and functions of many genes of both appendages are well characterized (Blair, 1995).

After a proliferation period in adult hosts, disc fragments were recovered and transplanted into metamorphosing larval hosts where they differentiated. Analyzing the adult cuticle, we found that transection through the dorsal leg disc led to TD only in specific fragments: the 1/2M, (Fig. 1a’) 1/4UM (Fig. 1b’), the anterior chip minus (Fig. 1c’), the 3/4L (Fig. 1d), and significantly less frequently in an even smaller fragment, the 1/4UM minus piece (Fig. 1e). TD was never observed with the 1/2L (Fig. 1a”), the 1/4UL (Fig. 1b”) and rarely with the posterior chip minus piece (Fig. 1c”). Therefore a cut through the dorsal aspect of the disc is not sufficient to induce TD. On the other hand, a cut through the anterior compartment is also not sufficient for TD (Fig. 1f, g, h). Thus, TD occurs only when cuts transect both the anterior compartment and the dorsal-proximal part of the disc, resulting in fragments containing anterior cells that are wg-responsive to Hh (hedgehog)-signalling (Fig. 1i), as determined by Basler and Struhl (1994).

Transdetermined wing structures were not observed adjacent to all leg structures. Direct contacts were only found between tarsal leg structures and wing (Fig. 2a1), and between dorsal proximal leg and ventral wing base (Fig. 2a2–3). This allowed us to map the site of TD to the anterior-dorsal aspect and to the tarsus anlage of the leg disc. It should be noted however that the transdetermined structures arise from the regenerate and not from the original fragment (Gehring, 1966; Mindek, 1968; Wildermuth, 1968). During regeneration the blastema will regenerate the proximal posterior segments as well as a tarsal duplicate. TD then occurs in the newly formed anterior-dorsal region and the duplicated distal anlage of the leg disc.

Fig. 2
Leg to wing TD

The anterior chip minus fragment is very similar to the 3/4L piece, yet has a much higher rate of TD. In the chip experiment the cut is shallower, cutting through more proximal regions, which stimulates much more growth. This is reflected in the high numbers of complete legs regenerated and the more complete inventory of wing structures differentiated. Recent work has taken advantage of vg expression as a marker for leg to wing TD (e.g., McClure and Schubiger, 2008). Vg is not expressed in the leg disc. Moreover its expression can be recognized in just a few cells in the disc and is thus a much more sensitive marker for TD than the adult cuticle patterns. When we determined TD by Vg expression, we now observed that TD frequency was the same for the two fragments (anterior chip minus, 52%, N=25; 3/4L piece, 53%, N=51). However the area of Vg was significantly larger in the chip minus ones (Mann Whitney rank sum two-tailed test P=0.02, N=28 and 51). We conclude that the smaller areas of TD in 3/4L fragments led to an underestimate of wing structures in the cuticle preparations.

The posterior chip minus test pieces in contrast, transdetermined only in 7% of fragments even though the fragments also showed good growth. But as stated above, a cut must also transect the anterior compartment for the TD process.

In addition to the type of cut, the duration of in vivo culture affected the amount and type of transdetermined structures. In the 1/4UM fragment we found that the area of Vg increased from 8% of the total implant area to 21% (Fig.2b–c), while the rate of TD stayed the same (79% after four days, 87% after 7 days). A similar observation was made from experiments with the 3/4L fragment. Here again, the fraction of the transdetermined area was more than twice as large after five days of in vivo culture than after four (data not shown). We speculate that growth and cell division stop upon completion of the missing structures, but continue on in the area of TD.

Regeneration of the posterior compartment from anterior cells

The enhanced plasticity of the anterior-dorsal cells is also manifested in that they are capable of regenerating a new posterior compartment (Abbott et al., 1981; Gibson and Schubiger, 1999). This means that cells in the blastema must change their compartment identity. Since cells in the anterior compartment express the selector gene cubitus interruptus (ci) and those of the posterior compartment en, we tested whether blastema cells must first lose Ci expression before en is activated. The 1/4UM fragment from the first leg disc contains only Ci-expressing cells in the columnar epithelium, but is overlaid by posterior (En expressing) peripodial cells. These cells transiently fuse with the columnar cells at the onset of wound healing (Reinhardt and Bryant, 1981) and provide a source of Hh, required for regeneration. The peripodial cells however do not participate in the blastema and the regeneration of the posterior compartment (Gibson and Schubiger, 1999). Analyzing the progress of regeneration of 1/4UM fragments we found 25% (n=20) of implants with En expression in columnar cells after one day of in vivo culture. However after two days of culture 70% (n=42) of implants expressed En in columnar cells. In over half of the cases we saw patches of cells expressing both Ci and En, and frequently observed that in such patches Ci was down regulated (Fig. 3). In some patches there was often no or only very little Ci, but higher En levels. We interpret these to be the cells of the regenerated posterior compartment. In no case did we detect cells that lacked both Ci and En, though we cannot exclude a few such cells. This result differs from recently published data (Bergantiños et al., 2010; Smith-Bolton et al., 2009) where, after genetically ablating patches of the wing pouch, cells only regenerated the compartment of their origin. In those experiments, however neither compartment was completely ablated. In such a case the restoration of compartments may occur by compensatory proliferation (Perez-Garijo et al., 2009). Our results indicate that in the regeneration blastema compartment identity becomes labile and allows both Ci and En to be co-expressed as the new posterior compartment is formed and that these cells do not revert to the blastoderm stage when expression of these selector genes is initiated.

A cis-regulatory sequence of wg is activated during regeneration

Previously we showed that ubiquitous wg expression without fragmentation is sufficient to induce a blastema and TD in the dorsal aspect of leg discs (Maves and Schubiger, 1995). Struhl and Basler (1993) reported that 12 hr after the activation of their act>y+>wg construct only low levels of Wg were observed throughout the disc. Here we re-investigated the response to ubiquitous Wg up to three days after induction. We heat shocked y, w, hs-flp; act>y+>wg larvae at 72 hours AED for one hour and fixed leg discs at varying times after wg induction. Six hours after the heat shock the discs showed weak uniform Wg (n=14), confirming Struhl and Basler’s results (Struhl and Basler, 1993). Twenty-four hours after the heat shock the expression of induced Wg was still uniform in most cases (9/12 discs; Fig. 4a), but in 25% of discs Wg was expressed at slightly higher levels dorsally (data not shown). At 47 hours after the heat shock, 73% of the discs (n=14) had faintly higher levels of Wg dorsally (Fig. 4a) and by 67 hours all discs (n=20) clearly had elevated levels of Wg dorsally where overgrowth and TD occurred (Fig. 4a). We refer to the elevated levels of dorsal Wg as ‘ectopic Wg’ to contrast it with the low levels of ubiquitously induced Wg. Frequently patches of dorsal Wg reached levels comparable to the endogenous level on the ventral side of the disc (Fig. 4b). In some cases, a mirror image formed (Fig. 4b), with both a ventral and dorsal wedge-shaped Wg pattern. To rule out the possibility that the ectopic Wg expression was stage dependent we dissected leg discs from wandering larvae that had been heat shocked 6, 24, 48, and 72 hours earlier. We found similar results with higher levels of Wg seen dorsally in all cases after 72 hours. Thus strong dorsal Wg levels depend on the duration of low ubiquitous wg expression.

Fig. 4
Low ubiquitous Wg activates the endogenous wg gene

We then asked if these high levels of ectopic Wg arise from the endogenous wg gene, as was reported after ectopically expressing wg using the dpp-Gal4 driver (Brook and Cohen, 1996). We heat shocked y, w, hs-flp; act>y+>wg larvae (72 hours AED) that also carried the wg-LacZ reporter gene (Struhl and Basler, 1993), and fixed discs 72 hours after the treatment. In Fig. 4c–c’’, the high levels of Wg overlap with high levels of β-Gal, indicating that the initial low levels of ubiquitous wg expression led to activation of the endogenous wg-gene and subsequently to regeneration and TD.

The results above suggest that there must be wg enhancers that respond to low ubiquitous levels of Wg. We tested BRV118-LacZ, a wg enhancer-reporter containing a cis-regulatory sequence located 3’ to the wg transcription start site (Fig. S1). In the wild-type leg disc BRV118 was expressed strongly in a small group of cells in the proximal region of the Wg domain (Fig. 5a–a”). In addition BRV118 was broadly expressed at low levels in much of the disc, including areas outside of the normal Wg domain (Fig. 5a–a”). In the wing disc we observed expression of BRV118 in the prospective wing margin. The expression pattern was broader than the thin stripe of Wg expression in the margin (Fig. 5b–b”). BRV118 expression was also seen along the midline that is not part of the normal Wg domain. When we induced ubiquitous wg expression in leg discs carrying BRV118 we found that in the dorsal areas of overgrowth BRV118 was strongly expressed and largely co-localized with the elevated levels of ectopic Wg (Fig. 5c–c”). Furthermore, Figure 5d–d” shows that BRV118 is also activated in transdetermining cells, as seen by the co-expression of Vg. Thus the BRV118 enhancer was activated directly or indirectly through signaling from ubiquitous Wg expression and suggests it may be involved in regeneration. Thus we tested if BRV118 is also induced in regenerating leg disc fragments. 3/4 L fragments were transplanted into adult hosts for three days and then fixed and labeled for Wg and BRV118 expression. In all fragments (n=36) BRV118 was strongly up regulated and in cases where the morphology allowed, BRV118 was expressed at the wound site, co-localizing with ectopic Wg expression in the regeneration blastema (Fig. 5e-e”).

To test if other wg cis-regulatory regions responded to ubiquitous wg, we screened a collection of 20 transgenic Gal4-reporter lines (Pfeiffer et al., 2008) containing overlapping non-coding regions 24 kb upstream and 35 kb downstream of the wg transcription start site. Only three lines (R25B09, R25C07 and R24G05, which together encompass about 10 kb) responded to ubiquitous wg. R25B09 and R25C07 overlap with, and 24G05 is adjacent to BRV118 (Fig. S1), confirming our results with BRV118.

Testing for BRV118 enhancer function

The original wg allele, wg1 is a small deletion in the region of BRV118 (van den Heuvel, 1993; A. Laughon, personal com). We have sequenced the deletion (Fig. S1) and determined that the 2416 bp deletion is covered by BRV118 except for the first 23 bp. BRV118 extends 567 bp beyond the end point of the deletion. The wg1 mutation leads to the loss of the wing blade and a duplication of the notum but only to minor defects in the leg that do not affect eclosion to adults (Held, 1993; Sharma and Chopra, 1976; G. Schubiger, unpublished observations). This mutation allowed us to test for the function of the BRV118 regulatory sequence by comparing ectopic Wg expression and rates of TD in wild-type and wg1/wg1 leg discs after ubiquitous wg expression. Larvae were heat shocked 60 hours AED and their first leg discs fixed 4 days later. We scored the discs for ectopic dorsal Wg expression, and classified them as having low to moderate or high levels of wg (see Fig. 4a, 67 hr, respectively Fig. 4c, disc on the right). In wg1/wg1 discs ectopic Wg was strongly reduced (Table 1), though not eliminated. This indicates that other regulatory elements must also respond to ubiquitous Wg, albeit less efficiently. We also found no case with a mirror image of the Wg domain in the dorsal aspect of the disc (Fig. 4b), in contrast to wild-type discs that showed this pattern in 20% of the cases. Surprising was our observation that wg1/wg1 discs were more likely to transdetermine (based on Vg expression). This result was unexpected since the wg1 mutation leads to the loss of wings. We will discuss this result later.

Table 1
Deleting the BRV118 enhancer alters amount of ectopic Wg and TD rate

Wg is sufficient to induce regeneration and TD in a non-regenerating and transdetermining fragment

In Figure 1 we identified fragments that failed to transdetermine. Here we asked if supplying Wg enables TD in such fragments. We tested the 1/2L fragment (Fig. 1a”) that will not only fail to transdetermine but is also unable to regenerate missing structures (Schubiger, 1971). First, in a set of control experiments we fragmented wild-type discs into 1/2M and 1/2L pieces (Fig. 6) and injected them into adult hosts for three days, followed by a pulse of BrdU to label cells in S-phase. Both pieces reached approximately the same size after in vivo culture, reflecting regeneration of the 1/2M piece and duplication in the 1/2L one. Both fragments showed areas of DNA replication (Fig. 6b), but in the 1/2M piece (n=12) the BrdU label extended to the proximal primordia of the disc and in addition, often to one side of the healing wound. In the 1/2L pieces (n=15) the label was concentrated in the distal primordia, and on both sides of the scar. Since the 1/2L piece does not contain wg-expressing cells (Fig. 6a) we asked if expressing wg ubiquitously in these fragments changes their response. Indeed activating act>y+>wg in fragments immediately after injection into the adult hosts led to increased proliferation in both 1/2M and 1/2L fragments (n=26, n=33, respectively). After three days of in vivo culture implants had grown to about double the size of halves without ubiquitous wg expression (Fig. 6c). In the 1/2L fragment BrdU was now found to extend to the periphery in a similar pattern to that of the 1/2M piece.

Fig. 6
Role of Wg in regeneration

To verify that these fragments can regenerate and transdetermine, we checked for the adult structures formed from 1/2L and 1/2M fragments after the induction of ubiquitous wg. After eight days of in vivo culture we recovered ten 1/2M and fourteen 1/2L pieces. The pieces had grown more than seen in any other experiment of this kind, and had to be cut into two to eight pieces for the 1/2 M and one to twelve for the 1/2 L fragment before injecting them into larval hosts. Because not all larval hosts survived, we only rarely got the complete inventory of all adult structures differentiated from one original disc fragment. Nevertheless for the 1/2 M fragment we found a very high rate of TD (six of the nine original 1/2M fragments). The normally non-regenerating and non-transdetermining 1/2L fragment now regenerated medial leg structures in 13 of 14 original fragments and transdetermined in 10 implants. Areas of unpigmented trichomes and shortened bristles were observed in seven cases, suggesting that the final steps of terminal differentiation had failed. In three of the seven cases we identified adult muscle. Since muscle arises from the adepthelial cells associated with the disc (Poodry and Schneiderman, 1970), it is unlikely that the muscle reflects TD. Remarkable was the variety of wing structures formed. In one case, in addition to wing margin and wing base, we also observed notum, a transdetermined structure rarely seen.

These experiments show that wg can induce regeneration and TD in fragments that normally lack this capacity. Almost all cells of the 1/2L fragment are in the posterior compartment, and do not express wg. Therefore we specifically tested whether posterior cells can be induced to regenerate and transdetermine by using en-Gal4 to drive a temperature sensitive allele of wg (UAS-wgts, see Material and Methods). Fragment growth after in vivo culture was not as extensive as was observed in the above experiment, and we could almost always implant the original piece into a single larval host. Even though the 1/2L pieces grew less than in the previous experiment, they regenerated and transdetermined with a low frequency (two cases of thirteen implants). It is likely that the wgts allele is unable to provide sufficient functional Wg (Johnston and Schubiger, 1996), and that this is the reason for the weak response. Nevertheless, these results clearly show that Wg can induce posterior cells to regenerate anterior cells and transdetermine, and demonstrates the significance of Wg in regeneration and TD.


Conditions for TD

The systematic analysis of a series of different cuts presented here has allowed us to conclude that for TD to occur a cut must go through the dorsal region of the disc and the fragment must also contain wg-responsive cells (Fig.1). After fragmentation the peripodial epithelium transiently fuses with the columnar epithelium (Reinhardt and Bryant, 1981). In the first leg disc the dorsal peripodial cells express hh, and will signal to the anterior columnar cells during initial wound healing (Gibson and Schubiger, 1999) to activate wg-responsive cells. In the posterior compartment in contrast, wg is not activated and thus this fragment will not transdetermine (Fig. 1b”).

Direct contacts between leg and wing have been mapped to the dorsal anterior aspect of the leg disc. This is the same region that transdetermines after ubiquitous wg expression (Fig. 5d, Maves and Schubiger, 1998). Recently, Salzer and Kumar (2010) showed that only cells in the dorsal region of the leg disc could be induced to form eyes after ectopically expressing retina-determining genes. Thus the cells in this region appear to have greater developmental plasticity, and the region has been called the ‘weak point’. This region of higher plasticity is overlaid by the hh expressing peripodial cells. As mentioned above a cut through hh positive peripodial cells is required for TD and experimentally defined the region capable of inducing TD. Yet after fragmentation transdetermined structures arise within the newly regenerated epithelium and not from dorsal cells in the original piece. This has led to some confusion about the term ‘weak point’. Why do the cells in the original fragment fail to transdetermine but can be induced to do so in the intact disc when wg or, for example, some selector genes are ectopically expressed? One reason may be that the cells at the dorsal cut rarely contribute to the regenerate (Abbott et al., 1981), indicating that they are not participating in the proliferating blastema. The induction of proliferation however is a prerequisite for regeneration and TD (Schubiger, 1973; Schweizer and Bodenstein, 1975). In addition WG and Dpp must interact for TD to occur (Maves and Schubiger, 1998), but at the dorsal cut wg is not expressed. In the blastema we have shown that ectopic Wg is expressed (McClure et al., 2008) and propose here that as new dorsal leg disc cells (i.e, dpp expressing cells) are regenerated, Wg and Dpp can interact in the blastema to allow TD.

WG, regeneration and transdetermination

Lost or damaged tissues are repaired by different mechanisms depending on the organism. But in many cases Wg or Wnt has an essential function. During morphallactic regeneration in Hydra, Wnt expression correlates with cell movements and head regeneration (Philipp et al., 2009). In planaria where regeneration depends on stem cells, blocking Wnt-signalling has no effect on blastema formation, but drastically changes the type of structures formed. In this case Wnt functions to pattern the regeneration blastema (Gurley et al., 2008; Petersen and Reddien, 2008).

Canonical Wg signaling is one of the early signals necessary and sufficient in epimorphic limb or fin regeneration (Kawakami et al., 2006, Stoick-Cooper et al., 2007). Similarly we observed that one of the first changes in disc fragments is the ectopic expression of Wg near the wound area preceding blastema formation (Gibson and Schubiger, 1999; McClure et al., 2008), recently confirmed by Smith-Bolton et al. (2009). The critical role of Wg signaling was shown in amputated limbs of chick embryos where wound healing but not regeneration occurs. However when activated β-catenin is induced in the wound cap, an apical ectodermal ridge is formed and regeneration now can proceed (Kawakami et al., 2006).

Here we demonstrated that the absence of Wg in the 1/2L fragment is a major reason for its failure to regenerate, and as in chick limb regeneration (Kawakami et al., 2006), induced Wg-signaling allows regeneration. Moreover, Wg- signaling led to TD, indicating that cells became more plastic.

Struhl and Basler (1993) reported duplicated leg structures after inducing ectopic wg expression. Twelve hours after they induced expression, Wg was observed at low levels throughout the disc, but surprisingly all defects in the adult leg were dorsal. Here we showed that after 2–3 days Wg becomes highly expressed in the dorsal leg disc where outgrowth occurs (Fig. 4a). We went on to show that the initial low levels of ubiquitous Wg activated the endogenous wg-promoter in this part of the leg disc (Fig. 5c). Thus uniform wg expression in the disc can activate the endogenous gene in specific regions. We do not know at this point if Wg signaling directly activates wg, as has been reported in the embryo (Hooper, 2002), or indirectly, but we were able to show that the wg-enhancer BRV118 reacts to the induced low levels of Wg. The BRV118 enhancer is also activated in the blastema after fragmentation, demonstrating that it responds to regeneration signals in general, and suggests that BRV118 regulates wg expression during regeneration.

The BRV118 enhancer is largely deleted in the wg1 allele. The wg1 mutation is not 100% penetrant, and although the majority of flies are missing one or both wings, some animals do have two perfect wings. Thus other enhancer regions must be able to promote wing blade formation at a low frequency. When we tested TD frequency in wg1/wg1 leg discs after ubiquitous wg expression we observed a significantly higher rate of TD compared to ubiquitous wg expression in a wild-type background (Table 1). At first glance this might seem unexpected since wg1/wg1 animals are characterized by the loss of wings. Previous work showed that varying amounts of ectopic Wg controlled the rate of leg to wing TD. Moderate ectopic Wg led to high TD rates, whereas high Wg signaling down-regulated dpp expression and led to ventralization, loss of dorsal structures and low TD rate (Johnston and Schubiger, 1996). Our results show that wg1 reduces ectopic Wg levels in the leg disc, most dramatically evident by the absence of discs with a mirror image Wg pattern. We propose that in the absence of the BRV118 regulatory sequence ectopic Wg reaches a threshold permissive for TD but too low for extensive loss of dorsal structures, structures from which TD will occur.

Regeneration of the posterior compartment by anterior cells

Cell clones induced in regenerating anterior leg disc fragments are composed of both En and non-En expressing cells (Gibson and Schubiger, 1999), thus clearly breaking their compartment identity. Once some posterior cells are formed the compartmental boundary is reestablished even before all missing pattern elements of the posterior compartment are reformed (Abbott et al., 1981). Here we have shown that during the early stages of regeneration of the 1/4UM fragment, cells co-expressed Ci and En. This indicates that cells in the blastema do not revert to an embryonic stage before compartment identity is established. Since anterior cells must be reprogrammed to make posterior cells we speculate that a novel pathway is activated as a consequence of wound healing and activation of the JNK-pathway that reduces polycomb function (Lee et al., 2005). Polycomb group (PcG) genes are required to maintain the anterior/posterior compartment boundary (Maschat et al., 1998). In the absence of the PcG gene ph, mutant clones in the anterior compartment will express en (Randsholt et al., 2000). We propose that during regeneration decreased PcG function destabilizes anteriorness and allows the expression of en. Once en is activated it will inhibit Ci expression leading to en only expression (Tabata et al., 1995). Since the anterior/posterior compartment is reestablished during regeneration, as mentioned above, PcG function is normalized again, and the new anterior/posterior compartment is maintained. Recently Gettings et al., (2010) reported that during normal dorsal closure an anterior cell at the leading edge of the ectoderm is reprogrammed and moves to the posterior compartment. The authors show that the JNK pathway is required for the change. During regeneration of the posterior compartment in disc fragments the JNK-pathway is also involved (see above), opening the possibility that in both processes the switch from anterior to posterior may be similarly regulated.

Our earlier work also supports the idea that new developmental programs are initiated during regeneration. For TD from leg to wing to occur, we showed that the vg boundary enhancer is essential, and not the quadrant enhancer which is required earlier in wing development, namely during the second instar (Maves and Schubiger, 1995). The clearest case for following a novel program during regeneration comes from transdetermining leg disc cells that transiently adopt a novel cell-cycle profile, different from the profile of younger disc cells (Sustar and Schubiger, 2005).

In summary these results indicate that successful regeneration does not require cells to re-capitulate embryonic programs, but is more likely to involve programs not normally used during disc development. In amphibians it was recently shown that in the regenerating axolotl limbs progenitor cells in the blastema do not dedifferentiate and reenter an embryonic state. Elegant labeling studies revealed that dedifferention leads to progenitor cells with restricted potential rather than to pluripotency (Kragl et al., 2009). It will be interesting to see if such cells also embark on a novel program during regeneration.

Supplementary Material


Fig. S1. Genomic map of the wg region:

The wg gene (open arrow, TSS 2L: 7307161, r5.28) and the 3’cis-regulatory sequences BRV118 (black box), R25B09, R25C07, and R24G05 (grey boxes) are indicated. The wg1 mutation is a 2416 deletion (first deleted base pair 7324252).


We would like to thank L. Buttitta, G. Struhl and the Bloomington Stock Center for flies, and to S. Carroll, T. Kornberg and the Developmental Studies Hybridoma Bank for antibodies. We are particularly grateful to S. Carroll for the BRV118 line, and to G. Rubin for the wg enhancer reporter lines prior to publication. We thank M. Markstein for discussions, and appreciate the comments on the manuscript from M. Bonvin, M. Gibson, R. Paro and T. Kornberg. We would like to recognize D. Erezyilmaz for her contribution during the early phase of this work. This work was supported by the National Institutes of Health NIH (GM058282) to G. S.


Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.


  • Abbott L, Karpen G, Schubiger G. Compartmental restrictions and blastema formation during pattern regulation in Drosophila imaginal leg discs. Dev Biol. 1981;87:64–75. [PubMed]
  • Ashburner M. Drosophila: A Laboratory Handbook and Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 1989.
  • Basler K, Struhl G. Compartment boundaries and the control of Drosophila limb pattern by hedgehog protein. Nature. 1994;368:208–214. [PubMed]
  • Bergantiños C, Corominas M, Serras F. Cell death-induced regeneration in wing imaginal discs requires JNK signalling. Development. 2010;137:1169–1179. [PubMed]
  • Blair S. Compartments and appendage development in Drosophila. Bioessays. 1995;17:299–309. [PubMed]
  • Bodenstein D. Beintransplantationen an Lepidopterenraupen. II. Zur Analyse der Regeneration der Brustbeine von Vanessa urticae-Raupen. Wilhelm Roux’ Arch EntwicklMech Org. 1933;130:747–770.
  • Bodenstein D. Beintransplantationen an Lepidopterenraupen. III. Zur Analyse der Entwickluugspotenzen der Schmetterlingsbein. Arch EntwMech Org. 1935;133:156–192.
  • Bohn H. Extent and properties of the regeneration field in the larval legs of cockroaches (Leucophaea maderae). I. Extirpation experiments. J Embryol Exp Morphol. 1974;31:557–572. [PubMed]
  • Bosch M, Serras F, Martín-Blanco E, Baguñà J. JNK signaling pathway required for wound healing in regenerating Drosophila wing imaginal discs. Dev Biol. 2005;280:73–86. [PubMed]
  • Brook WJ, Cohen SM. Antagonistic interactions between Wingless and Decapentaplegic responsible for dorsal-ventral pattern in the Drosophila leg. Science. 1996;273:1373–1377. [PubMed]
  • Brockes J, Kumar A. Appendage regeneration in adult vertebrates and implications for regenerative medicine. Science. 2005;310:1919–1923. [PubMed]
  • Bryant PJ. Cell lineage relationships in the imaginal wing disc of Drosophila melanogaster. Dev Biol. 1970;22:389–411. [PubMed]
  • Bryant PJ. Pattern formation in the imaginal wing disc of Drosophila melanogaster: fate map, regeneration and duplication. J Exp Zool. 1975;193:49–77. [PubMed]
  • Bryant PJ, Fraser SE. Wound healing, cell communication, and DNA synthesis during imaginal disc regeneration in Drosophila. Dev Biol. 1988;127:197–208. [PubMed]
  • Cohen S. Imaginal disc development. In: Bate M, Martinez Arias A, editors. The Development of Drosophila melanogaster. Vol. 2. Plainview, NY: Cold Spring Harbor Laboratory Press; 1993.
  • Gehring W. Übertragung und Änderung der Determinationsqualitäten in Antennenscheiben-Kulturen von Drosophila melanogaster. J Embryol Exp Morphol. 1966;15:77–111. [PubMed]
  • Gettings M, Serman F, Rousset R, Bagnerini P, Almeida L, Noselli S. JNK signalling controls remodelling of the segment boundary through cell reprogramming during Drosophila morphogenesis. PLoS Biol. 2010;8:e1000390. [PMC free article] [PubMed]
  • Gibson M, Schubiger G. Hedgehog is required for activation of engrailed during regeneration of fragmented Drosophila imaginal discs. Development. 1999;126:1591–1599. [PubMed]
  • Gurley K, Rink J, Sánchez Alvarado A. Beta-catenin defines head versus tail identity during planarian regeneration and homeostasis. Science. 2008;319:323–327. [PMC free article] [PubMed]
  • Hadorn E. Problems of determination and transdetermination. Brookhaven Symp Biol. 1965;18:148–161.
  • Hadorn E. Imaginal discs: transdetermination. In: Ashburner M, Wright T, editors. The Genetics and Biology of Drosophila. New York: Academic Press; 1978.
  • Haynie J, Bryant P. The effects of X-rays on the proliferation dynamics of cells in the imaginal wing disc of Drosophila melanogaster. Roux's Arch Dev Biol. 1977;183:85–100.
  • Held LI., Jr Segment-polarity mutations cause stripes of defects along a leg segment in Drosophila. Dev Biol. 1993;157:240–250. [PubMed]
  • Hooper J. Distinct pathways for autocrine and paracrine Wingless signalling in Drosophila embryos. Nature. 1994;372:461–464. [PubMed]
  • Johnston L, Schubiger G. Ectopic expression of wingless in imaginal discs interferes with decapentaplegic expression and alters cell determination. Development. 1996;122:3519–3529. [PubMed]
  • Kawakami Y, Rodriguez Esteban C, Raya M, Kawakami H, Martí M, Dubova I, Izpisúa Belmonte JC. Wnt/beta-catenin signaling regulates vertebrate limb regeneration. Genes Dev. 2006;20:3232–3237. [PMC free article] [PubMed]
  • Klebes A, Sustar A, Kechris K, Li H, Schubiger G, Kornberg T. Regulation of cellular plasticity in Drosophila imaginal disc cells by the Polycomb group, trithorax group and lama genes. Development. 2005;132:3753–3765. [PubMed]
  • Kragl M, Knapp D, Nacu E, Khattak S, Schnapp E, Epperlein H, Tanaka E. Novel Insights into the Flexibility of Cell and Positional Identity during Urodele Limb Regeneration. Cold Spring Harb Symp Quant Biol. 2008;73:583–592. [PubMed]
  • Kragl M, Knapp D, Nacu E, Khattak S, Maden M, Epperlein HH, Tanaka EM. Cells keep a memory of their tissue origin during axolotl limb regeneration. Nature. 2009;460:60–65. [PubMed]
  • Labour G. Étude de la régénération des pattes chez le Doryphore: Leptinotarsa decemlineata, au cours de son développement larvaire. Bull Soc Zool Fr. 1966;91:687–696.
  • Lender T, Grobocopateli A. Étude du territoire de régénération de la patte larvaire de Tenebrio molitor (Coléoptère) Bull Soc Zool Fr. 1967;92:213–222.
  • Lee N, Maurange C, Ringrose L, Paro R. Suppression of Polycomb group proteins by JNK signalling induces transdetermination in Drosophila imaginal discs. Nature. 2005;438:234–237. [PubMed]
  • Maschat F, Serrano N, Randsholt N, Géraud G. Engrailed and polyhomeotic interactions are required to maintain the A/P boundary of the Drosophila developing wing. Development. 1998;125:2771–2780. [PubMed]
  • Mattila J, Omelyanchuk L, Kyttälä S, Turunen H, Nokkala S. Role of Jun N-terminal Kinase (JNK) signaling in the wound healing and regeneration of a Drosophila melanogaster wing imaginal disc. Int J Dev Biol. 2005;49:391–9. [PubMed]
  • Maves L, Schubiger G. Wingless induces transdetermination in developing Drosophila imaginal discs. Development. 1995;121:1263–1272. [PubMed]
  • McClure K, Sustar A, Schubiger G. Three genes control the timing, the site and the size of blastema formation in Drosophila. Dev Biol. 2008;319:68–77. [PMC free article] [PubMed]
  • Mindek G. Proliferations- und Transdeterminationsleistungen der weiblichen Gemital-Imaginalscheiben von Drosophila melanogaster. Wilhelm Roux’ Arch EntMech Org. 1968;161:249–280.
  • Okubo T, Hogan BLM. Hyperactive Wnt signaling changes the developmental potential of embryonic lung endoderm. J Biol. 2004;3:11. [PMC free article] [PubMed]
  • Pérez-Garijo A, Shlevkov E, Morata G. The role of Dpp and Wg in compensatory proliferation and in the formation of hyperplastic overgrowths caused by apoptotic cells in the Drosophila wing disc. Development. 2009;136:1169–77. [PubMed]
  • Petersen C, Reddien P. Smed-betacatenin-1 is required for anteroposterior blastema polarity in planarian regeneration. Science. 2008;319:327–330. [PubMed]
  • Pfeiffer BD, Jenett A, Hammonds AS, Ngo TT, Misra S, Murphy C, Scully A, Carlson JW, Wan KH, Laverty TR, Mungall C, Svirskas R, Kadonaga JT, Doe CQ, Eisen MB, Celniker SE, Rubin GM. Tools for neuroanatomy and neurogenetics in Drosophila. Proc Natl Acad Sci USA. 2008;105:9715–9720. [PMC free article] [PubMed]
  • Philipp I, Aufschnaiter R, Ozbek S, Pontasch S, Jenewein M, Watanabe H, Rentzsch F, Holstein T, Hobmayer B. Wnt/beta-catenin and noncanonical Wnt signaling interact in tissue evagination in the simple eumetazoan Hydra. Proc Natl Acad Sci USA. 2009;106:4290–4295. [PMC free article] [PubMed]
  • Poodry CA, Schneiderman HA. The ultrastructure of the developing leg of Drosophila melanogaster. Wilhelm Roux’ Arch. 1970;166:1–44.
  • Poss K, Keating M, Nechiporuk A. Tales of regeneration in zebrafish. Dev Dyn. 2003;226:202–210. [PubMed]
  • Randsholt NB, Maschat F, Santamaria P. Polyhomeotic controls engrailed expression and the hedgehog signaling pathway in imaginal discs. Mech Dev. 2000;95:89–99. [PubMed]
  • Reinhardt C, Bryant P. Wound healing in the imaginal discs of Drosophila. II. Transmission electron microscopy of normal and healing wing discs. J Exp Zool. 1981;216:45–61. [PubMed]
  • Salzer CL, Kumar JP. Identification of retinal transformation hot spots in developing Drosophila epithelia. PLoS One. 2010;5:e8510. [PMC free article] [PubMed]
  • Schubiger G. Anlageplan, Determinationszustand und Transdeterminationsleistungen der männlichen Vorderbeinscheibe von Drosophila melanogaster. Wilhelm Roux’ Arch EntwMech Org. 1968;160:9–40.
  • Schubiger G. Regeneration, duplication and transdetermination in fragments of the leg disc of Drosophila melanogaster. Dev Biol. 1971;26:277–295. [PubMed]
  • Schubiger G. Regeneration of Drosophila melanogaster male leg disc fragments in sugar fed female hosts. Experientia. 1973;29:631–632. [PubMed]
  • Schweizer P, Bodenstein D. Aging and its relation to cell growth and differentiation in Drosophila imaginal discs: developmental response to growth restricting conditions. Proc Natl Acad Sci USA. 1975;72:4674–4678. [PMC free article] [PubMed]
  • Simcox A, Sang J. When does determination occur in in Drosophila embryos? Dev Biol. 1983;97:212–221. [PubMed]
  • Sharma RP, Chopra VL. Effect of the Wingless (wg1) mutation on wing and haltere development in Drosophila melanogaster. Dev Biol. 1976;48:461–4655. [PubMed]
  • Smith-Bolton R, Worley M, Kanda H, Hariharan I. Regenerative growth in Drosophila imaginal discs is regulated by Wingless and Myc. Dev Cell. 2009;16:797–809. [PMC free article] [PubMed]
  • Stoick-Cooper C, Weidinger G, Riehle K, Hubbert C, Major M, Fausto N, Moon R. Distinct Wnt signaling pathways have opposing roles in appendage regeneration. Development. 2007;134:479–489. [PubMed]
  • Struhl G, Basler K. Organizing activity of wingless protein in Drosophila. Cell. 1993;72:527–540. [PubMed]
  • Sustar A, Schubiger G. A transient cell cycle shift in Drosophila imaginal disc cells precedes multipotency. Cell. 2005;120:383–393. [PubMed]
  • Tabata T, Schwartz C, Gustavson E, Ali Z, Kornberg T. Creating a Drosophila wing de novo, the role of engrailed, and the compartment border hypothesis. Development. 1995;121:3359–3369. [PubMed]
  • Van den Heuvel M, Harryman-Samos C, Klingensmith J, Perrimon N, Nusse R. Mutations in the segment polarity genes wingless and porcupine impair secretion of the wingless protein. EMBO J. 1993;12:5293–5302. [PMC free article] [PubMed]
  • Wildermuth H. Differenzierungsleistungen, Mustergliederung und Transdeterminationsmechanismen in hetero- und homoplastischen Transplantaten der Rüsselprimordien von Drosophila. Wilhelm Roux’ Arch EntwMech Org. 1968;160:41–75.
  • Yechoor V, Liu V, Espiritu C, Paul A, Oka K, Kojima H, Chan L. Neurogenin3 is sufficient for transdetermination of hepatic progenitor cells into neo-islets in vivo but not transdifferentiation of hepatocytes. Dev Cell. 2010;16:358–373. [PMC free article] [PubMed]
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...