Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Dev Cell. Author manuscript; available in PMC 2009 March 1.
Published in final edited form as:
PMCID: PMC2277325

Distinct Mechanisms of Apoptosis-induced Compensatory Proliferation in Proliferating and Differentiating Tissues in the Drosophila Eye


In multi-cellular organisms, apoptotic cells induce compensatory proliferation of neighboring cells to maintain tissue homeostasis. In the Drosophila wing imaginal disc dying cells trigger compensatory proliferation through secretion of the mitogens Decapentaplegic (Dpp) and Wingless (Wg). This process is under control of the initiator caspase Dronc, but not effector caspases. Here, we show that a second mechanism of apoptosis-induced compensatory proliferation exists. This mechanism is dependent on effector caspases which trigger the activation of Hedgehog (Hh) signaling for compensatory proliferation. Furthermore, while Dpp and Wg signaling is preferentially employed in apoptotic proliferating tissues, Hh signaling is activated in differentiating eye tissues. Interestingly, effector caspases in photoreceptor neurons stimulate Hh signaling which triggers cell cycle re-entry of cells that had previously exited the cell cycle. In summary, dependent on the developmental potential of the affected tissue, different caspases trigger distinct forms of compensatory proliferation in an apparent non-apoptotic function.


Programmed cell death or apoptosis is a genetically controlled process with important roles during development of multicellular organisms. Molecular genetic studies have revealed that the basic principles of regulation and execution of apoptosis are conserved (Cashio et al., 2005). One common feature in the cell death program is the activation of caspases, a highly specialized class of cell death proteases. These enzymes are generally divided into two distinct classes: initiator and effector caspases. Upon activation, initiator caspases activate the downstream effector caspases via proteolytic processing, and activated effector caspases cleave key cellular substrates to promote apoptosis. In Drosophila, the pro-apoptotic genes head involution defective (hid), reaper and grim are both necessary and sufficient for the induction of apoptosis through activation of the initiator caspase Dronc and the effector caspases DrICE and Dcp-1 (Cashio et al., 2005). In addition to an essential role of caspases for apoptosis, recent findings have demonstrated that caspases also have important functions in non-apoptotic processes (Kuranaga and Miura, 2007; Lamkanfi et al., 2007). One such non-apoptotic process is the induction of compensatory proliferation in apoptotic tissue.

Coordination of cell death and cell proliferation is critical for the maintenance of tissue homeostasis. Excessive cell loss in a developing tissue can be compensated for by additional divisions of the remaining cells. For example, in the Drosophila wing imaginal disc irradiation-induced cell death is followed by compensatory cell proliferation which results in an adult wing of nearly normal size (Haynie and Bryant, 1977; James and Bryant, 1981). Further genetic manipulation using toxins to induce ectopic cell death in wing discs showed that the cells adjacent to the apoptotic cells undergo compensatory proliferation (Milan et al., 1997). This suggests that apoptotic cells induce compensatory proliferation of neighboring cells through secretion of mitogens.

More recently, in developing wing discs, in which apoptosis was induced by expression of the pro-apoptotic gene hid, loss of the caspase inhibitor DIAP1 or by X-ray treatment, the accumulation of two major mitogens, Decapentaplegic (Dpp) and Wingless (Wg), has been observed in dying cells (Huh et al., 2004; Perez-Garijo et al., 2004; Ryoo et al., 2004). Key for this finding was the simultaneous expression of the caspase inhibitor P35 (see also Figure 2A). Under these conditions, the dying cells were kept alive (‘undead’), allowing accumulation of Dpp and Wg. This accumulation appears to be dependent on the initiator caspase Dronc, because it cannot be blocked by expression of P35 which inhibits effector caspases, but not Dronc (Hawkins et al., 2000; Meier et al., 2000; Yoo et al., 2002; Yu et al., 2002). In addition, the Drosophila homolog of the tumor suppressor p53, Dp53, has been implicated downstream of Dronc for compensatory proliferation (Wells et al., 2006) (see also Figure 2A).

Figure 2
Effector caspases DrICE and Dcp-1 are required for GMR-hid-induced compensatory proliferation

Notably, these studies on mechanisms of compensatory proliferation were carried out in developing larval wing imaginal discs in Drosophila. Cells in wing discs proliferate extensively during larval stages, and the majority of these cells do not differentiate before they reach pupal development (Garcia-Bellido and Merriam, 1971; Milan et al., 1996). Hence, the mechanisms of compensatory proliferation have so far only been investigated in situations where most cells are proliferating. Interestingly, we have recently observed apoptosis-induced compensatory proliferation in differentiating eye tissue of third instar larvae (Srivastava et al., 2007). However, it is unclear whether this form of compensatory proliferation is controlled by a similar mechanism as reported for larval proliferating wing discs.

Unlike wing discs, cellular differentiation in eye discs starts during mid-third instar larval stage. An indentation known as the morphogenetic furrow (MF) forms at the posterior edge of the eye disc and sweeps across to the anterior of the eye disc. Anterior to the MF, cells are in an uncommitted state and keep proliferating. However, cells immediately before and in the MF arrest in the G1 phase of the cell cycle for an extended period. A subset of these G1-arrested cells starts differentiating into 5 of the final 8 photoreceptor neurons (R8, R2, R5, R3, R4) per ommatidium, while the remaining cells synchronously re-enter S phase and form the second mitotic wave (SMW) posterior to the MF (Wolff and Ready, 1993) (see also Figure 3A). After completion of mitosis in the SMW, all cells arrest in G1 indefinitely (Baker and Yu, 2001). From this pool of undifferentiated cells generated in the SMW, the remaining photoreceptors (R1, R6, R7), non-neuronal cone cells, and pigment cells are recruited into ommatidia (Wolff and Ready, 1993).

Figure 3
GMR-hid-induced compensatory proliferation occurs in undifferentiated cells and requires a short-range signal

Here, by taking advantage of the different developmental states of the third instar larval eye disc, we show that different mechanisms control apoptosis-induced compensatory proliferation in proliferating and differentiating eye tissue. As shown in the proliferating wing disc, Dpp and Wg are expressed in response to apoptotic activity in proliferating eye tissue. In contrast, compensatory proliferation in differentiating eye tissue requires Hh signaling. We also show that the effector caspases DrICE and Dcp-1 are required for compensatory proliferation in differentiating tissue, in contrast to proliferating tissue which requires the initiator caspase Dronc. Thus, depending on the developmental context, different caspases trigger distinct forms of compensatory proliferation in an apparently non-apoptotic function.


Apoptosis-induced compensatory proliferation in GMR-hid eye discs

Developing Drosophila larval eye discs are composed of the anterior proliferating tissue and the posterior differentiating tissue which are divided by the MF and the SMW (Figure 1A; see also Figure 3A) (Wolff and Ready, 1993). Notably, in wild type eye discs posterior to the SMW, only a few scattered cells label positive for the S-phase marker BrdU or the mitotic marker Phosphorylated Histone 3 (PH3) suggesting that most cells have exited the cell cycle (Figures 1A,A′,B). Similarly, only a few apoptotic cells as revealed by anti-cleaved, i.e. active, caspase-3 labeling (Cas3*), are scattered throughout the disc (Figure 1A).

Figure 1
GMR-hid induces a wave of compensatory proliferation

Expression of the pro-apoptotic gene hid under control of the eye-specific GMR enhancer (GMR-hid) (Grether et al., 1995) posterior to the MF causes strong caspase activity in two distinct waves posterior to the MF (Figure 1C). In the first apoptotic wave, cells in the SMW are dying, whereas photoreceptor neurons are differentiating normally and die in the second wave (data not shown). Surprisingly, compared to wild-type eye discs (Figure 1A′), a wave of compensatory proliferation forms posterior to the SMW in between the two apoptotic waves in GMR-hid discs as revealed by BrdU labelings (Figure 1C′). Moreover, cells in the wave of compensatory proliferation are also mitotically active as revealed by PH3 labelings (Figure 1D). This observation is remarkable because under normal conditions cells posterior to the SMW are indefinitely cell cycle-arrested (Baker and Yu, 2001), and it implies that at least some cells re-enter the cell cycle for GMR-hid-induced compensatory proliferation. We also analyzed whether cell death and compensatory proliferation are induced by GMR-hid at early stages following the movement of the MF. Although only one wave of cell death is observed around the SMW in mid-3rd instar eye discs, the wave of compensatory proliferation is detectable as soon as about six ommatidial columns have formed (Figure 1E). Thus, induction of hid posterior to the MF induces a wave of apoptosis-induced compensatory proliferation. Similarly, compensatory proliferation was also observed in GMR-reaper eye discs (data not shown).

Effector caspases are required for compensatory proliferation in GMR-hid

Previous studies in the wing imaginal discs have implicated the initiator caspase Dronc and Dp53, but not effector caspases, for apoptosis-induced compensatory proliferation (Figure 2A) (Huh et al., 2004; Kondo et al., 2006; Wells et al., 2006). Therefore, we tested whether similar requirements exist for GMR-hid-induced compensatory proliferation. Overexpression of DIAP1 or genetic inactivation of dronc results in loss of the wave of compensatory proliferation in GMR-hid eye discs, as expected (Figure 2B,C). Strikingly, in contrast to wing discs, GMR-hid-induced compensatory proliferation is blocked by simultaneous expression of the caspase inhibitor P35 (Figure 2D). Interestingly also, although Dp53 was found to be essential for apoptosis-induced compensatory proliferation in the wing (Figure 2A) (Wells et al., 2006), GMR-hid-induced compensatory proliferation was unaffected in dp53 mutants (Figure 2E) suggesting that dp53 is not required for compensatory proliferation in GMR-hid eye discs. The latter two findings imply that distinct mechanisms of compensatory proliferation exist in apoptotic eye and wing imaginal discs.

Because P35 blocks cell death through its ability to inactivate effector caspases in Drosophila, the role of the effector caspases Drosophila ICE (DrICE) and Dcp-1 for GMR-hid-induced compensatory proliferation was investigated. However, compensatory proliferation was not affected in drICE or dcp-1 single mutants (Figure 2F,G). Nevertheless, we showed recently that drICE and dcp-1 have overlapping functions in the apoptotic pathway in Drosophila (Xu et al., 2006). Thus, we analyzed GMR-hid-induced compensatory proliferation in dcp-1;drICE double mutants. The wave of compensatory proliferation in GMR-hid eye discs is blocked in dcp-1;drICE double mutants (Figure 2H). Statistical quantification suggests that simultaneous loss of DrICE and Dcp-1 reduces the number of proliferating cells posterior to the SMW in GMR-hid eye discs to almost wild-type levels, similar to dronc mutants or overexpression of Diap-1 or P35 (Figure 2I). This indicates that the effector caspases DrICE and Dcp-1 are required for compensatory proliferation in GMR-hid eye discs.

Cells undergoing compensatory proliferation are undifferentiated

Although the developing eye disc is a monolayer epithelium, the nuclei of various cell types are occupying distinct positions in the epithelium (Wolff and Ready, 1993) (Figure 3A). At late 3rd instar stage, the nuclei of undifferentiated cells dominate the basal side of eye discs, while the nuclei of differentiated photoreceptor neurons and cone cells move toward the apical side (Figure 3A). Although massive cell death is induced by overexpression of Hid, the differentiation of all cell types, such as photoreceptor neurons and cone cells, is largely normal in larval GMR-hid eye discs (Figure 3B; data not shown). The invariant positions of the cellular nuclei posterior to the MF allow determining which type of cells are proliferating in the wave of compensatory proliferation. As shown in Figure 3B,B′, the nuclei of BrdU-positive cells and of Elav-positive photoreceptor neurons are distinct. Nuclei of proliferating cells are close to the basal surface of GMR-hid eye discs (see cross sections in Figure 3B,B′) suggesting that the cells undergoing compensatory proliferation in GMR-hid eye discs are undifferentiated. To further confirm this, an antibody against Yan, a protein highly expressed in all undifferentiated cells posterior to the MF (Rohrbaugh et al., 2002), was used. All cells in the wave of compensatory proliferation are Yan-positive (Figures 3C,D) indicating that they are undifferentiated. Nevertheless, it should be stressed that although the cells undergoing compensatory proliferation are of undifferentiated origin, they had exited the cell cycle and were committed to adopt cell fate soon. However, under the apoptotic conditions of GMR-hid a signal may be produced that triggers these cells to re-enter the cell cycle.

Interestingly also, although hid expression under control of GMR is present in the entire posterior half of the disc, the location of proliferating cells is rather restricted (Figure 3B, double arrowheads). We detect the majority of BrdU-positive cells in the wave of compensatory proliferation between ommatidial columns 6 and 10 posterior to the MF (see also Figure 3B). Because columns emerge at a rate of 60 to 90 min (Wolff and Ready, 1993), this indicates that compensatory proliferation is induced approximately 6–15 hours after initial expression of hid (see Discussion).

Next, we determined the range of the signal emitted by dying cells by generating clones of cells expressing hid. However, when clones of hid-expressing cells are induced too early, they die before the MF starts moving. Therefore, we induced clones of GMR-hid in the eye (Stowers and Schwarz, 1999). Using this approach, GMR-hid clones are induced early, but they do not express hid before the MF. Additionally, due to the high activity of the Flipase (Flp) recombinase controlled by eyeless-GAL4 (ey-GAL4 UAS-Flp; EGUF), most cells in the eye disc had undergone mitotic recombination, and were either GMR-hid clones (2× GMR-hid) or wild type twin-spots (2× GFP) (Figure 3E). Some heterozygous tissue (1× GFP, 1× GMR-hid) persists and is apoptotic (yellow arrows in Figure 3E). The majority of GMR-hid-induced cell death and compensatory proliferation occurs within clones (Figure 3F). Importantly, a few proliferating cells were also found immediately adjacent to GMR-hid clones (see arrows in Figure 3F,F′). This suggests that the proliferating signal induced by GMR-hid is a short-range signal.

Hedgehog expression is increased in photoreceptor neurons in GMR-hid

To investigate which signaling pathway is required for GMR-hid-induced compensatory proliferation, various signaling cascades were analyzed in GMR-hid mosaic eyes. Because Wg and Dpp expression is induced in apoptotic wing discs, they were examined first. However, none of these signaling pathways is induced in GMR-hid clones (Figures 4A,B). In contrast, a hh-lacZ reporter transgene indicates that transcripts of Hedgehog (Hh) are upregulated in GMR-hid clones (Figures 4C and S1). Quantification revealed that the hh-lacZ reporter is increased by about 50% in GMR-hid clones (Figure S1C). To exclude the possibility that condensation and concentration of cell contents in apoptotic cells may cause the appearance of increased levels of β-galactosidase in GMR-hid clones, the expression of a different lacZ reporter gene (kek1-lacZ) in GMR-hid clones was analyzed as a control. The unchanged intensity of kek1-lacZ expression in GMR-hid clones (Figure S1D–F) indicates that the increase of hh-lacZ expression is specifically induced by GMR-hid. Notably, while cells undergoing compensatory proliferation are located at the basal side of the disc (Figures 3B–D), hh-lacZ expression is specifically increased in apically-located photoreceptor neurons (inset in Figure 4D). Moreover, Hh protein accumulates in GMR-hid clones (Figure 4E,F). This accumulation is best visible in the basal side of GMR-hid clones (Figure 4G) where compensatory proliferation occurs (Figure 3B). Together, these observations suggest that apically-located photoreceptor neurons produce and secrete Hh which may stimulate compensatory proliferation of basally-located undifferentiated cells (see also Figure 3A). Importantly also, hh expression is dependent on effector caspases as the inhibitor P35 blocks the increase of hh-lacZ expression (Figure 4H) suggesting that effector caspases trigger hh expression in photoreceptor neurons.

Figure 4
Effector Caspases trigger Hh expression in photoreceptor neurons

Hedgehog signaling is required for compensatory proliferation in GMR-hid

Secreted Hh functions as a short-range signal and is involved in many developmental processes including proliferation (Hooper and Scott, 2005). Thus, we tested a genetic requirement of Hh signaling for GMR-hid-induced compensatory proliferation in three ways. First, because Hh signaling controls the activity of the transcription factor Ci (Hooper and Scott, 2005), we tested a requirement of Hh signaling for compensatory proliferation in GMR-hid discs by manipulating Ci. Overexpression of truncated Ci proteins (Ci76 or CiCe), but not of full-length Ci (CiFL), constitutively represses Hh signaling (Aza-Blanc et al., 1997; Methot and Basler, 1999). Thus, we expressed these Ci transgenes under control of the GMR promoter in GMR-hid eye discs to score an effect on compensatory proliferation. In these genetic backgrounds, development of eye discs including differentiation of photoreceptor neurons is normal (data not shown). Strikingly, compensatory proliferation in GMR-hid eye discs expressing the repressor forms Ci76 or CiCe is reduced by more than 60%, whereas apoptosis is unaffected by these transgenes (Figure 5B–D). Expression of CiFL in GMR-hid does not affect or even slightly increase compensatory proliferation compared to GMR-hid alone (Figure 5A,D).

Figure 5
Hh signaling is required for GMR-hid-induced compensatory proliferation

Second, we tested a requirement of hh itself for GMR-hid-induced compensatory proliferation. The temperature-sensitive allele hhts2 (Ma et al., 1993) allows analyzing compensatory proliferation in a homozygous hh mutant background. However, incubating GMR-hid;hhts2/hhts2 animals at the non-permissive temperature (29°C) caused severe developmental defects including MF stop, precluding us from analyzing a requirement of hh for compensatory proliferation. Nevertheless, we found that incubating GMR-hid;hhts2/hhts2 animals at the permissive temperature (18°C) reduces Hh signaling enough to block compensatory proliferation more than 50% relative to controls (Figure 5D–F). Interestingly, under these conditions apoptosis is even increased (white arrows in Figure 5F), yet compensatory proliferation is reduced, further supporting a requirement of hh for GMR-hid-induced compensatory proliferation.

Finally, we analyzed GMR-hid-induced compensatory proliferation in mutant clones of smoothened (smo), an essential component in the Hh signal transduction pathway (Hooper and Scott, 2005). Two independently generated strong smo alleles, smo3 and smoD16 (Chen and Struhl, 1998), were used and similar results were obtained. Importantly, although Hh is expressed in developing photoreceptor neurons (Figure 4D), cells that lack Smo, and hence the ability to transduce the Hh signal, can develop into normal ommatidia (Strutt and Mlodzik, 1997) indicating that the normal differentiating process still takes place in smo mutant clones. Notably, proliferation in the SMW is delayed but not absent in smo clones (Firth and Baker, 2005) (Figures 5G and S2, yellow arrows). In contrast, compensatory proliferation is largely abolished in smo clones in GMR-hid discs (compare Figures 5G,H and S2C with S2B, indicated by orange arrows). As control, GMR-hid-induced apoptosis still occurs in smo clones (compare Figure S2C with S2B). Interestingly, in about 50% of smo clones (n>40) in GMR-hid eye discs, a few proliferating smo mutant cells can be found immediately adjacent to wild type tissue (Figures 5H and S2C, white arrows). A similar observation was made when smo clones were generated in wild type eye discs (Figure S2A). This suggests that a low level of proliferation can be induced non-autonomously in the absence of Smo. In summary, the results of these three sets of experiments demonstrate that Hh signaling is directly required for compensatory proliferation in GMR-hid.

Distinct mechanisms of compensatory proliferation in differentiating and proliferating eye tissues

Our analysis of GMR-hid eye discs revealed a novel mechanism of apoptosis-induced compensatory proliferation requiring a non-apoptotic function of the effector caspases DrICE and Dcp-1 which activate Hh signaling. Do the differences in the regulation of compensatory proliferation described for the wing disc and shown here for the eye disc reflect inherent differences between wing and eye discs? Or, alternatively, does this difference reflect a difference in the developmental potential of the tissue analyzed? In 3rd instar larval wing discs, compensatory proliferation was analyzed during a stage in which the disc was largely proliferating. In contrast, we have analyzed compensatory proliferation posterior to the SMW, where the cells are normally G1-arrested and committed to differentiate. Thus, the observed differences could be due to differences in developmental potential, i.e. proliferating versus differentiating stages. To test this possibility we expressed hid in proliferating eye tissue using ey-Gal4 which is expressed anterior to the MF (Hauck et al., 1999). However, because under these conditions hid expression occurs early, hid-expressing cells died before we were able to analyze them. Therefore, we co-expressed the caspase inhibitor P35 to block cell death. Under these conditions, hid-expressing cells survived and could be analyzed for their ability to induce compensatory proliferation.

Compared to control discs expressing only P35 (Figure 6A), co-expression of Hid and P35 induces tissue overgrowth anterior to the SMW in eye discs (arrows in Figure 6B), but not in the posterior differentiating tissue. Quantification revealed that the anterior part of ey-hid-p35 eye discs is on average 61% larger than the anterior part of ey-p35 control discs (Figure 6C). Interestingly, although photoreceptor neurons still differentiate (data not shown), the progression of the MF is delayed or hindered in the posterior differentiating eye tissue (see yellow dashed lines in Figures 6B,E,G,I) which is likely due to ectopic expression of wg (see below). Consistent with this and similar to control discs (Figure 6D), Hh signaling is only detectable in differentiated neurons but not in the anterior overgrowing tissue (Figure 6E) suggesting that Hh is not ectopically activated when compensatory proliferation is induced in the proliferating eye tissue. In contrast, compared to control discs (Figures 6F,H), Wg and Dpp signaling cascades are ectopically induced in the anterior proliferating tissue (Figures 6G,I). Taken together, these data suggest that the mechanism of apoptosis-induced compensatory proliferation in the anterior proliferating eye discs and developing larval wing discs is similar, and that Wg and Dpp signaling, but not Hh, are involved.

Figure 6
Expression of Dpp and Wg is induced in apoptotic proliferating eye tissue


This study revealed that there are at least two distinct mechanisms that promote compensatory proliferation in response to apoptotic activity. The general difference between these two mechanisms lies in the developmental context of the tissue in which compensatory proliferation occurs. In proliferating wing and eye tissues, compensatory proliferation induced by extensive apoptosis is dependent on Dronc and Dp53 which induce Dpp and Wg expression (Figure 7). In contrast, in differentiating eye tissue, apoptosis induces compensatory proliferation through a novel mechanism requiring the effector caspases DrICE and Dcp-1, which induce Hh signaling in a non-apoptotic function (Figure 7).

Figure 7
Model of apoptosis-induced compensatory proliferation in proliferating and differentiating tissues

When cells stop proliferating and become committed to adopt cell fate, dramatic changes in gene expression are occurring. Given these changes in developmental plasticity, it is not surprising that distinct mechanisms of apoptosis-induced compensatory proliferation are employed in proliferating versus differentiating tissues. However, it should be noted that the proliferating capacity of differentiating tissues is rather restricted. In GMR-hid eye discs, although hid is expressed in all cells posterior to the MF, compensatory proliferation occurs only in cells that are still undifferentiated (Figure 3). Yet, even though they are undifferentiated, they have withdrawn from the cell cycle and under normal developmental conditions (i.e. without GMR-hid) they would soon be recruited to adopt cell fate. However, the apoptotic environment causing increased Hh signaling appears to be able to trigger re-entry of these cells into the cell cycle.

Interestingly, the Hh signal is specifically increased in photoreceptor neurons requiring a non-apoptotic activity of effector caspases. Hh signaling can then non-autonomously induce proliferation of undifferentiated cells at the basal side of the eye disc (Figures 3 and and4).4). However, overexpression of Hh posterior to the MF in wild type eye discs alone is not sufficient to induce a comparable wave of compensatory proliferation as in GMR-hid eye discs (data not shown). This suggests that cell cycle reentry requires activation of additional factors/pathways stimulated in apoptotic cells.

Although hid can stimulate increased Hh expression in photoreceptor neurons throughout the posterior half of the eye disc, compensatory proliferation is restricted to a certain distance (6–10 ommatidial columns) from the MF. This corresponds to approximately 6–15 hours of developmental time (Wolff and Ready, 1993), and may be the time required for cell cycle re-entry. Similarily, when mammalian cells that have exited the cell cycle are stimulated to re-enter the cell cycle they need about 8 hours to do this (Coller, 2007). The reason for this delay is unknown. Studying compensatory proliferation in GMR-hid eye discs may provide a genetic model to address this interesting problem.

It is not clear whether this novel effector caspase-, Hh-dependent pathway of compensatory proliferation also applies to other, or even all, differentiating tissues. However, what this study shows is that there are at least two distinct mechanisms of apoptosis-induced compensatory proliferation. It is also possible that other mechanisms of compensatory proliferation in different developmental context are going to be uncovered in the future. Interestingly, in developing larval wing discs, P35-dependent compensatory proliferation has been implicated in cell competition (Li and Baker, 2007). This suggests that, even in tissue with the same developmental potential, compensatory proliferation can occur with distinct mechanisms.

How cells sense different developmental contexts and operate distinct proliferating mechanisms in response to apoptotic stress is unknown. Specifically, where is the specificity and selectivity for distinct caspases coming from in tissues of different developmental potential? What are the mechanisms engaged by these caspases to trigger secretion of either Dpp and Wg, or Hh? These are questions which need to be addressed in the future.

Implications for cancer

This study has several implications for tumorigenesis. First, many tumors develop when quiescent cells re-enter the cell cycle. The mechanisms for cell cycle re-entry are largely unknown. Second, evasion from apoptosis is a hallmark of cancer (Hanahan and Weinberg, 2000). Many tumor cells are induced to undergo apoptosis. However, they do not die because they down-regulate essential components of the apoptotic pathway such as Apaf-1 and caspases (Anichini et al., 2006; Hajra and Liu, 2004). Thus, these ‘undead’ tumor cells may secrete mitogens which may induce compensatory proliferation similar to the Drosophila case. In this way, ‘undead’ cells may contribute to the growth of the tumor. A similar argument can be made for chemotherapy which in many cases attempts to activate the apoptotic program in a tumor cell. If the death of the tumor cell is blocked, or slow, mitogens may be produced and the tumor growth may be even more severe. This is very obvious in the apoptotic wing or anterior eye discs in Drosophila when apoptosis is blocked by P35. Under these conditions, overgrown wing and eye tissues are observed (Huh et al., 2004; Perez-Garijo et al., 2004; Ryoo et al., 2004; Wells et al., 2006). Thus, evasion of apoptosis may directly contribute to tumor growth. Finally, although increased Hh signaling can lead to various cancers (Pasca di Magliano and Hebrok, 2003), how Hh induces cellular proliferation and tissue overgrowth is not well understood. Mutations in Patched1, a negative regulator of sonic Hh, frequently give rise to human tumors (Rubin and de Sauvage, 2006). The exact cause is unknown. Our data here imply that Hh signaling may be involved in cell cycle re-entry allowing cells to resume proliferation.


Fly Strains and Crosses

Either GMR-hid10 (Grether et al., 1995) or GMR-hid326 (this study) with insertions on the 2nd and 3rd chromosome, respectively, were used. Both transgenic lines give similar waves of compensatory proliferation in developing eye discs. p535a-1-4 (Rong et al., 2002), droncI24 (Xu et al., 2005), dcp-1Prev1 (Laundrie et al., 2003), and UASp-P35 (Werz et al., 2005) are as described. GMR-diap1, GMR-p35 and drICEDelta1 were kindly provided by B. Hay (Hay et al., 1995; Hay et al., 1994; Muro et al., 2006); smo3 and smoD16 by M. Atkins and G. Mardon (Chen and Struhl, 1998); UAS-CiFL, UAS-Ci76 and UAS-CiCe by D. Kalderon (Aza-Blanc et al., 1997; Methot and Basler, 1999); GMR-hid FRT19A by R.S. Stowers (Stowers and Schwarz, 1999). hsFLP, FRT19A P[ubiGFP], FRT40A P[ubiGFP], dpp-lacZBS3.0, kek1-lacZ15A6, hh-lacZP30 and hhts2 were obtained from the Bloomington Stock Center. The Gal4/UAS system was used to generate ‘undead’ cells in the anterior proliferating eye tissue by generating UAS-hid/+; UASp-P35/+; ey-Gal4/+ flies. The ey-GAL4 line was kindly provided by G. Halder.

Mosaic Analysis

To examine GMR-hid clones in eye discs, late third instar larvae of the following genotype were analyzed: GMR-hid FRT19A/P[ubiGFP] FRT19A; ey-GAL4 UAS-Flp (EGUF)/+. For mosaic analysis with smo3 or smoD16 clones, 8–20h posthatching larvae of the following genotypes were heat shocked for 1h at 37°C, raised at room temperature and analyzed at the late 3rd instar larval stage: (1) Generation of smo clones in wild type background: hsFLP; smo FRT40A/P[ubiGFP] FRT40A. (2) Generation of wt clones in GMR-hid background: eyFLP; FRT40A/P[ubiGFP] FRT40A; GMR-hid. (3) Generation of smo clones in GMR-hid background: hsFLP (or eyFLP); smo FRT40A/P[ubiGFP] FRT40A; GMR-hid. In each of these experiments, more than 20 representative clones were analyzed.

Immunohistochemistry and BrdU Labeling

Eye-antennal imaginal discs from late 3rd instar larvae were dissected and labeled as described (Fan et al., 2005). Rabbit anti-Hh was kindly provided by T. Tabata (Takei et al., 2004). Commercial antibodies used are mouse anti-BrdU (BD), rabbit anti-BrdU (ICL Inc.), rabbit anti-PH3 (Upstate), rabbit anti-cleaved Caspase-3 (Cell Signaling), rat anti-ELAV, mouse anti-ELAV, mouse anti-βGAL, mouse anti-Wg and mouse anti-Yan (U. of Iowa, DSHB). Secondary antibodies were donkey Fab fragments from Jackson ImmunoResearch. Images were taken with either a Zeiss AxioImager equipped with ApoTome technology or a confocal microscope.

Quantification of Compensatory Proliferation

All BrdU-positive cells posterior to the SMW were counted to indicate how much compensatory proliferation was induced in various genetic backgrounds. The wild type and GMR-hid eye discs were counted as control. For each genetic background, 10 representative eye discs were counted.

Signal Intensity and Size Measurement

The average intensity of hh-lacZ labeling and disc size was quantified through the “histogram” function in Adobe Photoshop CS. To compare the size of antennal discs, anterior eye discs and posterior eye discs, animals were reared under similar growth conditions and the eye-antennal discs of late 3rd instar stage larvae were analyzed. Images of eye-antennal discs were taken with a 10× objective lens after fixation and staining. Anterior and posterior eye discs are divided by the SMW indicated by PH3 antibody labeling. All measurements were repeated independently and their statistical significance was evaluated through a two-tailed, unpaired Student’s t test.

Supplementary Material


Supplemental Figure S1. Transcription of Hh is increased in GMR-hid clones.

Late 3rd instar eye discs. GMR-hid clones are labeled by absence of GFP. Experiments with more than 20 mosaic eye dics examined were repeated twice and results are consistent.

(A,B,B) Eye disc labeled with GFP and an enhancer trap line of hh (hh-lacZ) (A). Enlarged views of the outlined region are shown in (B,B). Compared to surrounding wild type tissues, in GMR-hid clones (outlined), expression of hh-lacZ is increased.

(C) Quantification of the average intensity of hh-lacZ labeling in (B). Compared to surrounding wild type tissues, expression of hh-lacZ increases about 1.5 fold in GMR-hid clones.

(D,E,E) Eye disc labeled with GFP and a kek1-lacZ reporter (D). Enlarged views of the outlined region are shown in (E,E). Compared to surrounding wild type tissues, in GMR-hid clones (outlined), expression of kek1-lacZ is unchanged.

(F) Quantification of the average intensity of kek1-lacZ labeling in (E). Compared to surrounding wild type tissues, expression level of kek1-lacZ is unchanged in GMR-hid clones.

Supplemental Figure S2. Additional smo mutant analysis in wild-type and GMR-hid background.

Late 3rd instar eye discs labeled with GFP, BrdU and Cas3*. smo or wild-type clones are labeled by lack of GFP. The two waves of dying cells induced by GMR-hid are indicated by yellow arrowheads.

(A–A″) A wild type eye disc with a smo3 clone (boundary is outlined) affecting the SMW. The SMW is delayed in the clone (yellow arrows). In about 20% of smo clones (n>20), a few BrdU-positive cells can also be observed in the smo3 clone adjacent to the wild type tissue (white arrows). A few dying cells are induced in smo3 clones (orange arrows).

(B–B″) A GMR-hid eye disc with wild type clones (boundaries are outlined). The SMW, the wave of compensatory proliferation and two waves of dying cells are not affected by wild type clones.

(C–C″) GMR-hid eye disc with a smoD16 clones (boundary is outlined). The SMW is delayed, but not disrupted, in the clone (yellow arrows). In contrast, the wave of compensatory proliferation is completely disrupted in the clone (indicated by orange arrows) and not delayed although the clone extends to the posterior edge of the disc. In about 50% of smo clones (n>20) a few BrdU-positive cells are detectable adjacent to wild type tissue (white arrows). However, these cells are also observed in smo clones generated in wild-type eye discs (Figure S2A, white arrows) suggesting that a low level of proliferation can be induced in the absence of Smo independently of apoptosis. Cell death induced by GMR-hid is not affected in the clone (C″).

(D,D′) GMR-hid eye disc with a large smo3 clone (boundary is outlined) covering a large portion of the SMW and a small part of the wave of compensatory proliferation. The SMW is missing in the center of the clone with delayed proliferating cells at the clone boundary (yellow arrows). However, the following wave of compensatory proliferation is only affected directly in the smo clone (orange arrows), and is not delayed in smo+ tissue at the same latitude where smo clones affect the SMW. A single proliferating cell remains adjacent to the wild type tissue (white arrows) similar to the ones observed in smo clones in otherwise wild-type discs (Figure S2A).

(E,E′) GMR-hid eye disc with multiple smo3 clones generated by eyeless-Flipase (ey-FLP) (clone boundaries are outlined). Although the SMW is not disrupted, proliferation in the SMW is delayed (yellow arrows). In contrast, in a large smo3 clone the wave of compensatory proliferation is strongly disrupted (orange arrows), and not delayed although the clones are extending to the posterior edge of the eye disc. A single proliferating cell remains adjacent to the wild type tissue (white arrows) similar to the ones observed in smo clones in otherwise wild-type discs (Figure S2A).


We would like to thank Mardelle Atkins, Danial Kalderon, Kent Golic, Bruce Hay, Georg Halder, Graeme Mardon, Kim McCall, Hermann Steller, Steven Stowers, Tetsuya Tabata, the Bloomington Stock Center in Indiana, and the Developmental Studies Hybridoma Bank (DSHB) in Iowa, for fly stocks and reagents. We thank Michael Galko, Georg Halder, and Bill Klein for discussions. The project was supported by Grant R01 GM068016 form the National Institute of General Medical Sciences. We gratefully acknowledge support by the Robert A. Welch Foundation (G-1496).


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.


  • Anichini A, Mortarini R, Sensi M, Zanon M. APAF-1 signaling in human melanoma. Cancer Lett. 2006;238:168–179. [PubMed]
  • Aza-Blanc P, Ramirez-Weber FA, Laget MP, Schwartz C, Kornberg TB. Proteolysis that is inhibited by hedgehog targets Cubitus interruptus protein to the nucleus and converts it to a repressor. Cell. 1997;89:1043–1053. [PubMed]
  • Baker NE, Yu SY. The EGF receptor defines domains of cell cycle progression and survival to regulate cell number in the developing Drosophila eye. Cell. 2001;104:699–708. [PubMed]
  • Cashio P, Lee TV, Bergmann A. Genetic control of programmed cell death in Drosophila melanogaster. Semin Cell Dev Biol. 2005;16:225–235. [PubMed]
  • Chen Y, Struhl G. In vivo evidence that Patched and Smoothened constitute distinct binding and transducing components of a Hedgehog receptor complex. Development. 1998;125:4943–4948. [PubMed]
  • Coller HA. What’s taking so long? S-phase entry from quiescence versus proliferation. Nat Rev Mol Cell Biol. 2007;8:667–670. [PubMed]
  • Fan Y, Soller M, Flister S, Hollmann M, Muller M, Bello B, Egger B, White K, Schafer MA, Reichert H. The egghead gene is required for compartmentalization in Drosophila optic lobe development. Dev Biol. 2005;287:61–73. [PubMed]
  • Firth LC, Baker NE. Extracellular signals responsible for spatially regulated proliferation in the differentiating Drosophila eye. Dev Cell. 2005;8:541–551. [PubMed]
  • Garcia-Bellido A, Merriam JR. Parameters of the wing imaginal disc development of Drosophila melanogaster. Dev Biol. 1971;24:61–87. [PubMed]
  • Grether ME, Abrams JM, Agapite J, White K, Steller H. The head involution defective gene of Drosophila melanogaster functions in programmed cell death. Genes Dev. 1995;9:1694–1708. [PubMed]
  • Hajra KM, Liu JR. Apoptosome dysfunction in human cancer. Apoptosis. 2004;9:691–704. [PubMed]
  • Hanahan D, Weinberg RA. The hallmarks of cancer. Cell. 2000;100:57–70. [PubMed]
  • Hauck B, Gehring WJ, Walldorf U. Functional analysis of an eye specific enhancer of the eyeless gene in Drosophila. Proc Natl Acad Sci U S A. 1999;96:564–569. [PubMed]
  • Hawkins CJ, Yoo SJ, Peterson EP, Wang SL, Vernooy SY, Hay BA. The Drosophila caspase DRONC cleaves following glutamate or aspartate and is regulated by DIAP1, HID, and GRIM. J Biol Chem. 2000;275:27084–27093. [PubMed]
  • Hay BA, Wassarman DA, Rubin GM. Drosophila homologs of baculovirus inhibitor of apoptosis proteins function to block cell death. Cell. 1995;83:1253–1262. [PubMed]
  • Hay BA, Wolff T, Rubin GM. Expression of baculovirus P35 prevents cell death in Drosophila. Development. 1994;120:2121–2129. [PubMed]
  • Haynie JL, Bryant PJ. The effects of X-rays on the proliferation dynamics of cells in the imaginal disc of Drosophila Melanagaster. Rouxs Arch Dev Biol. 1977;183:85–100.
  • Hooper JE, Scott MP. Communicating with Hedgehogs. Nat Rev Mol Cell Biol. 2005;6:306–317. [PubMed]
  • Huh JR, Guo M, Hay BA. Compensatory proliferation induced by cell death in the Drosophila wing disc requires activity of the apical cell death caspase Dronc in a nonapoptotic role. Curr Biol. 2004;14:1262–1266. [PubMed]
  • James AA, Bryant PJ. A quantitative study of cell death and mitotic inhibition in gamma-irradiated imaginal wing discs of Drosophila melanogaster. Radiat Res. 1981;87:552–564. [PubMed]
  • Kondo S, Senoo-Matsuda N, Hiromi Y, Miura M. DRONC coordinates cell death and compensatory proliferation. Mol Cell Biol. 2006;26:7258–7268. [PMC free article] [PubMed]
  • Kuranaga E, Miura M. Nonapoptotic functions of caspases: caspases as regulatory molecules for immunity and cell-fate determination. Trends Cell Biol. 2007;17:135–144. [PubMed]
  • Lamkanfi M, Festjens N, Declercq W, Vanden Berghe T, Vandenabeele P. Caspases in cell survival, proliferation and differentiation. Cell Death Differ. 2007;14:44–55. [PubMed]
  • Laundrie B, Peterson JS, Baum JS, Chang JC, Fileppo D, Thompson SR, McCall K. Germline cell death is inhibited by P-element insertions disrupting the dcp-1/pita nested gene pair in Drosophila. Genetics. 2003;165:1881–1888. [PubMed]
  • Li W, Baker NE. Engulfment is required for cell competition. Cell. 2007;129:1215–1225. [PubMed]
  • Ma C, Zhou Y, Beachy PA, Moses K. The segment polarity gene hedgehog is required for progression of the morphogenetic furrow in the developing Drosophila eye. Cell. 1993;75:927–938. [PubMed]
  • Meier P, Silke J, Leevers SJ, Evan GI. The Drosophila caspase DRONC is regulated by DIAP1. Embo J. 2000;19:598–611. [PubMed]
  • Methot N, Basler K. Hedgehog controls limb development by regulating the activities of distinct transcriptional activator and repressor forms of Cubitus interruptus. Cell. 1999;96:819–831. [PubMed]
  • Milan M, Campuzano S, Garcia-Bellido A. Cell cycling and patterned cell proliferation in the Drosophila wing during metamorphosis. Proc Natl Acad Sci U S A. 1996;93:11687–11692. [PubMed]
  • Milan M, Campuzano S, Garcia-Bellido A. Developmental parameters of cell death in the wing disc of Drosophila. Proc Natl Acad Sci U S A. 1997;94:5691–5696. [PubMed]
  • Muro I, Berry DL, Huh JR, Chen CH, Huang H, Yoo SJ, Guo M, Baehrecke EH, Hay BA. The Drosophila caspase Ice is important for many apoptotic cell deaths and for spermatid individualization, a nonapoptotic process. Development. 2006;133:3305–3315. [PubMed]
  • Pasca di Magliano M, Hebrok M. Hedgehog signalling in cancer formation and maintenance. Nat Rev Cancer. 2003;3:903–911. [PubMed]
  • Perez-Garijo A, Martin FA, Morata G. Caspase inhibition during apoptosis causes abnormal signalling and developmental aberrations in Drosophila. Development. 2004;131:5591–5598. [PubMed]
  • Rohrbaugh M, Ramos E, Nguyen D, Price M, Wen Y, Lai ZC. Notch activation of yan expression is antagonized by RTK/pointed signaling in the Drosophila eye. Curr Biol. 2002;12:576–581. [PubMed]
  • Rong YS, Titen SW, Xie HB, Golic MM, Bastiani M, Bandyopadhyay P, Olivera BM, Brodsky M, Rubin GM, Golic KG. Targeted mutagenesis by homologous recombination in D. melanogaster. Genes Dev. 2002;16:1568–1581. [PubMed]
  • Rubin LL, de Sauvage FJ. Targeting the Hedgehog pathway in cancer. Nat Rev Drug Discov. 2006;5:1026–1033. [PubMed]
  • Ryoo HD, Gorenc T, Steller H. Apoptotic cells can induce compensatory cell proliferation through the JNK and the Wingless signaling pathways. Dev Cell. 2004;7:491–501. [PubMed]
  • Srivastava M, Scherr H, Lackey M, Xu D, Chen Z, Lu J, Bergmann A. ARK, the Apaf-1 related killer in Drosophila, requires diverse domains for its apoptotic activity. Cell Death Differ. 2007;14:92–102. [PMC free article] [PubMed]
  • Stowers RS, Schwarz TL. A genetic method for generating Drosophila eyes composed exclusively of mitotic clones of a single genotype. Genetics. 1999;152:1631–1639. [PubMed]
  • Strutt DI, Mlodzik M. Hedgehog is an indirect regulator of morphogenetic furrow progression in the Drosophila eye disc. Development . 1997;124:3233–3240. study, t. [PubMed]
  • Takei Y, Ozawa Y, Sato M, Watanabe A, Tabata T. Three Drosophila EXT genes shape morphogen gradients through synthesis of heparan sulfate proteoglycans. Development. 2004;131:73–82. [PubMed]
  • Wells BS, Yoshida E, Johnston LA. Compensatory proliferation in Drosophila imaginal discs requires Dronc-dependent p53 activity. Curr Biol. 2006;16:1606–1615. [PMC free article] [PubMed]
  • Werz C, Lee TV, Lee PL, Lackey M, Bolduc C, Stein DS, Bergmann A. Mis-specified cells die by an active gene-directed process, and inhibition of this death results in cell fate transformation in Drosophila. Development. 2005;132:5343–5352. [PMC free article] [PubMed]
  • Wolff T, Ready DF. Pattern formation in the Drosophila retina. In: Martinez-Arias MBaA., editor. The Development of Drosophila melanogaster. Plainview, N.Y: Cold Spring Harbor Press; 1993. pp. 1277–1325.
  • Xu D, Li Y, Arcaro M, Lackey M, Bergmann A. The CARD-carrying caspase Dronc is essential for most, but not all, developmental cell death in Drosophila. Development. 2005;132:2125–2134. [PMC free article] [PubMed]
  • Xu D, Wang Y, Willecke R, Chen Z, Ding T, Bergmann A. The effector caspases drICE and dcp-1 have partially overlapping functions in the apoptotic pathway in Drosophila. Cell Death Differ. 2006;13:1697–1706. [PMC free article] [PubMed]
  • Yoo SJ, Huh JR, Muro I, Yu H, Wang L, Wang SL, Feldman RM, Clem RJ, Muller HA, Hay BA. Hid, Rpr and Grim negatively regulate DIAP1 levels through distinct mechanisms. Nat Cell Biol. 2002;4:416–424. [PubMed]
  • Yu SY, Yoo SJ, Yang L, Zapata C, Srinivasan A, Hay BA, Baker NE. A pathway of signals regulating effector and initiator caspases in the developing Drosophila eye. Development. 2002;129:3269–3278. [PubMed]