Introduction

Apoptosis affects many aspects of development, including shaping of body structures, controlling cell number and eliminating abnormal, misplaced or harmful cells 1, 2. Moreover, cell killing in response to death signal stimuli is essential to remove cells with neoplastic transformation potential 3, 4. The pathways controlling apoptosis are highly conserved from flies to humans. The core of the apoptosis cell death machine consists of members of the caspase family of proteases 5 whose activation is regulated either by an extrinsic or by an intrinsic signaling pathway. In Drosophila, the intrinsic pathway is stimulated by death-inducing signals that activate the RHG family (reaper, hid, grim, sickle and jafra-2) of pro-apoptotic proteins, leading to the activation of the effector caspases and the induction of apoptosis (reviewed in Domingos et al. 6). Likewise, in mammals the intrinsic pathway is stimulated by intracellular stimuli such as DNA damage and cytotoxic drugs that act inside the cell activating the pro-apoptotic proteins and leading to the activation of caspases 7. The extrinsic pathway in mammals, also referred to as the receptor-mediated apoptotic pathway, plays an important role in both development and disease 7. After an apoptotic stimulus, the TNF/TNFR (tumor necrotic factor/tumor necrotic factor receptor) signaling is activated and mediates cell death through both the activation of the JNK pathway and the activation of caspases. Additionally, in mammals, initiation of receptor-mediated apoptosis activates the formation of the death receptor complex, composed of TNFR associated factors (TRAFs), which function with the adaptor protein FADD to recruit and aggregate procaspase-8, thereby activating caspase-8 through mutual cleavage and initiating apoptosis 8. This pathway, however, is poorly understood in Drosophila. The Drosophila TNF/TNFR homologs are EIGER/WENGEN 9, 10. Three Drosophila members of the TRAF family have been identified 11, 12. The Drosophila homologs of the DED (Death Domain)-containing protein FADD and caspase-8 are dFADD and DREDD, respectively 13, 14. However, unlike its mammalian counterpart, the apoptotic effect of EIGER requires neither the activity of the caspase-8 homolog DREDD nor the adaptor dFADD protein, but is completely dependent on the activation of the JNK signaling pathway 10.

The dRYBP gene encodes a phylogenetically conserved 150-amino acid protein with a conserved amino terminus, which contains a NZF (Npl4 zinc finger) type zinc-finger 15 (Figure 1L) and a carboxy-terminus with so far no identified functional domains. Recently, a subgroup of NZF domains that includes that of RYBP has been shown to possess ubiquitin binding activity 16. We have previously shown that loss of dRYBP function produces phenotypes that are highly variable in both penetrance and expressivity 17. Moreover, we showed that dRYBP interacts genetically and molecularly with Polycomb (PcG) and trithorax (trxG) proteins and that its overexpression generates a small wing phenotype 17, 18.

Figure 1
figure 1

Exogenous dRYBP expression induces apoptotic cell death. (A) Activated caspase-3 (C3) staining (red) in a wild-type wing disc. (B) GFP expression (green) in a nub-Gal4, UAS-GFP wing disc showing the nub-Gal4 line domain of expression in the wing disc. (C) dRYBP expression (red) in a nub-Gal4, UAS-GFP/ UAS-dRYBP wing disc. (D) C3 staining (red) in a nub-Gal4, UAS-GFP/UAS-dRYBP wing disc. (E) C3 staining (red) in a nub-Gal4, UAS-GFP/UAS-dRYBP haltere disc. (F) Merged expression of GFP (green, showing the domain of expression of the nub-Gal4 line in the haltere disc) and C3 staining (red) in a nub-Gal4, UAS-GFP/UAS-dRYBP haltere disc. (G) C3 staining (red) in a dll-Gal4, UAS-GFP/UAS-dRYBP third leg disc. (H) GFP expression (green) in a dll-Gal4, UAS-GFP/UAS-dRYBP third leg disc, showing the dll-Gal4 line domain of expression in the leg disc. (I) C3 staining (red) in a nub-Gal4, UAS-GFP/UAS-dRYBP/+; UAS p35/+ wing disc. This wing disc illustrates the differential staining of the anti-C3 antibody when activated caspases are inhibited by overexpression of p35 in the nub-Gal4 domain (arrow, see detail in M) and when activated caspases are not inhibited by overexpression of p35 outside the nub-Gal4 domain (arrowhead, see detail in N). (J) C3 staining (red) in dll-Gal4, UAS-GFP/UAS-dRYBP antenna disc. (K) Merged expression of GFP (green) and C3 staining (red) in dll-Gal4, UAS-GFP/UAS-dRYBP antenna disc. (L) Scheme of the UAS-dRYBP and UAS-dRYBP-ΔZF constructs. The NZF (indicated in red) is located at the amino-terminus of the protein. (M) Detail of I – arrow showing staining of C3 in p35-apoptosis inactivated cells. (N) Detail of I – arrowhead showing staining of C3 in apoptotic cells. (O) UBX expression (blue) in nub-Gal4, UAS-GFP/dRYBP-ΔZF wing disc. (P) C3 staining in a nub-Gal4, UAS-GFP/dRYBP-ΔZF wing disc. (Q) Merged expression of UBX, C3 and GFP in nub-Gal4, UAS-GFP/dRYBP-ΔZF. The de-repression of UBX only takes place in a few cells within the nub-Gal4 domain.

Full size image

In mammals, high levels of the RYBP protein have been found to have tumor-specific killing activity, a characteristic that could potentially be exploited to develop anti-cancer treatments 19. Furthermore, the RYBP protein has been found to interact with Apoptin, a protein known to be active in killing tumor cells 20. As induction of apoptosis is a common mechanism underlying cancer therapy, understanding the mechanisms controlling tumor cell-specific killing may facilitate the development of improved therapies.

We show here that overexpression of dRYBP and overexpression of TRITHORAX, an epigenetic regulator of transcription 21, induce apoptosis in the imaginal discs. Moreover, we show that high levels of dRYBP and high levels of TRITHORAX activate the expression of a reaper-lacZ transgene. We also show that the pro-apoptotic gene reaper is a direct target of the dRYBP protein and that dRYBP-induced apoptosis requires the function of reaper and trithorax. These findings reveal the participation of a factor known to be an epigenetic regulator of gene transcription in the control of apoptosis. Furthermore, we show that dRYBP-induced apoptosis is independent of the JNK pathway but requires DREDD and dFADD, two factors previously shown not to be associated with receptor-mediated apoptosis in Drosophila 10.

Results

Endogenous dRYBP is not required for developmental apoptosis

Loss of function of the dRYBP gene results in phenotypes that are highly variable both in penetrance and in expressivity 17. In this report, we have investigated whether the inactivation of dRYBP function using RNAidRYBP transgenic lines and loss of function dRYBP mutations have an effect on apoptosis. We first performed caspase-3 22 and Terminal deoxynucleotidyl transferase-mediated dUTPNick End Labelling (TUNEL) staining in the wing imaginal discs of sd-GAL4/+; UAS-RNAidRYBP/UAS-RNAidRYBP larvae and the wing imaginal discs of dRYBP1/dRYBP1 larvae and found similar results to those in the wild-type wing imaginal discs. Second, we looked at Acridine Orange staining during embryonic development of the progeny of dRYBP1/CyO-GFP stock. Compared to wild-type embryos (Df (1)w67c23, see Materials and Methods), mutant (non-GFP) embryos displayed no appreciable changes in cell death throughout the central nervous system during nerve cord condensation in stage 16 embryos or in the embryonic head (data not shown). These results suggest that dRYBP function is not required for developmental apoptosis.

Exogenous dRYBP expression induces apoptotic cell death

Using the GAL4/UAS system, we observed that overexpression of the dRYBP protein in wing imaginal discs produces small adult wings 17, 18. We asked whether the reduction of wing size was due to the induction of apoptosis. We have used several GAL4 drivers (en-Gal4, ci-Gal4, sd-Gal4, nub-Gal4, 248-Gal4 and dll-Gal4; see Materials and Methods) to induce high levels of dRYBP and found that overexpression of dRYBP induces apoptosis in all imaginal discs as evidenced by the activation of caspase-3 (Figure 1D-1K). We have also investigated whether the murine RYBP/DEDAF protein can induce apoptosis in wing imaginal discs. Similar to the expression of the Drosophila protein (Figure 1D), we found that the wing imaginal discs of sd-Gal4/+; UAS-RYBP murine/+ show induction of apoptosis (not shown). To assess the contribution of apoptosis to the dRYBP overexpression phenotype, we used the baculovirus caspase inhibitor p35 to block apoptosis 23. Co-expression of p35 and dRYBP in nub-Gal4/UAS dRYBP; UAS p35/+ represses the dRYBP-induced apoptosis (Figure 1I, 1M and 1N). Moreover, the small-size wing phenotype observed when high levels of dRYBP are expressed in the wing is rescued when p35 is also overexpressed (data not shown). This result indicates that caspases are necessary for dRYBP-induced apoptosis.

We have shown that overexpression of dRYBP in wing discs causes the homeotic Ultrabithorax (UBX) protein 24 to be expressed in a limited number of wing disc cells 17, 18. This UBX expression is not dependent on the presence of the NZF domain of the dRYBP protein as overexpression of the dRYBP-ΔZF and full-length proteins (Figure 1L) produced the same pattern of UBX expression in wing discs (Figure 1O). Furthermore, in contrast to overexpression of full-length dRYBP, wing discs of nub-Gal4/UAS-dRYBP-ΔZF larvae do not show activation of caspase-3 (Figure 1P). As shown in Figure 1, the wing imaginal discs overexpressing the truncated dRYBP-ΔZF protein show expression of the homeotic UBX protein (Figure 1O) but do not show activation of caspase-3 (Figure 1P), indicating that the induction of apoptosis is not a consequence of UBX expression.

These results demonstrate that exogenous expression of either dRYBP or murine RYBP induces apoptosis in the Drosophila imaginal disc cells. Moreover, this induction is not a consequence of UBX expression. Finally, this induction of apoptosis does not require the cells to be tumoral as it occurs in normal imaginal disc cells.

dRYBP-induced apoptosis is dependent on both pro-apoptotic and anti-apoptotic factors

To gain insight into the molecular pathways controlling dRYBP-induced apoptosis, we investigated the requirement for pro-apoptotic factors 25 which act by inhibiting the activity of the Drosophila Inhibitor of Apoptosis (DIAP) anti-apoptotic proteins 26. Three of the pro-apoptotic genes described in Drosophila – reaper, hid and grim – are contained within a single genomic deficiency – Df (3L) H99 (27; see Materials and Methods). Imaginal discs from nub-Gal4/UAS dRYBP; Df (3L) H99/+ larvae show a marked reduction in dRYBP-induced apoptosis (Figure 2B) and a partial rescue of the wing size (data not shown). Furthermore, high levels of the DIAP1 protein in nub-Gal4/UAS dRYBP; UAS DIAP1/+ larvae led to a complete inhibition of apoptosis in wing imaginal discs (Figure 2C) and accordingly, rescue of the wing size (data not shown). Together, these results show that dRYBP-induced apoptosis requires the function of the pro-apoptotic Reaper, Hid and Grim proteins, and that this process is inhibited by high levels of DIAP1. This, in turn, suggests that dRYBP-induced apoptosis is dependent on the canonical apoptotic pathways in which these factors function.

Figure 2
figure 2

Factors involved in the dRYBP induction of apoptosis. (A) C3 staining (red) in nub-Gal4, UAS-GFP/dRYBP wing disc. (B) C3 staining (red) in nub-Gal4, UAS-GFP/dRYBP; Df(3L)H99/+ wing disc. Apoptosis is diminished. (C) C3 staining (red) in nub-Gal4, UAS-GFP/dRYBP; UAS-DIAP1/+. Apoptosis is almost absent. (D) C3 staining (red) in dreddD44/dreddD44; nub-Gal4, UAS-GFP/UAS-dRYBP wing imaginal disc. Apoptosis is greatly suppressed. (E) C3 staining (red) in nub-Gal4, UAS-GFP/UAS-RNAiFADDM8; UAS-dRYBP/+ wing disc. Apoptosis is nearly completely absent. (F) dRYBP expression (red) in a sd-Gal4/+; UAS-dRYBP/rpr-lacZ wing disc indicating the domain of expression driven by the sd-Gal4 line. (G) β-Gal expression (purple) of a wing disc sd-Gal4/+; rpr-lacZ showing the expression of the reaper-lacZ construct, which is not affected by the presence of the sd-Gal4 line. (H) C3 staining (red) in sd-Gal4/+; UAS-dRYBP/rpr-lacZ wing disc. (I) β-Gal expression (purple) in a sd-Gal4/+; UAS-dRYBP/rpr-lacZ wing disc. (J) ChIP with dRYBP antibody in 4-12 h embryos. dRYBP antibody (orange bars). Mock (green bars). The immunoprecipitated DNA was analyzed by qPCR using primers for the promoter region of reaper (rpr), for coding sequences (CDS) of reaper and for robo 3 gene as a control.

Full size image

The reaper gene promoter is a direct target of the dRYBP protein

To investigate whether dRYBP regulates the expression of apoptotic genes, we studied the effect of dRYBP on the expression of reaper by monitoring β-gal expression in sd-Gal4/+; UAS-dRYBP/rpr-lacZ wing imaginal discs. As shown in Figure 2G-2I, expression of rpr-lacZ in the wing discs is activated by dRYBP overexpression, indicating that high levels of dRYBP cause mis-regulation of reaper expression.

We next studied whether the reaper gene is bound by the dRYBP protein by performing chromatin immunoprecipitation (ChIP) experiments in wild-type Drosophila embryos using an antibody against dRYBP. Quantitative PCR was used to identify regions in and around the reaper gene enriched in dRYBP ChIP products and that are, therefore, likely to be bound by the dRYBP protein (Materials and Methods). As shown in Figure 2J, the reaper gene promoter is enriched in the dRYBP ChIP products compared to the control region.

These results clearly indicate that in wild-type imaginal discs the pro-apoptotic gene reaper is a direct target of dRYBP and suggest that its transcription is activated when dRYBP is expressed at high levels.

dRYBP-induced apoptosis is independent of the JNK pathway and is dependent on the DED-domain containing proteins dFADD and DREDD

Activation of the receptor-mediated extrinsic death pathway upon apoptotic stimuli induces cell death through activation of the JNK pathway 9, 10. In Drosophila, activation of the JNK pathway is independent of the DED-domain containing proteins DREDD (caspase-8 in humans) and dFADD (the Fas-associated death-domain containing protein, FADD in humans) 10. In humans, however, receptor-mediated apoptosis is dependent on FADD and caspase-8 28. By monitoring puckered expression using puckered-lacZ E69 transgenic flies 29, we studied whether dRYBP-induced apoptosis is able to activate the JNK pathway. We found that the β-gal expression pattern is identical in wing imaginal discs from sd-Gal4/+; UAS-dRYBP/+; puc-lacZ/+ and from control puc-lacZ/+ larvae, indicating that exogenous dRYBP does not activate the JNK pathway (not shown).

We next studied the requirement of DREDD and dFADD in dRYBP-induced apoptosis. dreddD44/dreddD44; nub-Gal4/UAS-dRYBP wing imaginal discs do not show activation of caspase-3 (Figure 2D). Additionally, nub-Gal4/ UAS-RNAiFADDM8; UAS-dRYBP/+ wing imaginal discs show no activation of apoptosis (Figure 2E). The overexpression of dRYBP in the absence of either dredd or dFADD results in flies with wings of a bigger size (data not shown) due to the absence of apoptosis in those mutant conditions. These results indicate that, like receptor-mediated apoptosis in mammals, exogenous dRYBP-induced apoptosis in fly imaginal discs is dependent on DREDD and dFADD expression.

dRYBP-induced apoptosis is trithorax dependent but Polycomb and Sex Comb extra independent

Several lines of evidence indicate that dRYBP interacts with the Polycomb group (PcG) and trithorax group (trxG) of proteins 17, 18. We used PcG and trxG mutant backgrounds to investigate whether apoptosis induced by dRYBP is dependent on these proteins. The activation of caspase-3 is clearly diminished in wing imaginal discs of nub-Gal4/UAS dRYBP; trxE2/+ larvae (Figure 3B), while there is no change in either nub-Gal4/UAS dRYBP; Pc3/+ larvae (Figure 3C) or nub-Gal4/UAS dRYBP; Sce1/+ larvae (Figure 3D). These results lead to the conclusion that trithorax, but not Polycomb or Sex comb extra, affects dRYBP-induced apoptosis, although it is possible that lack of only one dose of Pc or Sce is not sufficient to influence the induction of apoptosis. These results indicate that the trithorax protein functions in dRYBP-induced apoptosis and links the epigenetic regulation mediated by the trxG genes to the processes of programmed cell death.

Figure 3
figure 3

dRYBP-induced apoptosis is trithorax dependent and high levels of Trithorax induce apoptosis. (A) C3 staining (red) in a nub-Gal4, UAS-GFP/UAS-dRYBP wing disc. (B) C3 staining (red) in a nub-Gal4, UAS-GFP/UAS-dRYBP; trxE2/+ wing disc. Note the decrease, compared with (A), in the number of C3 stained cells. (C) C3 staining (red) in a nub-Gal4, UAS-GFP/UAS-dRYBP; Pc3/+ wing disc. (D) C3 staining (red) in a nub-Gal4, UAS-GFP/UAS-dRYBP; Sce1/+ wing disc. (E) C3 staining (red) in a sd-Gal4/+; UAS-TRX/+ wing disc. (F) β-Gal expression (purple) in a sd-Gal4/+; rpr-lacZ/+; UAS-TRX/+ wing disc.

Full size image

Exogenous TRITHORAX induces apoptosis and activation of reaper

We investigated whether developmental apoptosis is modified in trxE2 homozygous mutant embryos and found that activated caspase-3 staining in those embryos is similar to wild-type embryos (data not shown), suggesting that the Trithorax protein is not required in developmental apoptosis. We next investigated whether high levels of POLYCOMB, SEX COMB EXTRA and TRITHORAX are capable of inducing apoptosis. Wing imaginal discs from neither nub-Gal4/+; UAS-SCE/+ nor nub-Gal4/+; UAS-PC/+ larvae showed induction of apoptosis (data not shown). However, a high level of TRITHORAX-induced apoptosis was indicated by the clear activation of caspase-3 in sd-Gal4/+; UAS-TRX/+ imaginal discs (Figure 3E). Moreover, to study whether exogenous TRX is capable of inducing the expression of pro-apoptotic genes, we looked at the β-gal expression in imaginal discs of sd-Gal4/+; rpr-lacZ/+; UAS-TRX/+ larvae and found that rpr-lacZ expression is activated in the wing imaginal discs (Figure 3F). These results demonstrate the ability of exogenously expressed trithorax protein to both induce apoptosis as well as activate the expression of the pro-apoptotic reaper gene.

Discussion

High levels of exogenous dRYBP induce apoptosis

In the present work, we have shown that loss of function of dRYBP does not affect the apoptosis associated with development. However, high levels of dRYBP expression induce apoptosis in all the imaginal discs studied (Figure 1). In transfection experiments, murine and human RYBP/DEDAF proteins have been shown to have the ability to specifically kill mammalian tumor cells while leaving normal cells unaffected 19, 20. We have shown that the induction of apoptosis in Drosophila does not require the apoptotic cell to be in the transformed state, indicating that this may be a useful system to study the mechanisms controlling the tumor-specific cell killing.

Factors involved in the induction of dRYBP-mediated apoptosis

We have shown that dRYBP-induced apoptosis depends on the pro-apoptotic reaper, hid and grim genes (Figure 2C) and that it is inhibited by DIAP1 (Figure 2C). How dRYBP-induced cell death is regulated by these factors is not known. However, we have shown that the promoter region of the reaper gene is bound by the dRYBP protein (Figure 2J) and that dRYBP overexpression activates the expression of a reaper-lacZ construct (Figure 2I). These results suggest that dRYBP regulates the transcription of reaper and that dRYBP-induced apoptosis occurs via the canonical cell death pathways.

We have also shown that dRYBP-induced apoptosis is dependent on the DREDD (Figure 2D) and dFADD proteins (Figure 2E). DREDD is the Drosophila homolog of the mammalian apical caspase-8 13, while dFADD is the Drosophila homolog of mammalian FADD, an adaptor protein involved in the TNFR signaling complex 14. Both caspase-8 and FADD are required for the receptor-mediated apoptosis in mammals. However, apoptosis induced by the Drosophila TNF homolog EIGER 10 does not require dFADD or DREDD 10. These results raise the possibility that dRYBP, dFADD and DREDD define a novel Drosophila extrinsic apoptotic pathway. Moreover, studying the mechanisms of dRYBP-mediated cell death in Drosophila might help to define the cell death functions of dFADD and DREDD as well as to better characterize the TNFR apoptotic pathways.

Finally, our results show that dRYBP-induced apoptosis requires the function of trithorax and that TRITHORAX itself is able to induce apoptosis. These findings connect the mechanisms of epigenetic regulation of gene expression with the mechanisms of apoptosis.

Are dRYBP and TRX involved in the epigenetic regulation of stress-induced apoptosis?

Several observations lead us to propose that dRYBP might be involved in apoptosis associated with cellular stress response. First, we have found no evidence indicating dRYBP involvement in developmental apoptosis. Second, loss of function of the dRYBP gene results in phenotypes which are highly variable both in penetrance and in expressivity 17. Third, high levels of RYBP/DEDAF have been found in a number of different cancers, including T-cell and Hodgkin´s lymphomas, pituitary adenomas and oligodendrogliomas 30. Fourth, it has been shown that the ability of v-Fos transformed cells to invade other tissues is dependent upon repression of RYBP expression. Moreover, re-expression of RYBP in Fos-transformed cells is sufficient to strongly inhibit invasion 31. Finally, it has been shown in transfection experiments that high levels of RYBP/DEDAF induce tumor cell-specific killing. Therefore, either by induction of apoptosis or by inhibition of invasion, the RYBP protein appears to be able to detect aberrant cell behavior and initiate apoptotic pathways.

We show that high levels of TRX induce apoptosis and, furthermore, that dRYBP-induced apoptosis requires the function of trithorax. These results indicate that the TRX protein might be involved in control of apoptosis and suggest that dRYBP and TRITHORAX could function in the regulation of stress-induced apoptosis.

In our model (Figure 4), under normal conditions of development (Figure 4, left panel), the function of the dRYBP protein remains latent and developmental apoptosis takes place through the canonical pathways of apoptosis proceeding through the activation of the effector caspases followed by cell death. Under conditions (Figure 4, right panel) where high levels of dRYBP are induced – and similar to conditions in tumor cells where high levels of RYBP have been found 30 – the reaper gene is transcriptionally activated via the dRYBP-dependent recruitment of trxG/PcG complexes. Critical to our model is the establishment of the correct balance between PcG and trxG proteins so that appropriate control of apoptotic gene expression is maintained. This model is consistent with the observed requirement of TRX protein for dRYBP-induced apoptosis (Figure 3B). Although a high level of TRITHORAX can itself activate apoptosis, we propose that the activation of transcription of the apoptotic genes, mediated by the TRX protein, is faster and more accurate when dRYBP is mediating and balancing the recruitment of the PcG and trxG proteins.

Figure 4
figure 4

Role of dRYBP and TRX in the epigenetic regulation of apoptosis. Hypothetical model. Developmental apoptosis takes place through the activation, under death stimuli, of the pro-apoptotic genes, which removes the inhibitory action of the DIAP proteins, which in turn allows the initiator and the effector caspases to be activated and induce apoptosis. During developmental apoptosis the function of the dRYBP protein is not required and remains latent. The function of neither dFADD nor the initiator caspase DREDD is required in this process. We hypothesize that the levels of the dRYBP protein are increased, perhaps in response to cellular stress (see text), resulting in the induction of apoptosis. (1) High levels of dRYBP cause the aggregation of the dFADD and the DREDD proteins, thus promoting the activation of DREDD and triggering apoptosis. (2) High levels of dRYBP activate the expression of the pro-apoptotic gene reaper through the recruitment of the TRX protein, resulting in the induction of apoptosis.

Full size image

We have shown that dRYBP-induced apoptosis requires the function of DREDD and dFADD (Figure 2D and 2E). In mammals the FADD and procaspase-8 proteins are required for the apoptotic response mediated by the TNF/TNFR complex. This is not the case in Drosophila as the fly TNF/TNFR homologs EIGER/WENGEN do not require dFADD and DREDD to trigger apoptosis. Moreover, in mammals the RYBP/DEDAF protein mediates the interaction between pro-caspase-8 and FADD to activate caspase-8. This leads us to propose that an additional function of the dRYBP protein is to mediate the interaction between dFADD and DREDD to activate the caspase triggering in the apoptotic pathway. dRYBP and TRITHORAX functions may define a novel apoptotic induction pathway that responds to environmental-stress and mutant-stress signals and that requires dFADD and DREDD.

dRYBP-induced apoptosis does not take place in the absence of pro-apoptotic genes reaper, hind and grim (Figure 2B) or in the absence of DREDD/dFADD (Figure 2E and 2D). It has been previously shown that overexpression of REAPER and overexpression of HID require the presence of dredd to induce apoptosis. Thus, DREDD is an effector of REAPER and HID signaling pathways 13. Moreover, expression of Reaper, Grim and Hid was found to trigger the processing of the DREDD protein precursor 13. These results support our model and explain why induction of apoptosis by dRYBP is abolished in the absence of either the pro-apoptotic proteins or the adaptor dFADD protein.

How the activation of reaper expression and the activation of DREDD processing function in a pathway induced upon dRYBP overexpression is not known. Nor is it known what stress signals would lead to increased levels of dRYBP and thereby promote apoptosis. We have attempted to identify treatments which would result in increased levels of dRYBP expression in the imaginal discs. None of the treatments investigated – X-ray irradiation, overexpression of EIGER 10, overexpression of HEMIPTEROUS 32, overexpression of REAPER 33, overexpression of TRITHORAX or overexpression of p53 34 – resulted in increased expression of dRYBP (data not shown). The study of factors controlling dRYBP expression might help to clarify the role of the mammalian RYBP gene in cancer development. Furthermore, elucidating the mechanisms of dRYBP-dependent cell killing and the role of epigenetic regulation by TRX might provide a path to the development of new cancer therapies.

Materials and Methods

Drosophila strains and handling

Flies used include stocks containing mutations Pc3, Sce1, trxE2 and Df (1) w67c23 (y-, w-), dreddD44, Df(3L)H99 (a deletion that eliminates the pro-apoptotic genes reaper, grim and hid), and they are all described in FlyBase (http://flybase.org). Additionally, the null mutation dRYBP1 17, 35 was used; the stock dRYBP1/ CyO-GFP was used to select homozygous dRYBP1 embryos and larvae. Transgenic flies containing either puckered-lacZE69 (puc-lacZ) 29 or reaper-4kb-lacZ (rpr-lacZ) 36 were used to detect the expression of puckered and reaper, respectively. For overexpression experiments, the GAL4/UAS system was used 37 at 25 ºC and 29 ºC with the following lines (described in Calleja et al. 38): scalloped-Gal4 (sd-Gal4), nubbin-Gal4 (nub-Gal4), apterus-Gal4 (ap-Gal4), distal less-Gal4 (dll-Gal4, MD23 and EM212) and 248-Gal4. Finally, transgenic lines containing UAS-RNAidRYBP 18, UAS-RNAiFADDM8 39, UAS-DIAP1 40, UAS-EIGER 10, UAS-p35 23, UAS-Dmp53 34, UAS-TRX (P-EP-TRX, gift of Isabel Guerrero), UAS-SCE 41, UAS-RYBP murine and UAS-PC (both constructed in this work), UAS-dRYBP 18 and UAS-dRYBP-ΔZF 17 were used. For imaginal disc irradiation, eggs were collected during a 24-h laying period, allowed to develop until the formation of second instar larvae and then irradiated. After 24 h, the larvae were collected for dissection and immunostaining. Irradiation was performed using a Philips X-ray instrument at a dose of 1 500R.

Staining procedures

Imaginal discs were dissected in PBS and fixed with 4% paraformaldehyde in PBS for 30 min at room temperature. After blocking in a PBT solution containing 1% bovine serum albumin (BSA) and 0.3% Triton X-100 for 30 min, discs were incubated overnight at 4 °C with the primary antibodies. Following three 10-min washes with PBT (PBS and 0.3% Triton X-100), they were incubated in the dark with the appropriate fluorescently labelled secondary antibody for 1 h at room temperature, washed in PBT and mounted in Vectashield (Vector Laboratories). The primary antibodies used were rabbit anti-dRYBP (1:100) 18, rabbit anti-C3 (1:200) (Cell Signalling Technologies) raised against the cleaved form of human caspase-3, which has been shown to cross-react with cleaved Drosophila caspase-3 22, rabbit anti-β-Gal (1:200) (Promega), mouse anti-β-Gal (1:200) (Promega), mouse anti-UBX 24 and rabbit anti-phospho-Histone H3 (1:100) (Ser10) (Upstate). Images were generated using either a MicroRadiance (Bio-Rad) or an LSM510 META (Zeiss) confocal microscope, and subsequently processed using Adobe Photoshop.

TUNEL staining was performed using imaginal discs dissected, fixed and incubated with the primary and secondary antibodies as described above for immunostaining. After incubation with the secondary antibody, they were incubated at 65 °C with NC-T solution (495 μl of sodium citrate 100 mM, 5 μl of Triton X-100 10%) for 30 min, washed three times for 10 min in PBT, incubated twice with 100 μl of Dilution buffer (Roche) for 5 min and treated with 50 μl of labelling solution (in situ cell death detection TMR Red kit, Roche) for 30 min at 37 °C. Next, 5 μl of terminal transferase (Roche) was added and the reaction was allowed to proceed for 2 h at 37 °C, stopped with PBT, washed three times with PBT and the discs were mounted in Vectashield (Vector Laboratories).

BrdU (bromodeoxyuridine) staining was performed using imaginal discs dissected in cold PBS, incubated in 0.01 mM BrdU for 15 min at 37 °C, washed three times with PBS, fixed for 2 min with Carnoy (3:1-ethanol:acetic acid) and washed four times for 5 min in PBS. Discs were then treated with 2 M HCl for 10 min, washed three times with PBS for 10 min and incubated with the primary antibody at 4 °C (BrdU labelling and detection Kit I-Roche). The anti-BrdU antibody staining was performed according to the protocol of the supplier.

Acridine Orange staining was performed following the procedure of Abrams 42. Embryos were dechorionated with bleach, rinsed with water and incubated in 500 μl of heptane and 500 μl of 5 mg/ml of Acridine Orange (Sigma). After 5 min of shaking, embryos were removed from the interface and mounted in Halocarbon oil (Merck).

Expression constructs

Procedures for the cloning and generation of transgenic flies containing the p[UAS-dRYBP] and p[UAS-dRYBP-ΔZF] proteins have been described previously 17. The p[UAS-RYBPmurine] construct was generated by cloning the NotI/XbaI fragment from the murine RYBP cDNA (accession number AF101779) 43 into the p[UAST] vector 44. Likewise, p[UAS-PC] was generated by cloning the PC cDNA (NM 079475) EcoRI fragment into the p[UAST] vector 44. Transgenic flies were obtained by standard procedures using Df (1) w67c23 flies (y-, w-) as hosts.

Chromatin immunoprecipitation

ChIP assays using whole embryos were performed as described previously 45. Chromatin was immunoprecipitated using a 1:20 dilution of the anti-dRYBP antibody 18. For quantitative ChIP, following immunoprecipitation and DNA purification, enrichment for specific DNA fragments was analyzed by real-time PCR, using the Roche Light Cycler instrument and accessories as described in Negre et al. 45. Data are expressed as the percentage of input chromatin precipitated for each region examined. The mean values and the standard deviations of two independent ChIP experiments are shown. As a negative control the robo3 (roundabout homolog 3) (http://flybase.org) gene region was included in the RT-PCR experiment. The following primers were used. For reaper CDS: rpr CDS 5′-AGG CGA CTC TGT TGC GGG AG-3′ (loc. 3L: 18391288-18391307; Flybase release 5.10) and rpr CDS 5′-TGC GAT GGC TTG CGA TAT TTG-3′ (loc. 3L: 18391142-18391162; Flybase release 5.10). For reaper promoter: rpr promoter 5′-AAA AAC ACG CTT GGC AAC AG-3′ (loc. 3L: 18391768-18391787; Flybase release 5.10) and rpr promoter 5′-GCT ATT TAT ACC TGG TTC TCT CAC G-3′ (loc. 3L: 18391573-18391597; Flybase release 5.10). For robo3: 5′-ATA GCC TCA ACA CTG AGG AAG G-3′ (loc. 2L: 1285910-1285931; Flybase release 5.10) and 3′-GTA GGG TTT GAT TAA CCG GAC C-5′ (loc. 2L: 1286083-1286104; Flybase release 5.10). PCR reactions were performed following Comet et al. 46.