Temporal Dynamics of Apoptosis-Induced Proliferation in Pupal Wing Development: Implications for Regenerative Ability

The ability of animals to regenerate damaged tissue is a complex process that involves various cellular mechanisms. As animals age, they lose their regenerative abilities, making it essential to understand the underlying mechanisms that limit regenerative ability during aging. Drosophila melanogaster wing imaginal discs are epithelial structures that can regenerate after tissue injury. While significant research has focused on investigating regenerative responses during larval stages, particularly regarding the regulation and function of the JNK pathway, our comprehension of the regenerative potential of pupal wings and the underlying mechanisms contributing to the decline of regenerative responses remains limited. This study explores the temporal dynamics during pupal development of the proliferative response triggered by the induction of cell death, a typical regenerative response. Our results indicate that the apoptosis-induced proliferation response can be initiated as late as 30 hours after pupa formation (APF), when in normal circumstances cell proliferation ceases at around 20 hours APF. Furthermore, our data revealed that after 35 hours APF, cell death alone fails to induce further proliferation. Interestingly, the failure of reinitiating the cell cycle beyond this time point is not attributed to an incapacity to activate the JNK pathway. Instead, one of the constraining factors in the apoptotic-induced proliferation process during pupal development seems to be the activity level of ecdysone-responsive genes. Author Summary Animals have the remarkable ability to regenerate damaged tissues, but this regenerative potential diminishes with age. Understanding the mechanisms underlying age-related decline in regenerative abilities is crucial. Drosophila melanogaster wing imaginal discs provide a valuable model for studying tissue regeneration. While significant research has focused on regenerative responses during larval stages, our understanding of the regenerative potential and mechanisms in pupal wings remains limited. In this study, we investigate the temporal dynamics of the proliferative response triggered by cell death during late during the development, in pupal development. Our findings reveal that the apoptosis-induced proliferation response can occur during pupal development, even after normal cell proliferation has ceased. However, at late stages of pupal development this response does not occur. We have found that, the inability to reinitiate the cell cycle beyond this time point is influenced by the activity of the hormone ecdysone and its-responsive genes. These findings shed light on the dynamic processes involved in tissue regeneration during pupal development. This study expands our understanding of the complex interplay between cell death, proliferation, and gene activity during tissue regeneration, providing valuable insights for future research in regenerative biology.


Abstract
The ability of animals to regenerate damaged tissue is a complex process that involves various cellular mechanisms. As animals age, they lose their regenerative abilities, making it essential to understand the underlying mechanisms that limit regenerative ability during aging. Drosophila melanogaster wing imaginal discs are epithelial structures that can regenerate after tissue injury. While significant research has focused on investigating regenerative responses during larval stages, particularly regarding the regulation and function of the JNK pathway, our comprehension of the regenerative potential of pupal wings and the underlying mechanisms contributing to the decline of regenerative responses remains limited. This study explores the temporal dynamics during pupal development of the proliferative response triggered by the induction of cell death, a typical regenerative response. Our results indicate that the apoptosis-induced proliferation response can be initiated as late as 30 hours after pupa formation (APF), when in normal circumstances cell proliferation ceases at around 20 hours APF. Furthermore, our data revealed that after 35 hours APF, cell death alone fails to induce further proliferation. Interestingly, the failure of reinitiating the cell cycle beyond this time point is not attributed to an incapacity to activate the JNK pathway.
Instead, one of the constraining factors in the apoptotic-induced proliferation process during pupal development seems to be the activity level of ecdysoneresponsive genes.

Introduction
Regeneration is a remarkable ability present across the animal kingdom that enables multicellular organisms to repair damaged tissues and maintain tissue homeostasis (Tanaka and Reddien, 2011;Brockes and Kumar, 2008). This intricate process involves various cellular mechanisms, including regenerative growth. As animals age, they progressively lose their regenerative abilities, including some salamanders with boundless regenerative capacities.
Understanding the underlying mechanisms that limit regenerative ability during aging is a crucial developmental biology question.
During development, intrinsic cellular and physiological changes occur that limit the ability of signaling pathways to promote cellular plasticity and induce regenerative growth, which are crucial to support regeneration (McCusker and Gardiner, 2011;Seifert and Voss, 2013). For instance, differentiated cells lose the ability to re-enter the cell cycle, a process necessary for limb regeneration in vertebrates. Tumor suppressor proteins like the Retinoblastoma protein (Rb) likely play a crucial role in maintaining plastic cellular states and cell cycle reentry. The levels of this factor vary during regeneration in salamanders and across developmental stages in mammals (Seifert and Voss, 2013). However, the loss of the ability to regenerate limb buds can be delayed experimentally in Xenopus laevis, indicating that the loss of regenerative ability during aging may also be postponed or reversed.
These sac-like structures have the remarkable ability to regenerate during larval stages but lose this ability at the end of the larval stage or during pupal development, which coincides with the cessation of cell proliferation (Smith-Bolton et al., 2009;Diaz-Garcia and Baonza, 2013;Buttitta et al., 2007).
During pupal development, cells exit the cell cycle as a result of a decrease in Cyclin E (CycE) and an increase in Rb factor activity (Buttitta et al., 2007). Rb represses the transcription of genes required for the G1-S transition, including CycE, by binding and inhibiting the transcription factor E2F1 (Du et al., 1996a;Du et al., 1996b). CycE, in turn, promotes the G1-S transition by phosphorylating and inhibiting Rb. Although cell cycle exit occurs around 24 hours APF, the positive feedback loop between CycE and E2F1 is maintained until around 30-35 hours APF. Therefore, over-expression of either CycE or E2F1 between 24-35 hours after pupa formation (APF) is sufficient to induce proliferation. However, after 30-35 hours, co-expression of CycE and E2F1 is necessary to induce proliferation (Buttitta et al., 2007). The positive feedback between CycE and E2F1 ends around the onset of epigenetic shutdown of regulatory regions of key genes involved in cell cycle control, such as CycE, E2F1, and string (Buttitta et al., 2007). Recent research has revealed that ecdysone-responsive transcription factors regulate the temporal changes in chromatin accessibility during wing disc development (Uyehara et al., 2017).
The c-Jun N-terminal kinase (JNK) signalling pathway has emerged as a critical signal in the process of regeneration. Upon injury, the JNK signalling pathway is initiated at the wound site (Bosch et al., 2005). This pathway plays a key role in regulating several biological processes associated with regeneration (Bosch et al., 2005;Bergantinos et al., 2010a;Bogoyevitch et al., 2010) (Lee et al., 2005) (Chen, 2012). In studies of disc regeneration, inhibition of JNK function was found to negatively affect wound healing and reduce regenerative proliferation (Bosch et al., 2005;Mattila et al., 2005;Ramet et al., 2002). During regeneration, JNK signalling promotes the activation of several downstream pathways, including JAK/STAT and Wingless (Wg) (Harris et al., 2016;Katsuyama et al., 2015;Smith-Bolton et al., 2009;Pastor-Pareja et al., 2008).
Despite extensive research on regenerative responses during larval stages, our understanding of the regenerative abilities of pupal wings and the mechanisms involved in the decline of regenerative responses remains limited. It is unclear whether the signals activated by tissue damage during larval stages also operate during pupal development.
In this study, we investigated the proliferative response triggered by the induction of cell death during pupal development. Our findings indicate that apoptosis-induced proliferation (AiP) response can be triggered up to 30 hours APF. As cell proliferation normally ceases at 20-24 hours APF during normal development, these results suggest that cell death can extend the proliferative phase of cells in the wing discs during pupal development. We found that after 35 hours APF cell death is not sufficient to induce proliferation. Our data suggest that the inability to reactivate the cell cycle after this time point is not due to an inability to trigger the JNK pathway. Rather, one of the limiting factors in the apoptosis-induced proliferation process during pupal development appears to be the activity of ecdysone-responsive genes.

Ionizing Irradiation (IR) can trigger an apoptosis-induced proliferation response up to 30-35 hrs APF
During the larval stage, the cells in the wing disc proliferate asynchronously.
However, once the larva-to-pupa transition occurs (0 hours APF), the cells in the wing disc undergo a G2 phase arrest, which stops their proliferation. By 6 hours APF, the majority of cells in the pupal wing have entered this G2 phase arrest. Between 12-18 hours APF, a fraction of these cells resumes the cell cycle and undergo the last round of division before being arrested in G1. By 24 hours APF, approximately 95% of the wing disc cells are arrested in G1 phase and become quiescent before initiating terminal differentiation (Milán et al., 1996b;Buttitta et al., 2007;Schubiger and Palka, 1987;Guo et al., 2016).
We analyse the proliferative response of wing disc cells following the induction of apoptosis during pupal development. To induce cell death, we employed Xray irradiation on pupae at three different time intervals: 0-4, 10-14, and 20-24 hours APF. Subsequently, we dissected the pupae 20 hours after the irradiation process ( Fig. 1A). To assess the proliferative response, we conducted an analysis of the mitotic marker Phospho-Histone 3 (PH3) and determined the mitotic index of wing discs in the pupal stage (see methods).
In wing discs examined at 20-24 hours (APF), we observed the presence of mitotic cells in both irradiated and non-irradiated pupae. However, the irradiated pupal wings had a higher mitotic index ( Fig. 1B-D'). Surprisingly, we also observed a relatively high number of mitotic cells in irradiated pupal wings examined at 30-34 hours APF, even though there were no dividing cells in control discs at that time point (Fig. 1 E-F'). At 40-44 hours APF, neither the control nor the irradiated discs showed any mitotic cells ( Fig. 1 G-H). These findings suggest that irradiation-induced apoptosis may prolong the proliferative period up to 30-34 hours APF.
IR-induced apoptosis depends on the proliferative status of the cells, with proliferating cells being more sensitive to irradiation than differentiating cells (Ruiz-Losada et al., 2022;. Therefore, the lack of a proliferative response at 40-44 hours APF upon IR might be due to cells no longer being sensitive to irradiation at this stage. To investigate this possibility, we examined radiation-induced apoptosis in pupal wing irradiated at 0-4 hours and 20-24 hours APF and examined 20 h later (20-24h and 40-44h APF, respectively). Our study revealed a notable increase in apoptosis in irradiated pupal wing cells at 20-24 hours APF (irradiated at 0-4 hours APF), compared to control discs, as indicated by the apoptotic marker Dcp-1 (see Fig. S1).
However, we did not detect apoptotic cells in wing discs at 40-44 h APF (irradiated at 20-24 hours APF Fig. S1). These findings suggest that wing disc cells at 40 hours APF are insensitive to irradiation-induced apoptosis.
To further pinpoint the time at which pupal disc cells become insensitive to irradiation, we selected pupae at 2-hour intervals, starting at 20 hours APF ( Fig Therefore, these results suggest that the cells of pupal wing become radioresistant around 26 h APF concomitant to their definitive cell cycle arrest.

Genetic ablation expands the period of damage-induced proliferative response during pupal development
Our results indicate that the cells in the pupal wing exhibit insensitivity to IRinduced apoptosis after 26 hours (APF). Consequently, it is plausible to consider that the absence of observed proliferative responses in irradiated pupae during the later stages of pupal development may be attributed to the ineffectiveness of IR in inducing apoptosis. To test this, we employed an alternative approach to effectively induce apoptosis. To this end we used the Gal4/UAS binary system with Gal80 ts to induce genetic ablation at different times APF. We overexpressed the pro-apoptotic gene reaper ( could not be attributed to the lack of apoptosis. Therefore, our findings suggest that apoptosis triggers a proliferative response that lasts up to 30-34 hours APF, after which it declines. The mitotic signals produced in the damaged area may extend the proliferative period of pupal wing cells, but only until 30-34 hours (APF).
In summary, our study provides evidence that the duration of the proliferative response in pupal wing cells is limited and can be influenced by proliferative signals produce by dead cells.

JNK signalling is activated in response to damage in late stages of pupal development
The JNK pathway plays a crucial role in triggering apoptosis-induced proliferative response during the larval stages (Bosch et al., 2005;Mattila et al., 2005;Ramet et al., 2002). To investigate the evolution of JNK pathway activity in response to cell death induction after metamorphosis, we examined the expression of the TRE-GFP reporter (Chatterjee and Bohmann, 2012) following the over-expression of rpr under the control of hh-Gal4 at different times APF.
We selected pupae at 0 hours APF and raised them at 17°C until they reached the equivalent of 10-15 h APF at 25°C. Next, we induced ectopic expression of Our experiments using IR to induce apoptosis (Fig. S2) yielded similar results.
We observed JNK activation in response to irradiation, regardless of whether it occurred during early stages, when cell proliferation is activated, or late stages, when it is not. These findings suggest that the loss of proliferative response during late stages of pupal development is not caused by an inability of cells to activate JNK. To test this idea, we overexpressed a constitutively activated form of the JNK-kinase Hemipterous (hep CA ) in late stages of pupal development.
Our results showed that hep CA overexpression induced a proliferative response only until 30-35 hours AFP, and not at later stages (Fig. 5).

The over-expression of CycE or E2F1 in damaged pupal wing is not sufficient to promote a proliferative response after 35-40 hrs APF
Previous studies have shown that the overexpression at pupal stages of key regulators of the G1-S transition, such as CycE, CycE/Cdk2, CycD/Cdk4 or E2F1 maintains cell division until 30 h APF, but not later (Buttitta et al., 2007). This is due to a positive feedback mechanism between CycE and E2F1 that is active during larval and early pupal development, but ceases after 30 hrs APF (Buttitta et al., 2007). Similarly, our observations indicate that apoptosis can trigger proliferation up to 30-35 hours AFP, but not beyond. Thus, it is plausible that the mitogenic signals generated by apoptotic cells specifically activate CycE/Cdk2 or E2F1, thereby triggering a proliferative response that is sustained only until approximately 30-35 hours APF. This time frame coincides with the period when the feedback mechanism between CycE and E2F1 is still functioning. To test this hypothesis, we utilized the double transcriptional transactivator system sal E/Pv -LHG/lexOp in combination with Gal4/UAS to overexpress E2F1 dp or CycE while inducing apoptosis. The sal E/Pv -LHG transgene contains a Gal80 suppressible form of LexA (Yagi et al., 2010). To conditionally express both the sal E/Pv -LHG/lexOp and Gal4/UAS binary systems, we used tub-Gal80 ts . We induced apoptosis in the sal E/Pv -LHG domain of the wing disc by using lexOp-rpr, while simultaneously overexpressing UAS-CycE or UAS-E2F1 in the overlapping anterior compartment with the cubitus interruptus (ci)-Gal4 line (Fig. 6A).
We selected pupae between 0-5 hours old APF and maintained them at 17ºC until they reached the developmental stage equivalent to 25-30 hours APF at 25ºC. Following this, we shift them to 29ºC to trigger ectopic expression of UAS-CycE (ci-Gal4, UAS-CycE), while simultaneously inducing apoptosis (sal E/Pv -LHG/lexOp-rpr). The discs were analysed at the developmental stage equivalent to 35-40 hours APF or beyond 40 hours APF. Our observations revealed a high number of mitotic cells in the anterior compartment of discs expressing CycE, or co-expressing rpr and CycE, at the developmental stages equivalent to 35-40 hours and 40-45 hours AFP (as illustrated in Fig. 6).
However, at 45 hours APF, we did not observe any mitotic cells in either the CycE-expressing discs or the discs co-expressing CycE and rpr. We obtained similar results when we induced apoptosis and overexpressed E2F1 simultaneously (Fig. S4).
All together these results suggest that damage can only induce a proliferative response while the positive feedback mechanism between CycE and E2F1 operates.

Ecdysone-responsive transcription factor E93 blocks the proliferative response during pupal development
The end of the positive feedback loop between CycE and E2F, as well as the ending of the proliferative response following damage, coincide with the initiation of epigenetic silencing of the regulatory regions of critical genes involved in cell cycle regulation, including CycE, E2F1, and string (stg) (Uyehara et al., 2017;Guo et al., 2016;Ma, 2019). In Drosophila, developmental transitions are regulated by the hormone ecdysone, and ecdysone-responsive transcription factors control temporal changes in chromatin accessibility during wing disc development (Uyehara et al., 2017 ;Guo et al., 2016;Ma, 2019). Specifically, the E93 transcription factor is transcriptionally activated at 18 hours and 24 hours APF (Uyehara et al., 2017).
Loss-of-function mutations in the E93 gene results in chromatin accessibility changes at several genome regions in 24 h and 44 h APF wing discs (Uyehara et al., 2017). Importantly, the progressively closed chromatin status observed at regulatory regions of the CycE, E2F1 and stg genes between third instar and 44 hours APF is attenuated in E93 mutants (Uyehara et al., 2017) (Fig S5). Given that this temporal window coincides with the end of apoptosis-induced proliferation during pupal development, it is plausible that E93, by blocking the chromatin accessibility at these cell cycle regulators, prevents the induction of this response. To explore this idea, we have irradiated E93 mutant larvae at 20 hours APF and examined the proliferative response 20 hours later (40 hours APF). However, in non-irradiated E93 mutant pupal wings, we still observed some PH3 positive cells, confirming the role of E93 in suppressing the proliferative capability of wing disc cells at this specific time point (Fig. 7A and   B). This finding highlights the importance of E93 in regulating cell proliferation during pupal wing development ( Fig. 7A and B). Our results strongly suggest that E93, through its involvement in the ecdysone response, acts as a key orchestrator of the intricate cellular processes that regulate and limit the proliferative response of pupal cells towards the end of their development.

Discussion
Regeneration is a complex process that relies on multiple cellular and molecular mechanisms working in concert to uphold tissue homeostasis and effectively respond to external challenges. One critical aspect of this process is the induction of regenerative growth, which is often lost during development or aging. In Drosophila development, multiple studies have shown that cell proliferation is one of the primary responses during discs regeneration in larval stages (Adler and Macqueen, 1984;Dale and Bownes, 1981;Kiehle and Schubiger, 1985;Bryant and Fraser, 1988;Sustar and Schubiger, 2005;Bosch et al., 2008;Worley and Hariharan, 2022). When damage occurs, apoptotic cells generated by the insult can initiate a process known as Apoptosis-induced proliferation (Perez-Garijo and Steller, 2015; Ryoo HD, 2012). Different studies using third instar larval imaginal discs have suggested that this ability is lost at the end of larval stage or during pupal development (Smith-Bolton et al., 2009;Diaz-Garcia and Baonza, 2013). However, our work shows that damage induction in the wing disc produces a proliferative response that is maintained up to 30 hours APF.
Previous studies have demonstrated that overexpression of key regulators of the G1-S transition, such as CycE, CycE/Cdk2, CycD/Cdk4, or E2F1, in pupal stages can maintain cell division until 30 hours APF. However, overexpression of these regulators after that time is not sufficient to activate cell proliferation. To keep cells proliferating until at least 40-44 hours APF, ectopic activation of both CycE and E2F1 is necessary. It has been proposed that this is due to the existence of a positive feedback mechanism between CycE and E2F that is active during larval and early pupal development, but finishes after 30 hours APF (Buttitta et al., 2007;Guo et al., 2016;Ma, 2019). Our findings reveal that induction of apoptosis in wing imaginal discs before 24 hours APF induces cell proliferation up to 30 hours APF, which coincides with the end of the positive feedback between CycE and E2F1. This suggests that mitogenic signals released by apoptotic cells would specifically act on CycE or E2F1, stimulating cell division during the active positive feedback mechanism. However, these signals alone are insufficient to counteract the signals that drive cell cycle exit at the conclusion of the proliferative phase.

Hormones block apoptosis-induce proliferation
In Drosophila developmental transitions are regulated by the hormone Ecdysone. The signals induce by this hormone are mediated by the Ecdysoneresponsive transcription factors, that are involved in regulating temporal changes in chromatin accessibility that occur throughout wing disc development (Uyehara et al., 2017;Guo et al., 2016;Ma, 2019). Specifically, the E93 transcription factor is expressed at 18 hours and 24 hours PFD APF (Uyehara et al., 2017) During pupal development, the induction of cell death triggers the activation of JNK pathway, both in early (Bosch et al., 2005) and late stages, even when apoptosis-induced proliferation is not observed. Although the JNK pathway is known to suppress the Polycomb group of proteins (Lee et al., 2005), our results suggest that its activation in pupal discs is not enough to overcome epigenetic silencing of key cell cycle regulators.
Our data provide evidence supporting a model in which ecdysone-mediated developmental programming induces specific alterations in chromatin accessibility of crucial cell cycle regulator genes (Uyehara et al., 2017;Guo et al., 2016;Ma, 2019). These alterations, in turn, govern the process of cell cycle exit and subsequently suppress the proliferative response that is normally activated in response to damage (Fig 7 E). This model is likely to be applicable to other organisms, such as urodeles, zebrafish, and mice. For instance, Lin-28, an RNA-binding protein that regulates Let-7, has been shown to inhibit the expression of thyroid hormone target genes, delay development, and prolong regenerative potential after damage in urodeles (Faunes et al., 2017). Similarly, inhibiting thyroid hormone signaling at the level of its synthesis and at the receptor level in neonatal mice rescues the proliferative capacity of cardiomyocytes and the regenerative potential of the adult heart (Hirose et al., 2019). Conversely, exogenous administration of thyroid hormones in zebrafish inhibits the regenerative capacity of the heart and caudal fin (Sun et al., 2019).
The result presented in this work highlights the intricate interplay between developmental cues, chromatin modifications, and the regulation of cell cycle dynamics, shedding light on the mechanisms underlying the fine-tuned control of cellular proliferation and regeneration

Fly strain
The following strains were used in this study. Unless otherwise was indicated, strain descriptions can be found at http://flybase.bio.indiana.edu.

Protocol for Genetic ablation and irradiation experiments
The development of pupae is observed to be 2.5 times slower at 17˚C compared to 25˚C and 1.3 times faster at 29˚C compared to 25˚C, as per our experimental conditions. To ensure equivalent stages of development, all time points for pupae grown at 17˚C were adjusted to match the stages at 25˚C.

Irradiation
Pupae were given a dose of 4000 R using Philips-MG-102 irradiation unit.

Immunocytochemistry
Immunostaining of the wing discs was performed according to standard protocols. The following primary antibodies were used: rabbit anti-Phospho-      analysed. The discs were stained with anti-PH3 antibody (in Green A-C, and grey A'-C') and Phalloidin to reveal F-Actin (in red A-C ). (E) A proposed model for the dynamic response to damage during pupal development. Statistical analysis was conducted using One-way ANOVA Tukey's test ,**** p<0.0001.