(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Mutual repression between JNK/AP-1 and JAK/STAT stratifies senescent and proliferative cell behaviors during tissue regeneration [1]

['Janhvi Jaiswal', 'Hilde-Mangold-Haus', 'University Of Freiburg', 'Freiburg', 'Spemann Graduate School Of Biology', 'Medicine', 'Sgbm', 'Janine Egert', 'Institute Of Medical Biometry', 'Statistics']

Date: 2023-06

Epithelial repair relies on the activation of stress signaling pathways to coordinate tissue repair. Their deregulation is implicated in chronic wound and cancer pathologies. Using TNF-α/Eiger-mediated inflammatory damage to Drosophila imaginal discs, we investigate how spatial patterns of signaling pathways and repair behaviors arise. We find that Eiger expression, which drives JNK/AP-1 signaling, transiently arrests proliferation of cells in the wound center and is associated with activation of a senescence program. This includes production of the mitogenic ligands of the Upd family, which allows JNK/AP-1-signaling cells to act as paracrine organizers of regeneration. Surprisingly, JNK/AP-1 cell-autonomously suppress activation of Upd signaling via Ptp61F and Socs36E, both negative regulators of JAK/STAT signaling. As mitogenic JAK/STAT signaling is suppressed in JNK/AP-1-signaling cells at the center of tissue damage, compensatory proliferation occurs by paracrine activation of JAK/STAT in the wound periphery. Mathematical modelling suggests that cell-autonomous mutual repression between JNK/AP-1 and JAK/STAT is at the core of a regulatory network essential to spatially separate JNK/AP-1 and JAK/STAT signaling into bistable spatial domains associated with distinct cellular tasks. Such spatial stratification is essential for proper tissue repair, as coactivation of JNK/AP-1 and JAK/STAT in the same cells creates conflicting signals for cell cycle progression, leading to excess apoptosis of senescently stalled JNK/AP-1-signaling cells that organize the spatial field. Finally, we demonstrate that bistable separation of JNK/AP-1 and JAK/STAT drives bistable separation of senescent signaling and proliferative behaviors not only upon tissue damage, but also in Ras V12 , scrib tumors. Revealing this previously uncharacterized regulatory network between JNK/AP-1, JAK/STAT, and associated cell behaviors has important implications for our conceptual understanding of tissue repair, chronic wound pathologies, and tumor microenvironments.

Funding: Funding for this work was provided by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany ́s Excellence Strategy (CIBSS – EXC-2189), the Emmy Noether Programme (CL490/1-1), the Heisenberg Program (CL490/3-1, CL490/4-1), the SFB850 / A08, and by the Boehringer Ingelheim Foundation (BIF Plus3) to AKC. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

A stress-induced cell cycle arrest is often associated with senescence. Senescence was originally defined as the irreversible cessation of cell proliferation due to age, cellular damage, or oncogenic signaling and was suggested to drive age-dependent decline in regenerative capacity [ 52 – 56 ]. Senescent cells often exhibit a complex senescence-associated secretory phenotype (SASP) [ 2 , 57 – 59 ], which is linked to persistent production of inflammatory paracrine molecules and secretion of ECM degrading enzymes. However, senescent cells also exhibit up-regulation of autophagy, unfolded protein response (UPR), and ROS [ 2 , 57 – 59 ]. Importantly, senescent cells remain viable for long periods of time, indicating that they are resistant to apoptosis, despite extensive cellular damage signals. However, recent studies reveal that the senescence program may also act transiently during wound healing of the mouse epidermis, where it may promote wound closure and cell plasticity [ 60 – 62 ]. Our previous work also started to suggest that G2-arrested cells may display senescent features that allow them to act as a central driver of physiological and pathological wound healing processes and contribute to the oncogenic potential of the tumor microenvironment [ 29 ]. Yet the precise features of this cell population are not described and competition between senescent cell cycle stalling and regenerative proliferation was not explored.

At the level of a single cell, these paradoxical cell behaviors are integrated by cell cycle progression. Using Drosophila wing imaginal discs ( S1A Fig ) , we recently reported that high JNK/AP-1 signaling in wounds and tumors induces a cell-autonomous arrest or stalling of cells in the G2-phase of the cell cycle. G2-stalling is mediated by the JNK-dependent up-regulation of the G2/M pseudokinase tribbles (trbl) and the down-regulation of the G2/M phosphatase string (stg/cdc25). Knock-down of tribbles or ectopic expression of string is sufficient to suppress G2-stalling and forces these cells into G1. This causes a substantial increase in apoptosis, which suggests that the G2-arrest protects high JNK-signaling cells from cell death and that reentering the cell cycle in wound environments is associated with cell lethality [ 29 ]. This model is supported by the idea that wounds produce reactive oxygen species (ROS); which can exacerbate cellular damage and cell death [ 26 , 35 ]. However, the molecular damage sensor p53 is specifically activated by the G2/M kinase Cdk1 to induce competence for damage-driven apoptosis only upon exit from G2 ( S1A’ Fig ) [ 36 ]. Importantly, JNK/AP-1 signaling cells also produce mitogens belonging to the Unpaired (Upd) family [ 37 – 41 ], which activate the pro-proliferative JAK/STAT pathway essential for compensatory proliferation [ 16 , 18 , 26 , 42 – 51 ]. Paradoxically, this reveals that a JNK-signaling cell population exists at the center of wounds, which produces pro-proliferative signals but itself is prohibited from proliferation. These findings highlight the necessity for mechanisms, which ensure that high JNK/AP-1-signaling cells do not cycle despite being in the presence of mitogenic signals and that JAK/STAT-driven compensatory proliferation occurs only in cells that do not experience active JNK/AP-1-signaling. The regulatory network and molecular effectors that resolve this paradox have not been described. As JNK/AP-1 and JAK/STAT drive contradictory yet critical cell behaviors, characterizing their regulatory interactions is essential to understand how these pathways cooperatively organize tissue stress responses.

Upon tissue damage, wound-derived factors initiate signaling pathways which drive cellular responses like apoptosis, proliferation, survival and tissue remodeling [ 1 – 4 ]. While these responses are essential for tissue repair, it is critical that they remain spatio-temporally restricted to avoid pathological consequences, such as the establishment of chronic wounds [ 5 – 7 ]. Chronic wounds are characterized by sustained inflammation, deregulated proliferation, and apoptosis, which are also hallmarks of tumors [ 8 – 11 ]. Different experimental models have identified the signaling pathways that control wound repair and tissue regeneration, many of which also drive diseases like cancer. Specifically, the JNK/AP-1 and the JAK/STAT pathways are consistently implicated in regeneration [ 12 – 15 ] and tumor growth [ 16 – 21 ]. In Drosophila, JNK/AP-1 is one of the earliest pathways activated in response to damage and indispensable for wound healing and regeneration [ 22 – 26 ]. In this role, JNK/AP-1 regulates a variety of conflicting cell behaviors including apoptosis [ 27 , 28 ], survival [ 29 ], and compensatory proliferation [ 30 , 31 ]. These paradoxical behaviors have been extensively characterized individually [ 32 – 34 ]. Yet, how they are spatially organized on a larger tissue scale to ensure regeneration is less well understood.

Results

JNK/AP-1 and JAK/STAT separate into distinct spatial domains with distinct proliferative potential Expression of egr induces upd1-3, the pro-mitogenic cytokines for the JAK/STAT pathway (S2 Fig; [72]) [18,26,48,50]. Yet, high JNK-signaling cells in egr-expressing discs that up-regulated a transcriptional upd3 reporter did not exhibit JAK/STAT activity, as assessed by the 10xStat92E-dGFP reporter (Fig 3A–3G) [73,74]. This is in stark contrast to wing disc development where ligand expression and pathway activation patterns largely coincide and suggests that a specific regulatory network controls JAK/STAT activity upon tissue damage [74–77]. Consistent with paracrine activity of upd1-3 [78,79], JAK/STAT signaling was broadly induced in the pouch periphery and hinge of egr-expressing discs (Fig 3E). Importantly, as proliferation was inhibited in the JNK-signaling domain at the wound center, proliferation was restricted to domains of JAK/STAT activation in the wound periphery (Fig 3D–3G). These observations highlight a pronounced spatial separation of signaling pathways and regenerative cell behaviors linked to JNK-induced tissue damage responses. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 3. JNK/AP-1 and JAK/STAT separate into distinct spatial domains with different proliferative potentials. (A) egr-expressing disc also expressing the upd3.1-lacZ reporter (gray) and the JNK/AP-1 reporter TRE-RFP (magenta). Yellow arrow illustrates the principle of line traces used to generate graphs in (C, F). (B) Schematic representation of JNK/AP1 activity (magenta) after 24 h of egr-expression (R0). Arrow illustrates the vector of line traces used to generate graphs in (C, F). (C) Fluorescence intensity profiles for upd3.1-lacZ (gray) and TRE-RFP (magenta) reporters were traced along the axis from the pouch center (PC) to the disc periphery (DP, hinge) (arrows in A, B). Graph represents mean ± SEM of reporter fluorescence intensity values from n = 15 tracks across a representative disc, scaled to the maximum measured value. A total of n = 27, control and n = 33, egr-expressing discs were evaluated from N = 3 independent experiments. (D, E) A control (D) and egr-expressing disc (E) also expressing the TRE-RFP (magenta) and Stat92E-dGFP (green) reporters and assessed for S-phase activity using EdU incorporation assays. A total of n = 15, control and n = 19, egr-expressing discs were evaluated from N = 3 independent experiments. (F) Fluorescence intensity profiles for TRE-RFP (magenta), Stat92E-dGFP (green) reporters, and EdU (yellow) intensities traced from the pouch center (PC) to the disc periphery (DP) (black arrow in B). Graph represents mean ± SEM of reporter fluorescence intensity values from n = 15 tracks across a representative disc, scaled to the maximum measured value. (G) A smoothening function (see Materials and methods) applied to observed reporter patterns in (F) gives rise to a simple bistable pattern. Domain with dead cells was excluded from the graph. (H) Expression of egr using the en-GAL4 driver. The domain of expression in the posterior compartment was tracked using UAS-RFP, and efficiency of JNK-activation was assessed by staining for the JNK-target MMP-1 (magenta). Discs activate the JAK/STAT Stat92E-dGFP reporter (green) nonautonomously but not in the JNK-signaling domain. A total of n = 17, control and n = 17, egr-expressing discs were evaluated from N = 2 independent experiments. (I-K) A time-course analysis of TRE-RFP (magenta) and Stat92E-dGFP (green) reporter activity in egr-expressing discs after 7 h (I), 14 h (J), and 24 h (K) of expression. A total of n = 21, 7 h; n = 32, 14 h; and n = 31, 24 h control and n = 25, 7 h; n = 20, 14 h; and n = 25, 24 h egr-expressing discs were evaluated from N = 2 independent experiments. (L-O) Quantification of the fluorescence intensity of the JNK-reporter TRE-RFP (L, M) and the JAK/STAT reporter Stat92E-dGFP (N, O) within the JNK/AP1 signaling domain of the pouch (“pouch”) and in a 20-μm band outside the JNK/AP1-signaling domain (“periphery”) in egr-expressing disc (see also S3W Fig) after 7 h, 14 h, and 24 h of inductive temperature shift to 30°C. Graphs displays mean ± SEM for n = 7, 7 h egr-expressing discs; n = 7, 14 h egr-expressing discs; and n = 14, 24 h egr-expressing discs. One-way ANOVA with Holm–Sidak’s multiple comparisons test (periphery STAT) or a Brown–Forsythe and Welch ANOVA with Dunnett’s T3 multiple comparisons test (other data sets) was performed to test for statistical significance. Source data for quantifications provided in S1 File. Maximum projections of multiple confocal sections are shown in D, E, and H. Discs were stained with DAPI (magenta) to visualize nuclei. Scale bars: 50 μm. https://doi.org/10.1371/journal.pbio.3001665.g003 To further support this conclusion, we carried out a series of control experiments. We first wanted to rule out the possibility that the observed JAK/STAT activation pattern in the periphery of egr-expressing discs perpetuated from early hinge patterns of developing discs due to a Dilp8-induced developmental delay (S3A and S3B Fig) [80]. Hence, we monitored temporal changes in JAK/STAT activity via a time course of 0, 7, and 14 h of egr-expression under the control of rn-GAL4. Compared to developmental levels, we consistently observed ectopic induction of the JAK/STAT reporter in the hinge confirming that JAK/STAT activity is induced de novo in response to egr-expression (S3A–S3M Fig). We also wanted to test if the same spatial separation of JNK/AP-1 and JAK/STAT activity can be induced by different patterns of egr-expression. We thus expressed egr in the posterior compartment of the wing disc using en-GAL4 or in the hinge domain using 30A-GAL4. To track the ability of these drivers to express egr and thus activate JNK, we monitored expression of the JNK/AP-1 target gene MMP-1 [81]. In both instances, cells that were expressing MMP-1 did not activate JAK/STAT signaling but JAK/STAT signaling was activated in the neighboring domains (Figs 3H, S3N, and S3O). These results confirm that JAK/STAT activity can be induced nonautonomously and de novo in the hinge and pouch upon tissue damage. Yet, they also strongly indicate that cell-autonomous repression of JAK/STAT signaling may be an important consequence of JNK/AP-1 activation. This is consistent with recent observations of JNK/AP-1 and JAK/STAT separation in other contexts, such as epidermal wounds [12]. To gain more insight into the dynamics of spatial separation, we analyzed the temporal evolution of JNK/AP-1 and JAK/STAT activity in the JNK-signaling domain of control and egr-expressing discs. This analysis revealed that JAK/STAT was mildly elevated in the hinge (nonautonomously) and pouch (autonomously) after 7 h of egr-expression and even more by 14 h. At this point, many egr-expressing pouch cells displayed activation of either JNK/AP-1 or JAK/STAT reporter, or coexistence of both pathways. After 24 h of sustained egr-expression and high JNK/AP-1 activity, JAK/STAT activity remained high in the hinge periphery (nonautonomous activity) but had largely disappeared from the egr-expressing, JNK-signaling pouch cells (lack of autonomous activity) (Figs 3I–3O and S3P–S3W). These observations suggest that short-term or moderate activity of JNK/AP-1 may facilitate local, intermixed activation of JAK/STAT. However, sustained or high JNK/AP-1 activity represses JAK/STAT activity cell autonomously. As a result, JNK/AP-1 and JAK/STAT signaling become separated into distinct cell populations within the tissue. Importantly, these patterns of JNK/AP-1 and JAK/STAT separation strongly correlate with spatial separation of regenerative cell behaviors, specifically senescent G2 stalling and compensatory proliferation, into spatially distinct signaling domains. These data suggest that prolonged egr-expression and JNK activation organizes spatial patterns of damage-induced signaling and regenerative cell behaviors into a separated bistable field.

JNK/AP-1 signaling represses JAK/STAT activity Next, we wanted to understand if JNK/AP-1 signaling when activated by egr-expression can directly repress JAK/STAT activation cell-autonomously. We therefore analyzed clonal expression of a constitutively activated form of the JNKK Hep and coexpressed p35 to prevent cell death [82,83]. As expected, hepact clones activated JAK/STAT nonautonomously in the pouch, hinge, or peripodium but, importantly, strongly repressed JAK/STAT activation cell-autonomously (Fig 4A–4D). When clones were placed within the developmentally activated JAK/STAT domain in the hinge [76], cell-autonomous repression could also be observed (Fig 4E). These experiments confirm that JNK/AP-1 has the ability to activate JAK/STAT nonautonomously and strongly demonstrate that JNK/AP-1 represses JAK/STAT signaling cell-autonomously. In these mosaic experiments, central pouch cells displayed lower levels of JAK/STAT activation. This could be due to the expression of Nubbin (Nub) and Rotund (Rn), 2 transcription factors described to reduce developmental Stat92E activity in the pouch [76,84]. To first exclude the possibility that JNK/AP-1 represses JAK/STAT by inducing pouch-specific transcription factors [84], we assessed Nub and Rn levels within hepact clones. hepact clones did not ectopically induce Nub or Rn; in fact, Nub was distinctly down-regulated (S4A–S4C Fig). To exclude that activation of apoptosis and cell death in JNK-signaling cells contributed to repression of JAK/STAT, we also analyzed egr-expressing discs heterozygous for the rpr,hid,grim deficiency Df(3L)H99 (Fig 4F–4K) or coexpressing p35 (S4D and S4E Fig). These genotypes are expected to inhibit caspase activity at initiator or effector levels, respectively. Yet, spatial separation of JNK/AP-1 and JAK/STAT signaling could still be observed. These data strongly suggest that the cell-autonomous repression of JAK/STAT by JNK/AP-1 is neither dependent on developmental mechanisms of JAK/STAT repression nor on JNK-activated cell death pathways. Lastly, to test if activation of JNK/AP-1 by different methods also drives separation of JNK/AP-1 and JAK/STAT signaling domains, we examined wing discs with reduced function of the well-characterized tumor-suppressor gene scrib [85–87]. Expression of scrib-RNAi with or without p35 under the control of rn-GAL4 activated both JNK/AP-1 and JAK/STAT in the pouch as described previously [85–87]. However, both signaling pathways were activated in an almost mutually exclusive pattern (Figs 4L, 4M, S4F and S4G). These experiments suggest that JNK-mediated repression of JAK/STAT and spatial separation of signaling domains may be a robust feature of JNK/AP-1-driven processes. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. JNK/AP-1 signaling represses JAK/STAT activity. (A-D) RFP expression marks p35-expressing control clones in the pouch (magenta, A, C) or p35,hepact-coexpressing clones (JNK+) (magenta, B, D), in the peripodium (A, B) or in the pouch proper (C, D) at 28 h after clone induction in discs also expressing the stable Stat92E-GFP reporter (green). Yellow arrowheads indicate example clones of interest. Note the cell-autonomous repression of Stat92E-GFP reporter activity in JNK+ clones in the pouch and peripodium, when compared to control clones. Note the absence of non-cell-autonomous Stat92E-GFP reporter activity around control clones. A total of n = 8, control discs were evaluated from N = 2 independent experiments, and n = 13, p35,hepact-coexpressing discs were evaluated from N = 3 independent experiments. (E) p35,hepact-coexpressing clones in a disc expressing the stable Stat92E-GFP reporter (green) reporting developmental JAK/STAT activation patterns 48 h after clone induction. The disc was stained for MMP-1 (magenta)—a downstream target of the JNK/AP-1 signaling pathway to confirm pathway activation. The yellow circle highlights a clone, which shows strong cell-autonomous repression of the developmentally regulated JAK/STAT activity in the hinge. A total of n = 7, p35,hepact-coexpressing discs were evaluated. (F, G) A control (F) and egr-expressing disc (G) assessed for JAK/STAT reporter activity when caspase activity is reduced. Both discs a heterozygous for Df(3L)H99, a chromosomal deletion of hid, rpr, and grim, all central regulators of DIAP and thus caspase activity. Discs also express the TRE-RFP (magenta) and Stat92E-dGFP (green) reporters. (H-K) An egr-expressing control disc (H) and an egr-expressing disc heterozygous for Df(3L)H99 (I) expressing the Stat92E-dGFP (green) reporter. Quantification of the Stat92E-dGFP reporter fluorescence intensity measured within egr- and egr, Df(3L)H99-coexpressing central pouch domain. Black square in schematic (J) shows measured region. Graph (K) displays mean ± SEM for n = 13, egr-expressing discs and n = 13, egr-expressing disc heterozygous for Df(3L)H99. Unpaired t test was performed to test for statistical significance. (L, M) scrib-RNAi, p35-coexpressing discs stained for MMP-1 to detect JNK/AP-1 activity (magenta). Disc also expresses the Stat92E-dGFP (green) reporter. Note distinct cell-by-cell segregation and nonoverlapping patterns between the MMP-1 staining (magenta) and the Stat92E-dGFP (green) reporter. A total of n = 4, control and n = 6, scrib-RNAi,p35-coexpressing discs were evaluated. Source data for quantifications provided in S1 File. Discs were stained with DAPI (magenta) to visualize nuclei. Scale bars: 50 μm. https://doi.org/10.1371/journal.pbio.3001665.g004

JAK/STAT repression is required for cell cycle stalling in G2 and to protect arrested cells from apoptosis As spatial separation of JNK/AP-1 and JAK/STAT signaling domains was a robust feature of egr-expressing discs, we wondered if it was necessary for regeneration. We thus asked if ectopic reactivation of JAK/STAT altered the behavior of JNK/AP-1 signaling cells. We reactivated JAK/STAT signaling in egr-expressing cells by overexpression of Stat92E, and by knock-down of Ptp61F or Socs36E, and closely monitored cell and tissue level responses, such as cell survival and proliferation. Reactivation of JAK/STAT signaling in all 3 egr-expressing genotypes led to a pronounced increase in apoptosis in the tissue (Fig 7A–7I). Expression of UAS-GFP in egr,Stat92E-coexpressing cells revealed that apoptotic cells originated from this coexpressing cell population (S7A–S7D Fig), demonstrating that JNK/AP-1 and JAK/STAT coactivation in the same cell is detrimental for cell survival. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 7. JAK/STAT repression is required for cell cycle stalling in G2 and to protect arrested cells from apoptosis. (A-F) Egr-expressing (A, C, E), an egr,Stat92E-coexpressing (B), an egr,Ptp61F-RNAi-coexpressing (D), and an egr,Socs36E-RNAi-coexpressing (F) disc after 24 h of expression, stained for DAPI (cyan), and cleaved Dcp-1 (grey or red) to visualize apoptosis. (G-I) Quantification of percentage of volume occupied by cleaved Dcp-1 for all genotypes shown in (A-F). Graph in (G) displays mean ± SEM for n = 10, egr-expressing and n = 9, egr,Stat92E-coexpressing discs. Unpaired t test was performed to test for statistical significance. Graph in (H) displays mean ± SEM for n = 11, egr-expressing and n = 16, egr,Ptp61F-RNAi-coexpressing discs. Unpaired t test was performed to test for statistical significance. Graph in (I) displays mean ± SEM for n = 11, egr-expressing and n = 9, egr,Socs36E-RNAi-coexpressing discs. Mann–Whitney U test was performed to test for statistical significance. (J, K) An egr-expressing (J) and egr,Stat92E-coexpressing disc (K) after 24 h of expression. Discs also express the FUCCI reporter. Dashed white squares highlight the central pouch domain of rn-GAL-expressing cells. Note the distinct shift from most cells in the G2-phase (yellow nuclei) to also cells in the G1-phase (green nuclei) within the central pouch domain. (L, M) Bar graph (L) representing the cumulative proportion of gap phase cells in G1 (green) and G2 (yellow) for each genotype. Graph (M) displays G1-phase:G2-phase ratios obtained from n = 5, egr-expressing and n = 3, egr,Stat92E-coexpressing discs. Dashed white squares in (J, K) indicate the position of the analyzed domain. Welch’s t test was performed to test for statistical significance. (N-Q) EdU incorporation assays to detect DNA replication of cells in S-phase (gray) in egr-expressing (N, P), or egr,Stat92E-coexpressing (O) and egr,Ptp61F-RNAi-coexpressing (Q) disc. Discs were stained with DAPI (red) to visualize nuclei. A total of n = 10, egr-expressing discs and n = 10, egr,Stat92E-coexpressing discs were evaluated. A total of n = 10 egr-expressing discs and n = 10, egr,Ptp61F-RNAi-coexpressing discs were evaluated. (R, S) An egr-expressing disc after 24 h of expression at R0 (R) and at recovery time point R48h (S) coexpressing an inducible permanent lineage label for rn-GAL4-expressing cells using the G-TRACE system (gray or green). Discs also express the TRE-RFP reporter (magenta) and were stained with DAPI to visualize nuclei. Note how the surviving rn-GAL4 population increases in size suggesting that stalled cells have started cycling again. A total of n = 8, egr-expressing discs at R0 and n = 9, egr-expressing discs at R48h were evaluated. Source data for quantifications provided in S1 File. Maximum projections of multiple confocal sections are shown in A-F and N-S. Scale bars: 50 μm. https://doi.org/10.1371/journal.pbio.3001665.g007 We hypothesized that the pro-proliferative function of JAK/STAT interfered with the G2 cell cycle arrest of JNK/AP-1-signaling cells. Previous work demonstrated that genetically overriding the G2-arrest and forcing JNK/AP-1-signaling cells to cycle increases the probability of these cells to undergo apoptosis, due to activation of p53 by G2/M kinases (see S1A’ Fig) [29,36]. We thus analyzed the ability of Stat92E to override the cell cycle arrest in the G2-phase. In undamaged control discs, targeted expression of Stat92E using rn-GAL4 altered the proportion of cells in S-phase consistent with its reported role in cell cycle acceleration [133], yet importantly, it did not alter the proportion of cells in G1 or G2 (S7E–S7H Fig). In contrast, a cell cycle analysis of egr,Stat92E-HA coexpressing domains revealed a substantial increase in G1-phase cells, indicating that JNK-signaling cells did not stall in G2 anymore (Fig 7J–7M). Moreover, cells undergoing active DNA replication could now be observed in the JNK/AP-1-signaling domain, indicating that JNK-signaling cells were actively cycling (Fig 7N and 7O). Knock-down of Ptp61F in egr-expressing discs also led to an increase in DNA-replicating cells (Fig 7P and 7Q). We conclude that forcing Stat92E activation in high JNK-signaling cells is sufficient to overcome the JNK-dependent G2 arrest, forcing cells to transition through G2/M into G1 and thereby increases the probability of these cells to undergo apoptosis. Our finding that the arrested G2-state could, in principle, be overcome was at odds with the cell population’s senescent features, which would normally be associated with a terminally arrested cell cycle state. To understand if the JNK-induced cell cycle arrest is a transient cellular state, which can be reversed by the right signaling environment, we analyzed the proliferative potential of egr-expressing cells after egr-expression was terminated. Consistent with published lineage tracing results for this population [50], we found previously that egr-expressing cells started to proliferate within 48 h, which correlated with decreasing JNK/AP-1-signaling within the tissue (Fig 7R and 7S). Our findings highlight the necessity of separating JNK/AP-1 and JAK/STAT signaling within the tissue. The JNK-induced mutual repression network ensures the establishment of 2 distinct and indispensable cell populations: (1) a JNK/AP-1-signaling cell population that stalls in G2, which prevents their apoptosis. Their function is to secrete necessary pro-mitogenic factors like Upds, or other paracrine effectors like Dilp8 and ImpL2 [18,80]; and (2) a JAK/STAT-signaling cell population that is able to respond to mitogenic signals and undergoes regenerative proliferation. This spatial separation would ensure that tissue damage sensing can be achieved by an apoptosis-resistant signaling center that survives wound-associated ROS or cellular damage and that regenerative proliferation can occur in an environment not exposed to wound-associated ROS, cellular damage, or inflammatory defense processes. In agreement with the idea that repression of JAK/STAT by JNK/AP-1 and thus bistable patterns are required for regeneration, we find that adult wings developing from egr,Ptp61F-RNAi coexpressing wing discs are smaller than adult wings from egr-expressing discs (S7I Fig).

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3001665

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.