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



Actin remodeling mediates ROS production and JNK activation to drive apoptosis-induced proliferation [1]

['Luchi Farrell', 'University Of Birmingham', 'School Of Biosciences', 'Birmingham', 'United Kingdom', 'Aleix Puig-Barbe', 'Md. Iqramul Haque', 'Department Of Physiology', 'Faculty Of Veterinary Science', 'Bangladesh Agricultural University']

Date: 2022-12 

Stress-induced cell death, mainly apoptosis, and its subsequent tissue repair is interlinked although our knowledge of this connection is still very limited. An intriguing finding is apoptosis-induced proliferation (AiP), an evolutionary conserved mechanism employed by apoptotic cells to trigger compensatory proliferation of their neighboring cells. Studies using Drosophila as a model organism have revealed that apoptotic caspases and c-Jun N-terminal kinase (JNK) signaling play critical roles to activate AiP. For example, the initiator caspase Dronc, the caspase-9 ortholog in Drosophila, promotes activation of JNK leading to release of mitogenic signals and AiP. Recent studies further revealed that Dronc relocates to the cell cortex via Myo1D, an unconventional myosin, and stimulates production of reactive oxygen species (ROS) to trigger AiP. During this process, ROS can attract hemocytes, the Drosophila macrophages, which further amplify JNK signaling cell non-autonomously. However, the intrinsic components connecting Dronc, ROS and JNK within the stressed signal-producing cells remain elusive. Here, we identified LIM domain kinase 1 (LIMK1), a kinase promoting cellular F-actin polymerization, as a novel regulator of AiP. F-actin accumulates in a Dronc-dependent manner in response to apoptotic stress. Suppression of F-actin polymerization in stressed cells by knocking down LIMK1 or expressing Cofilin, an inhibitor of F-actin elongation, blocks ROS production and JNK activation, hence AiP. Furthermore, Dronc and LIMK1 genetically interact. Co-expression of Dronc and LIMK1 drives F-actin accumulation, ROS production and JNK activation. Interestingly, these synergistic effects between Dronc and LIMK1 depend on Myo1D. Therefore, F-actin remodeling plays an important role mediating caspase-driven ROS production and JNK activation in the process of AiP.

In multicellular organisms, damaged cells are frequently removed via apoptosis, the major form of programmed cell death. Intriguingly, these apoptotic cells can emit signals to induce proliferation of their neighboring cells for the maintenance of tissue homeostasis, a phenomenon termed apoptosis-induced proliferation (AiP). Caspases, a family of cysteine proteases well known to execute apoptosis, also play a critical role to trigger AiP via activation of JNK, a stress response kinase. In this study, we identified the actin cytoskeleton, a dynamic structural network of the cell, as a key mediator of JNK activation in AiP. During this process, actin filaments undergo increased polymerization which depends on LIM domain kinase 1 (LIMK1). Genetically, caspases and LIMK1 work together to promote actin polymerization, which in turn drives production of reactive oxygen species (ROS) and the subsequent activation of JNK. Therefore, our study discovered a role of actin remodeling in dying cells which mediates the non-apoptotic roles of caspases in AiP and tissue repair.

Funding: This work was supported by the grants BB/M010880/1 and BB/S015701/1 from the UKRI BBSRC to YF and a MIRA award R35 GM118330 from the NIH/NIGMS to AB. LF and MIH are supported by the Midlands Integrative Biosciences Training Partnership (MIBTP) programme and the Commonwealth Scholarship Commission in the UK (CSC), respectively, for their PhD studies. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Introduction

In multicellular organisms, damaged cells are frequently removed by apoptosis, the major form of programmed cell death. Intriguingly, these stress-induced dying cells, prior to their removal, can actively induce proliferation of neighboring cells to compensate for their loss and maintain tissue integrity [1–8]. This phenomenon has been termed as apoptosis-induced compensatory proliferation or apoptosis-induced proliferation (AiP) [9, 10]. Importantly, AiP is found not only to promote tissue recovery and regeneration, but also to drive tumor development and cancer recurrence [4, 11–13]. Therefore, understanding the mechanisms of AiP has significant clinical and therapeutic implications.

Caspases, an evolutionary conserved family of cysteine proteases, are essential for both apoptosis and AiP. Whilst caspases mediate execution of apoptosis, they can also promote AiP via activation of mitogenic signals such as cytokines or growth signaling pathways in a context-dependent manner [1, 4, 5, 14–16]. Studies in Drosophila have revealed distinct mechanisms of AiP mediated by different groups of caspases. In differentiating tissues where cells have exited mitosis, the caspase-3-like effector caspases DrICE and Dcp-1 activate Hedgehog signaling to drive cell cycle re-entry [14]. In contrast, in proliferating tissues where cells are actively dividing, the caspase-9-like initiator caspase Dronc activates the c-Jun N-terminal kinase (JNK), an evolutionary conserved stress response molecule, which in turn induces growth signals such as Wg, Dpp and EGFR signaling resulting in AiP [1, 15, 17].

Notably, in the proliferating larval eye and wing epithelia, Reactive Oxygen Species (ROS), the oxygen-containing free radicals produced during cellular metabolism, accumulate in apoptotic tissues and trigger activation of JNK [18, 19]. Recent studies further showed that ROS and JNK in apoptotic cells also damage the epithelial basement membrane and act as signals to recruit macrophage-like hemocytes, which in turn contribute to further activation of JNK signaling cell non-autonomously [20–22]. Interestingly, the initiator caspase Dronc translocates from cytosol to the plasma membrane, where it exerts its non-apoptotic function to activate Duox, a NADPH oxidase, for ROS production [20]. Myo1D, an unconventional myosin, is critical for this process through its interaction with Dronc. However, it remains unknown what mediates this non-apoptotic action of Dronc and Myo1D to drive ROS production and JNK activation within the dying cells.

To study the mechanisms of AiP in proliferating tissues, we have previously developed two genetic assays, an overgrowth assay and a regeneration assay, to identify regulators of AiP [15]. In the overgrowth assay, hid, a pro-apoptotic gene, and p35, an inhibitor of effector caspases DrICE and Dcp-1, are simultaneously expressed under control of the ey-GAL4 driver (ey>hid-p35, Fig 1A) in the developing larval eye epithelium which is composed of the anterior proliferating and the posterior differentiating tissues. While the posterior differentiating tissue develops to the adult eye, the anterior proliferating tissue develops to adult head appendages including cuticle and sensory organs such as ocelli and bristles. Expression of P35 inhibits the effector caspases DrICE and Dcp-1, therefore AiP in the differentiating tissue, but not the initiator caspase Dronc [14, 23]. Hence, Dronc-dependent AiP occurs specifically in the proliferating eye tissues in ey>hid-p35. Cells in this tissue are kept ‘undead’ because apoptotic responses are activated but execution of cell death is blocked. These ‘undead’ cells continuously drive AiP leading to a head overgrowth phenotype, a convenient readout of AiP for genetic screens (Fig 1A). In the regeneration assay, hid is expressed in a temporally and spatially controlled manner to induce a pulse of apoptosis therefore tissue ablation. This tissue is then allowed to recover to observe AiP and study its regulation during tissue regeneration (Fig 2A).

PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 1. LIMK1 and Cofilin, regulators of F-actin polymerization, are required for AiP. (A) A schematic representation of the AiP-dependent overgrowth assay. ey-GAL4 is expressed in the developing larval eye disc which is composed of the anterior proliferating tissue and the posterior differentiating tissue, separated by a morphogenic furrow (MF, the grey bar). Compared to the control ey>p35 (ey-GAL4 UAS-p35), simultaneously expression of hid and p35 in ey>hid-p35 (ey-GAL4 UAS-hid UAS-p35) results in AiP-dependent overgrowth of the anterior proliferating tissue, which leads to an adult head overgrowth phenotype characterized by expanded head cuticle with ectopic sensory organs such as ocelli and bristles. Consequently, sizes of the differentiating eye tissue and adult eye are reduced in ey>hid-p35 animals. These larval and adult phenotypes were used as readouts of AiP in our analyses. (B-I) Representative adult fly head images of the indicated genotypes. Compared to the control ey>p35 (B), which is similar to wildtype, ey>hid-p35 fly head capsules display overgrowth phenotypes which can be grouped into three categories (C-E): severe (S), moderate (M) and weak (W, including wildtype-like), as previously described [15]. A majority of ey>hid-p35 flies show either severe (57%) or moderate (34%) overgrowth phenotype, characterized by overgrown head capsules with duplications of sensory organs including bristles and ocelli (C and D, arrows). (F) Knockdown of LIMK1 by RNAi does not cause any defects. (G-I) LIMK1RNAi-57012 strongly reduces the percentage of ey>hid-p35 flies displaying severe (4%) and moderate overgrowth (34%) phenotypes, with a large increase of flies (62%) showing a weak phenotype or wildtype-like appearance. (J) Summary of the suppression of the ey>hid-p35 overgrowth phenotype by a LIMK1 hypomorphic mutant (LIMK12), expressing three independent LIMK1RNAi lines (26294, 42576 and 57012) or cofilin. Black indicates severe, dark grey indicates moderate, and light grey indicates weak or wildtype-like phenotypes. (K) Specificity and efficiency of LIMK1 RNAi lines were determined by measuring LIMK1 transcript levels with RT-qPCR. Compared to the control without expression of LIMK1RNAi, two independent LIMK1RNAi lines, 26294 and 57012, suppress LIMK1 transcript levels to less than 50% and 30%, respectively. These reductions are statistically significant (***P<0.001). (L-O) Late 3rd instar eye discs labelled with the mitotic marker PH3 (green) and the photoreceptor neuron marker ELAV (magenta), anterior is to the left. ELAV is used to mark the posterior differentiating portion of the eye discs. White dotted lines indicate the anterior proliferating portion of the eye discs. Compared to the control ey>p35 (L), the size of the anterior proliferating portion and the number of mitotic cells increase in the ey>hid-p35 eye discs (M). These increases are suppressed by expressing LIMK1RNAi-57012 (N) or cofilin (O). (P) Quantification of the number of PH3+ cells in the anterior portion of the eye discs of the indicated genotypes. Compared to ey>p35, the number of PH3+ cells are significantly (**p < 0.01) increased in the ey>hid-p35 discs. This increase is significantly (**p < 0.01) reduced in response to expression of droncRNAi, LIMK1RNAi or cofilin. (Q and R) Expression levels of LIMK1 (Q) and cofilin (R) were measured by RT-qPCR. Compared to the control ey>p35-mCherry (ey-GAL4 UAS-p35 UAS-mCherry) eye discs, expression of LIMK1 is increased at about 1.6-fold in ey>hid-p35 (***P<0.001). In contrast, expression of cofilin is not significantly changed. https://doi.org/10.1371/journal.pgen.1010533.g001

PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 2. LIMK1 is required for complete tissue regeneration in response to apoptosis. (A) A schematic representation of the AiP-dependent regeneration assay. Conditional expression of GFP and the pro-apoptotic gene hid is under control of the Dorsal Eye-GAL4 (DE-GAL4) driver, which is expressed in the dorsal half of the eye disc, together with tub-GAL80ts, a ubiquitously expressed and temperature sensitive (ts) inhibitor of GAL4. tub-GAL80ts inhibits DE-GAL4 (DEts) at 18°C, but not at 30°C. In this regeneration assay, a temperature shift to 30°C for 12 hours during second instar larval stage induces expression of GFP together with or without hid. This is then followed by a recovery period at 18°C. Compared to the control with GFP expression only, a 12-hour expression of hid results in tissue ablation in the dorsal eye half, which is regenerated completely after a recovery period of 72 hours (R72h). (B-C’) Early 3rd instar eye discs, anterior is to the left, were labelled with GFP (green in B, C) and an apoptosis marker cDcp1 (magenta in B, C and grey in B’, C’). Dashed lines highlight the discs. Compared to the control (DEts>GFP, B, B’), conditional expression of hid (DEts>hid, C, C’) for 12 hours under the control of DEts results in a strong induction of apoptosis (arrows) and loss of bilateral symmetry of the disc which are visible after a recovery period of 24 hours (R24h). (D-F’) Late 3rd instar eye discs, anterior is to the left. ELAV (magenta in D, E, F and grey in D’, E’, F’) labels photoreceptor neurons and is used to outline the shape of the discs. Conditional expression of LIMK1RNAi (D, D’), hid (E, E’) or hid and LIMK1RNAi (F, F’) was under control of DEts and indicated by GFP (green in D, E, F). A temperature shift to 30°C for 12 hours was followed by a recovery period of 72 hours at 18°C (R72h). (D, D’) Following this protocol, expression of LIMK1RNAi alone (DEts>LIMK1RNAi) does not affect the eye disc morphology indicated by the normal ELAV pattern in the dorsal half of the eye disc. (E, E’) Tissue damage induced by expression of hid (DEts>hid) for 12 hours has fully recovered after 72 hours recovery (R72h) at 18°C as indicated by the largely normal ELAV pattern in the late third instar eye discs. (F, F’) A DEts>hid eye disc that simultaneously expresses LIMK1RNAi (DEts>hid-LIMK1RNAi). The arrow in (F’) highlights the incomplete ELAV pattern on the dorsal half of the disc indicating that the regenerative response was partially impaired by reduction of LIMK1. (D) Quantification of the dorsal/ventral size ratio in representative eye discs of the indicated genotypes. Compared to DEts>hid, expression of LIMK1RNAi or cofilin significantly (*P < 0.5) reduces the size of the dorsal half of the eye disc. As the controls, disc sizes of DEts>LIMK1RNAi and DEts>cofilin are not significantly (n.s.) different from those of DEts>hid. https://doi.org/10.1371/journal.pgen.1010533.g002

The actin cytoskeleton, a structural network found in all eukaryotic cells, consists of actin and actin-binding proteins [24, 25]. Actin exists in two forms: globular monomers (G-actin) and filamentous polymers (F-actin). Actin filaments are highly dynamic and constantly undergo assembly and disassembly. The spatial and temporal regulation of F-actin mediates multiple cellular signaling responses including cell division, motility and phagocytosis. The evolutionary conserved LIM domain kinase 1 (LIMK1) and Cofilin are essential for the formation of actin filaments [26, 27]. Cofilin, also called actin-depolymerizing factor (ADF), promotes rapid F-actin turnover by severing actin filaments [28]. LIMK1 phosphorylates Cofilin and inhibits its activity, therefore promotes F-actin polymerization. Although AiP relies on cell-cell interactions and signal transduction, it is not yet clear whether and how the F-actin network is involved in this process.

Using both overgrowth and regeneration assays, in this study, we showed that knockdown of LIMK1 or overexpression of Cofilin suppresses both AiP-dependent tissue overgrowth and regeneration. Consistent with this, F-actin polymerization, regulated by LIMK1 and Cofilin, was observed in AiP. It depends on the initiator caspase Dronc and is required for activation of JNK and its upstream ROS production. Furthermore, Dronc and LIMK1 genetically interact and drive F-actin remodeling in AiP via Myo1D.

[END]
---
[1] Url: https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1010533

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/