Cullin-4 regulates Wingless and JNK signaling-mediated cell death in the Drosophila eye

In all multicellular organisms, the fundamental processes of cell proliferation and cell death are crucial for growth regulation during organogenesis. Strict regulation of cell death is important to maintain tissue homeostasis by affecting processes like regulation of cell number, and elimination of unwanted/unfit cells. The developing Drosophila eye is a versatile model to study patterning and growth, where complex signaling pathways regulate growth and cell survival. However, the molecular mechanisms underlying regulation of these processes is not fully understood. In a gain-of-function screen, we found that misexpression of cullin-4 (cul-4), an ubiquitin ligase, can rescue reduced eye mutant phenotypes. Previously, cul-4 has been shown to regulate chromatin remodeling, cell cycle and cell division. Genetic characterization of cul-4 in the developing eye revealed that loss-of-function of cul-4 exhibits a reduced eye phenotype. Analysis of twin-spots showed that in comparison with their wild-type counterparts, the cul-4 loss-of-function clones fail to survive. Here we show that cul-4 clones are eliminated by induction of cell death due to activation of caspases. Aberrant activation of signaling pathways is known to trigger cell death in the developing eye. We found that Wingless (Wg) and c-Jun-amino-terminal-(NH2)-Kinase (JNK) signaling are ectopically induced in cul-4 mutant clones, and these signals co-localize with the dying cells. Modulating levels of Wg and JNK signaling by using agonists and antagonists of these pathways demonstrated that activation of Wg and JNK signaling enhances cul-4 mutant phenotype, whereas downregulation of Wg and JNK signaling rescues the cul-4 mutant phenotypes of reduced eye. Here we present evidences to demonstrate that cul-4 is involved in restricting Wg signaling and downregulation of JNK signaling-mediated cell death during early eye development. Overall, our studies provide insights into a novel role of cul-4 in promoting cell survival in the developing Drosophila eye.

death through the activation of caspases. 11,14,[25][26][27] JNK belongs to a conserved MAP kinase super-family, which is involved in cell proliferation and cell survival, and is activated through a cascade of phosphorylation by MAP kinases. 26,[28][29][30] In Drosophila, JNK signaling is activated by binding of the tumor necrosis factor (TNF) Eiger (Egr) to its receptor Wengen (Wgn), and a conserved signaling cascade of Tak 1 (TGFβ activating kinase 1, a Jun kinase kinase kinase (JNKKK), hemipterous (hep) (Jun kinase kinase), basket (bsk) (Jun kinase) and jun. Activation of the pathway leads to expression of the downstream target puckered (puc), a dual phosphatase, which participates in a negative feedback loop by downregulating JNK activity. 29,30 We argued that during early eye development, Wg or JNK levels must be tightly regulated to allow differentiation to proceed, and to prevent premature cell death that results in small or reduced eye phenotype. In a genetic screen, we identified cullin4 (cul-4) as a modifier that rescues the reduced eye phenotype. 31 During development cul-4 is globally required. Analysis of cul-4 function revealed its new role in promoting cell survival during early eye development. The cul-4 gene belongs to an evolutionary conserved class of Cullin-family E3 ubiquitin ligases. 32 Earlier studies showed that cul-4 is involved in maintenance of genomic integrity by promoting the ubiquitylation and subsequent degradation of key regulators of cell cycle progression. [33][34][35][36] Here, we report that cul-4 promotes cell survival by preventing Wg and JNK signaling-mediated cell death in the developing eye.

Results
Gain-of-function of cul-4 rescues reduced eye mutant phenotype. In comparison with the wild-type larval eye disc and the adult eye (Figures 1a and b), L mutant exhibits reduced eye phenotype in larval eye disc ( Figure 1c) and adults ( Figure 1d). 11,37 Misexpression of cul-4 using Gal4/UAS approach 38 (L 2 ; ey4cul-4) resulted in the rescue of L 2 -reduced eye phenotype (Figures 1e and f). Misexpression of cul-4 (ey4cul-4) does not affect the eye size (not shown) suggesting that cul-4 may not promote cell proliferation. We analyzed loss-of-function phenotype of cul-4 to understand its role during eye development.
Loss-of-function of cul-4 fail to survive and exhibit reduced eye phenotype. We generated cul-4 loss-offunction clones by cell lethal (cl) approach, which results in an eye disc comprising of nearly 80% mutant cells due to elimination of the wild-type twin spot by a cl mutation. 39 Loss-of-function clones of cul-4 alleles (Figure 2a) 40 like cul-4 ExG1 − 3 (Figure 2c) or cul-4 JJ11 (Figures 2d and e) resulted in a small eye phenotype as compared with the wildtype eye (Figure 2b). It is to be noted that both cul-4 JJ11 and cul-4 ExG1 − 3 loss-of-function phenotypes were similar in the eye. Downregulation of cul-4 expression in the dorso-ventral (DV) margins of developing eye disc by using bi-Gal4 driver (Figure 2f; bi4GFP) resulted in reduction of eye field on DV margins (Figures 2g and h, arrows). 41,42 It suggests that there is no domain constraint in cul-4 function in the eye. It is possible that reduced eye phenotype may be due to induction of cell death. To test this, we used ey-Flippase (ey-Flp) to induce somatic 'twin clones' using Ubi-GFP (1XGFP), where cul-4 −/− mutant cells (GFP-negative) were adjacent to their wild-type twin spot (2XGFP). However, we found that only the wild-type twin clone (2XGFP) and the heterozygous cells (express 1XGFP) could be identified. However, we found wild-type twin clones (2XGFP) but no cul-4 mutant clones (GFP-negative) in the third-instar eye disc (not shown), suggesting that the cul-4 mutant cells failed to survive. We The heat shock was administered in the first instar larva and the resultant clones were examined in the second-(24 h after clone formation (ACF)) and third-instar (48-72 h after clone formation) disc to determine whether or not these clones could survive. Very small clones were detected at second instar stage (within 24 h after clone formation). However, within 48 h after clone formation, the mutant clones were lost. In comparison with the wild-type clones (Figure 3a), the mutant clones generated at later time points (second instar (48 h)  cul-4 prevents cell death in the developing eye. We, therefore, tested if Drosophila Caspases-Dronc and Drice activation is part of the mechanism. In the eye disc, cul-4 JJ11 loss-of-function clones generated by 'cell-lethal' clonal approach, 39   significantly upregulated in semi-quantitative western blots on total protein isolated from eye imaginal discs from wild-type, and cul-4 ExG1 − 3 and cul-4 JJ11 (Figure 5e). These data suggest that cul-4 may downregulate Wg in the eye imaginal disc. Misexpression of wg on DV margins of eye disc bi4wg results in ectopic wg transcription suggesting that wg is a target of Wg pathway in developing eye ( Supplementary Fig. S1). We then tested if aberrant Wg signaling is responsible for cul-4 mutant phenotypes.
Wg signaling pathway alters the cul-4 mutant phenotype.
Arm translocates to the nucleus in response to Wg signaling and binds with the transcription factor dTCF (LET/TCF family  work in conjunction with Wg in multiple contexts including correction of morphogen gradient discontinuities; 26 and differential levels of JNK signaling are associated with cell survival. 26,28 cul-4 prevents JNK-mediated cell death in the developing eye. We tested if JNK pathway is associated with the cul-4 loss-of-function phenotypes (Figure 7). We tested JNK levels in cul-4 loss-of-function background using puc (puc-lacZ) the downstream target, which serves as the functional readout for JNK pathway activation. 30 In wild-type eye disc, puc is expressed in differentiated photoreceptor neurons of eye disc (Figure 7a). In cul-4 loss-of-function background, ectopic induction of puc was seen in the eye as well antenna disc (Figures 7b and b'), suggesting that JNK signaling is activated in cul-4 mutant cells. To confirm, we checked levels of p-JNK, a reporter for activated JNK signaling, in western blots and found significant upregulation of p-JNK levels in cul-4 mutants compared with wild-type eye disc (Figure 7c). Blocking JNK signaling in the developing eye by misexpression of puc (ey4puc, Figure 7d) or bsk dominant negative (bsk DN ) (ey4bsk DN , Figure 7g) 25 (Figure 2). Conversely, in cul-4 loss-of-function background activation of the JNK signaling pathway by misexpression of activated Jun (jun aspv7 ) (cul-4 JJ11 − / − , ey4jun aspv7 ) in the eye disc, strongly enhanced the reduced eye-to a 'no-eye' phenotype (Figures 7k and k'). Misexpression of jun aspv7 (ey4jun aspv7 ) alone in the eye results in a highly reduced eye field (Figure 7j). It suggests that loss of cul-4 leads to activation of JNK signaling in the eye.
To confirm that activation of Wg /JNK signaling pathway are both associated with the induction of cell death observed in cul-4 mutant cells, we monitored cell death using TUNEL assay when Wg ( Figure 8) and JNK (Figure 9) levels are modulated in the wild-type, and in cul-4 mutant eye discs. We found that cell death is reduced when Wg or JNK signaling is downregulated in cul-4 mutant background. However, cell death is elevated when Wg/ JNK signaling is activated. Thus, cul-4 may be involved in limiting JNK as well as Wg activation in the developing eye disc, and thereby promote cell survival during development.

Discussion
Cul-4, an E3 ligase, is involved in regulation of chromatin function through heterochromatin gene silencing, maintenance of genomic integrity by promoting the ubiquitylation  51 and Mind Bomb (Mib), are E3 ligases that are components of N signaling pathway; and are required for Drosophila eye development. 52 Recently other functions for E3 ligases are being recognized. For example, DIAP1 regulates Dronc/Hid caspases, 20,53 and is transcriptionally regulated by yorkie (yki) for survival function. 54 DIAP-1 in turn, is regulated by Cul-3 in the developing eye to regulate apoptosis. 71 Our studies provide evidences for a new function for cul-4 in cell survival during eye development.
Homozygous larvae of some cul-4 alleles are larval lethal that can survive until early third instar and produce smaller imaginal discs than wild-type discs at comparable developmental age. 40 These phenotypes were attributed to problems with cell division. Our twin spot analysis revealed an interesting result that cul-4 mutant tissues in the developing eye imaginal disc failed to survive (Figure 2), and are eliminated by activation of caspases (Figure 3). Generating cul-4 mutant clones by using multiple approaches (for example, eyeless and heat-shock flippase) validated that  Figures 4c and d), supporting a role for cul-4 in cell survival. We tested several cell signaling pathways in cul-4 loss-offunction background and found aberrant activation of Wg and JNK signaling (Figure 10). Wg is required for patterning, growth regulation and cell survival in multiple tissues including the eye discs. Ectopic induction of Wg induces cell death. 9,10,11,55 We found that cells lacking cul-4 function also undergo cell death and they express high levels of Wg. Arm, the nuclear effector of the Wg signaling pathway, is a target of E3 ubiquitin ligase-mediated degradation. 56 Loss-of-function phenotype of cul-4 mutants could be modified by modulating the levels of canonical Wg signaling (Figures 6 and 7). Our data suggests that cul-4 may downregulate Wg signaling in the eye to promote cell survival in the eye disc. Since the cul-4 mutant phenotype was not completely rescued by blocking Wg signaling, we also tested the JNK signaling in the cul-4 mutant clones. The possibility of indirect consequences responsible for the mutant phenotype can be refuted because these phenotypes can be rescued by blocking Wg as well as JNKmediated cell death. We found that Wg levels were affected when JNK signaling was modulated in cul-4 mutant background (Figure 10a-c). However, the converse did not show effect on phospho-JNK levels (Figure 10d-f). Our studies generate insights into genetic mechanisms that regulate cell  Figure 10). A recent study showed that loss of Godzilla, a member of the RNF family of membrane-anchored E3 ubiquitin ligases regulates Wg levels on the basolateral surface of the tissues through dynamin-dependent endocytosis from the apical surface and subsequent trafficking from early apical endosomes to the basolateral surface. 57 Our studies also generate mechanistic insights into genetic mechanisms that regulate cell survival during normal development.
Numerous studies have shown the role of ubiquitinmediated proteolysis in a broad array of cellular processes like defects in organogenesis, growth, differentiation, metabolism and aging in all organisms. 58 Abnormal protein homeostasis underlies various disorders ranging from growth defects to neurodegenerative disorders. 59 Our studies introduced new role of cul-4 in cell survival in the developing Drosophila eye. Since Drosophila serves as an excellent model to study /CyO, which lack 340 amino acids from amino terminal. cul-4 JJ11 / twi4GFP, CyO carries a nonsense mutation at Trp199 position. 40 Other stocks include Canton-S, y w eyFLP, 39 L 2 /Cyo; 37,62 wg-lacZ/CyO, 63 UAS-wg, 64 UAS-sgg S9A , 65 UAS-arm, 66 UAS-dTCF DN , 44 UAS-P35, 24 Df(3L)H99/TM6B, 43 puc E69 , UAS-puc, 30 UAS-bsk DN,25 and UAS-DJun aspv7 . 67 The Gal4/UAS system was used for targeted misexpression studies 38 using ey-Gal4, 65  Twin spot analysis. We used hsFlp; FRT42D ubi-GFP to generate loss-offunction clones of cul-4 ExG1 − 3 and cul-4 JJ11 in the eye imaginal disc at different larval development stages. Egg laying were collected from synchronous cultures maintained at 25°C. The cultures were heat shocked at 24 and 48 h after egg laying (AEL) at 37°C for 50 min in order to induce loss-of-function clones. Eye discs were dissected in second and third-instar stages to analyze/identify cul-4 loss-offunction clones marked by the absence of GFP expression.
Immunohistochemistry. Eye-antennal imaginal discs were dissected from wandering third-instar larvae and stained following the standard protocol. 62 Antibodies used were rat anti-Elav TUNEL assays. Apoptotic cell death was assayed using TUNEL assays in the mutant clones generated via twin spot analysis and cell lethal approach. Eye discs, after secondary antibody staining, 62 were blocked in 10% Normal Goat Serum in Phosphate Buffered Saline with 0.2% Triton X-100.TUNEL assays were done using the Cell-death Detection Kit from Roche Diagnostics following the standardized protocol. 11,17 The TUNEL-positive nuclei were counted from five eye imaginal discs for each genotype using Image-J and statistical analysis was performed using Microsoft Excel 2013. The P-values were calculated and the error bars represent Standard Deviation.
Adult eye imaging. Adult Drosophila eye images were taken 70 using a Zeiss Apotome Imager.Z1 microscope (Carl Zeiss Microscopy GmbH, Jena, Germany). The flies were prepared by freezing them at − 20°C for~2 h. The legs and wings of the flies were removed and flies were mounted on a dissection needle, and the fly was positioned on a glass slide using mounting putty. Images were captured by using extended depth of focus function of the Axiovision software version 4.6.3 (Carl Zeiss Microscopy GmbH, Jena, Germany) by compiling the individual stacks from the Z-sectioning approach. The final images and figures were prepared using Adobe Photoshop CS4 software.
Developmental Studies Hybridoma Bank (DSHB) for the antibodies. We thank members of the Singh laboratory for critical comments on the manuscript. Confocal microscopy was supported by Biology Department central core facility. MK-S is supported by Knight's Templar Ophthalmology Research Foundation and startup support from UD. This work is supported by start-up support from UD, and Ohio Cancer Research Associates to AS.

Author contributions
MT performed majority of the experiments and the data analysis. AS performed experiments. SB performed experiments. MK-S did the data analysis and the manuscript writing. AS developed the concept and did the data analysis and manuscript writing.