Novel initiator caspase reporters uncover previously unknown features of caspase-activating cells

ABSTRACT The caspase-mediated regulation of many cellular processes, including apoptosis, justifies the substantial interest in understanding all of the biological features of these enzymes. To complement functional assays, it is crucial to identify caspase-activating cells in live tissues. Our work describes novel initiator caspase reporters that, for the first time, provide direct information concerning the initial steps of the caspase activation cascade in Drosophila tissues. One of our caspase sensors capitalises on the rapid subcellular localisation change of a fluorescent marker to uncover novel cellular apoptotic events relating to the actin-mediated positioning of the nucleus before cell delamination. The other construct benefits from caspase-induced nuclear translocation of a QF transcription factor. This feature enables the genetic manipulation of caspase-activating cells and reveals the spatiotemporal patterns of initiator caspase activity. Collectively, our sensors offer experimental opportunities not available by using previous reporters and have proven useful to illuminate previously unknown aspects of caspase-dependent processes in apoptotic and non-apoptotic cellular scenarios.

. Labelling of caspase-activating cells through DBS-S A) The overexpression of P53 under the control of spalt-Gal4 induces prominent cell death. XY images at different focal planes and cross-sections of the wing discs (Zsection) are shown in A. The discontinuous yellow line indicates the location of the Zsection in the XY image. Caspase-activating cells are positively labelled by the nuclear translocation of a histone-GFP fragment from DBS-S (green); actin is labelled with LifeAct-Ruby in red; nuclei are stained with Dapi. B) Time-lapse of a wing imaginal disc ex vivo after irradiation (apical view of the wing discs; time is indicated in minutes; DBS-S is show in green; LifeAct-Ruby shows actin in red). Notice the progressive accumulation of GFP positive nuclei in apical positions of the wing epithelia. The recording started 90 minutes after irradiation and it lasted for 40 minutes. Images were acquired with a 2 minutes interval. C) Nuclei trajectories observed in time-lapses from irradiated discs (real information shown in movie 3 and movie 4). Blue arrows indicate apical interkinetic nuclear movement, while red arrows highlight the nuclear trajectory of delaminating cells. D) Time-lapse series of an abdominal larval epithelial cell undergoing apoptosis during metamorphosis (white arrow, DBS-S > green). Notice that the GFP signal is initially located in the membrane bound cytoplasmic structures (arrowhead points ER/Golgi) and progressively translocates to the nucleus (arrow). Strong nuclear GFP signal is only distinguishable 20 minutes after movie starts, when presumably high levels of caspase activation are achieved. Nuclear GFP accumulation is correlated with the concomitant depletion of the signal from the cytoplasm. Time is indicated in minutes in all figure panels. Development: doi:10.1242/dev.170811: Supplementary information Figure S3. Design and functional characterization of DBS-S-QF. A) Schematic diagram illustrating the rational design of DBS-S-QF. B) Induced cell death and DBS-S-QF activation in a wing imaginal disc expressing P53 under the regulation of spalt-Gal4. Nuclear beta-gal staining indicates the activation of DBS-S-QF (QUAS-nucLacZ, anti-betagal, red); UAS-CD8-GFP is shown in green; Dapi staining labels the nuclei. C and D) Induced cell death and DBS-S-QF activation in eye or wing imaginal discs expressing P53, under the regulation of either GMR promoter or spalt-Gal4, respectively. QUAS-Tomato-HA signal indicates the activation of DBS-S-QF (QUAS-tomato-HA, anti-HA, red); caspase-3 staining is shown in green; Dapi staining labels the nuclei. E) Transient labelling of caspaseactivating cells in the wing discs obtained with DBS-S-QF (QUAS-tomato-HA, anti-HA, red); sensory organ precursors are labelled by neuralized-nuclear lac-Z staining in green). F and G) The ectopic expression of Diap-1 induced by DBS-S-QF activation compromises the apoptosis of larval cells in pupal stages (GFP cells indicated by white arrow in E), and ultimately causes dorsal closure defects in adults (abnormal accumulation of larval cells in adult abdomen is shown in green in F). To obtain a full description of genotypes, please see MM. H) Diagram showing the labelling system used in combination with DBS-S-QF to visualize the temporal patterns of apical caspase activation. The combination of different markers allows the labelling of caspase-activating cells transiently or permanently. The current/ongoing activation of caspases is shown in red upon induction of QUAS-Tomato-HA (anti-HA in red). Concomitantly, QF translocation can also induce the expression of a Gal4 transcription factor (QUAS-Gal4). Gal4 production can then trigger the expression of a second cellular marker (UAS-CD8-GFP). The appearance of the GFP signal is delayed in respect to the QUAS-Tomato since it demands a second round of transcriptional events; however, both markers can co-exist (yellow cells; yellow indicates either sustained caspase activation or alternatively, sensor activation in the recent past). If apical caspase activation terminates QF translocation finishes; however, the transcriptional amplification obtained with the Gal/UAS transcription loop can maintain the GFP signal for long periods of time after the QUAS-tomato signal disappears (Old caspase activation, green cells). Lineage-tracing (permanent labelling) of caspase activating cells is also achievable by using DBS-QF. In this case the QF transcription factor activates a recombinase (QUAS-flipase), which subsequently mediates the genomic excision of a FRT-stop cassette. Before recombination, the FRT-stop cassette prevents the activation of a cellular marker (nuclear-lacZ) under the regulation of an actin ubiquitous promoter (permanent labelling). I) Representative example of H. To obtain a full description of genotypes, please see MM. J) An accumulative plot of clones growing in the wing disc obtained through the permanent labelling of caspase-activating cells with DBS-QF (number of wing discs analysed n=18, average number of lacZ clones per disc 21.5 ± 2.64). Warmer areas (red) in the heat map indicate higher concentration of clones. Notice that clones tend to preferentially grow in the proximal regions of the wing discs (wing hinge), remaining distal parts (wing blade) less crowded.

Movie 1. Time-lapse ex vivo of a wing imaginal disc after irradiation.
Nuclei marked in green indicate DBS-S activation, LifeAct-Ruby shows actin in red. The time-lapse is showing an apical focal plane of the wing disc. Notice the progressive accumulation of GFP-positive nuclei in apical focal planes. The recording started 90 minutes after irradiation and it lasted for 40 minutes. Images were acquired within a 2 minutes interval.

Movie 2. Detail of nuclear apical migration in apoptotic cells.
Nuclei marked in green indicate DBS-S activation, LifeAct-Ruby shows actin in red. The time-lapse is showing a cross-section of a wing imaginal disc. Notice the progressive accumulation of GFP signal in the nuclei, the changes in their shape (roundness increase) and the movement towards the apical cell cortex. Notice also the contraction of the actin bundles over time (blue arrowheads). The recording started 90 minutes after irradiation and it lasted for 40 minutes. Snapshots were acquired every 2 minutes.
Movie 3. General view of the wing disc shown in movie 2. Nuclei marked in green indicate DBS-S activation, LifeAct-Ruby shows actin in red. The time-lapse is showing a cross-section of a wing imaginal disc. Notice the progressive accumulation of GFP signal in the nuclei, the changes in their shape (roundness increase) and the movement towards the apical cell cortex (white arrowheads). Notice also the contraction of the actin bundles over time (blue arrowheads). The recording started 90 minutes after irradiation and it lasted for 40 minutes. Snapshots were acquired every 2 minutes. Movie 5. Schematic diagram summarizing the apoptotic events uncovered with DBS-S. DBS-S translocation into the nucleus is represented in green. Changes in actin dynamics are illustrated in red. Development: doi:10.1242/dev.170811: Supplementary information Movie 6. Tracking of apoptotic larval epithelial cells (LECs) in the fly abdomen cells during pupal stages with DBS-S. In non-dying cells the GFP signal from DBS-S is retained at the cellular membranes (arrowhead point ER-Glogi structures), remaining the nuclei unlabelled. Arrow highlights a dying LEC that progressively accumulates the GFP signal into the nucleus. Strong nuclear GFP signal is only distinguishable 20 minutes after movie starts, when presumably high levels of caspase activation are achieved. Nuclear accumulation of GFP is correlated with the concomitant depletion of the signal from the cytoplasm. At the end of the movie, the nucleus of the apoptotic LEC is fragmented. A region of a hemi-segment of segment A2 is shown. The time-length of the movie was 2.6 hours and frames were acquired every 150 seconds. Anterior is to the left. Dorsal is up.
Movie 7. Tracking of apoptotic histoblasts in the fly abdomen during pupal stages with DBS-S. In non-dying histoblasts the GFP signal from DBS-S is retained at the cellular membranes and the nuclei remain unlabelled. Arrows highlight dying histoblasts that progressively accumulates the GFP signal into the nucleus. Eventually the nucleus of the apoptotic cells is fragmented and the dying/dead cell delaminates. A region of a hemi-segment of segment A2 is shown. The time-length of the movie was 52.5 minutes and frames were acquired every 150 seconds. Anterior is to the left. Dorsal is up.