Live imaging of muscle histolysis in Drosophila metamorphosis
The contribution of programmed cell death (PCD) to muscle wasting disorders remains a matter of debate. Drosophila melanogaster metamorphosis offers the opportunity to study muscle cell death in the context of development. Using live cell imaging of the abdomen, two groups of larval muscles can be observed, doomed muscles that undergo histolysis and persistent muscles that are remodelled and survive into adulthood.
To identify and characterize genes that control the decision between survival and cell death of muscles, we developed a method comprising in vivo imaging, targeted gene perturbation and time-lapse image analysis. Our approach enabled us to study the cytological and temporal aspects of abnormal cell death phenotypes.
In a previous genetic screen for genes controlling muscle size and cell death in metamorphosis, we identified gene perturbations that induced cell death of persistent or inhibit histolysis of doomed larval muscles. RNA interference (RNAi) of the genes encoding the helicase Rm62 and the lysosomal Cathepsin-L homolog Cysteine proteinase 1 (Cp1) caused premature cell death of persistent muscle in early and mid-pupation, respectively. Silencing of the transcriptional co-repressor Atrophin inhibited histolysis of doomed muscles. Overexpression of dominant-negative Target of Rapamycin (TOR) delayed the histolysis of a subset of doomed and induced ablation of all persistent muscles. RNAi of AMPKα, which encodes a subunit of the AMPK protein complex that senses AMP and promotes ATP formation, led to loss of attachment and a spherical morphology. None of the perturbations affected the survival of newly formed adult muscles, suggesting that the method is useful to find genes that are crucial for the survival of metabolically challenged muscles, like those undergoing atrophy. The ablation of persistent muscles did not affect eclosion of adult flies.
Live imaging is a versatile approach to uncover gene functions that are required for the survival of muscle undergoing remodelling, yet are dispensable for other adult muscles. Our approach promises to identify molecular mechanisms that can explain the resilience of muscles to PCD.
KeywordsLive imaging Drosophila Metamorphosis Muscle atrophy Muscle remodelling Cell death Histolysis
after head eversion
AMP-activated protein kinase α subunit
cysteine proteinase 1
Dorsal external oblique muscle
dorsal internal oblique muscles
green fluorescent protein
laser scanning confocal microscopy
maximum intensity projection
monomeric kusabira orange
mammalian target of rapamycin
programmed cell death
prepupal to pupal transition
time of death
target of rapamycin toxic effector domain
Skeletal muscles are essential for mobility and metabolism. Therefore, the preservation of muscles mass and strength plays an important role in improving the quality of life in sickness and old age. The two most common types of muscle wasting are sarcopenia, the ageing related loss of skeletal muscles, and cachexia, a metabolic syndrome associated with diseases such as cancer, heart failure and HIV . Muscle degeneration can also result from heritable muscular dystrophies, such as Duchenne muscular dystrophy  or centronuclear myopathies .
Muscle wasting can result from two cellular processes; the reversible reduction of muscle fiber size by atrophy or the irreversible elimination of muscles by programmed cell death (PCD). The changes of muscle size are controlled by conserved signaling pathways that regulate the rates of protein synthesis and degradation. Protein synthesis and cell growth are activated by a signaling cascade consisting of insulin-like growth factor-1 (IGF-1), the kinase Akt1 and the mammalian target of rapamycin (mTOR) . Protein breakdown and atrophy are activated by a pathway comprising Myostatin, Smad2 and the FoxO transcription factors . Proteins are degraded by two processes, the autophagy lysosomal pathway  and the ubiquitin proteasome system .
Less understood is the contribution of PCD to muscle wasting. Satellite cells are myogenic stem cells that fuse with muscles to repair injuries and drive hypertrophy. In old age, satellite cells are more prone to apoptosis, thus promoting sarcopenia by affecting the repair of damaged muscles . The relationship between cell death in postmitotic, multi-nucleated muscle fibers and muscle wasting is less clear. Skeletal muscles are considered to be more resistant to PCD than proliferating and mono-nucleated cells. In cultured C2C12 cells treated with the apoptosis inducers H2O2 or staurosporine, apoptosis was significantly reduced to 10 % in myotubes compared to 50 % in myoblasts, although several pro-apoptotic proteins like caspases, EndoG and AIF were induced . The protective effect in differentiated muscles was proposed to be mediated by yet undefined anti-apoptosis mechanisms. Muscle fibers exhibit morphological changes during PCD that are different from those observed in mono-nucleated cells. The structural analysis of rat skeletal muscles during post-denervation atrophy revealed cell shrinkage and chromatin condensation. However, the occurrence of DNA fragmentation, a hallmark apoptotic phenotype, was negligible . Furthermore, apoptosis associated with atrophy in rat muscles was found outside of but not inside muscle fibers .
Most knowledge about PCD in mammalian muscles is derived from experimental interventions like denervation, immobilization or exposure to toxins. As such, progress is hampered by the lack of models to study naturally occurring muscle cell death. Non-mammalian models can fill this knowledge gap since degeneration of skeletal muscles can be studied without experimental manipulations during amphibian and insect development. In Xenopus laevis metamorphosis, thyroid hormone induces apoptosis of tail muscles in the tadpole , which is associated with the formation of muscle apoptotic bodies (sarcolytes) and chromatin fragmentation. During metamorphosis of Lepidoptera (moths and butterflies), most muscles of caterpillars are destroyed, while a few persistent muscles survive into adulthood . Major insights in moths were obtained from research on the giant intersegmental muscles (ISM) of the abdomen that undergo hormonally induced atrophy prior to eclosion and die after eclosion . ISM PCD, which is triggered by changes in juvenile hormone 20-hydroxyecdysone, does not display a variety of morphological features associated with classical apoptosis, such as chromatin condensation or phagocytosis [15, 16].
In previous studies, we demonstrated that in vivo microscopy could image abdominal muscle development throughout the 5 days of Drosophila metamorphosis [19, 23]. Here we extend our methodology to identify and characterize genes that are involved in the PCD of doomed muscles and the survival of remodeled muscles. We used a macro zoom microscope to screen for candidate gene perturbations that either prevent or induce cell death, while multi-location laser scanning confocal microscopy (LSCM) helped us to elucidate their phenotypic consequences in more detail. Loss of function of TOR, the RNA helicase Rm62, the lysosomal protease cathepsin-L homolog Cp1 and the master regulator of energy metabolism AMPKα caused histolysis of persistent larval muscles. However, none of these gene perturbations affected the survival of newly formed adult muscles. In contrast, reducing the expression of Atrophin inhibited histolysis of a subset of doomed larval muscles. Since this approach was able to identify genes that protect Drosophila persistent muscles from degeneration, it is conceivable that some conserved genes may also promote resilience of mammalian skeletal muscles to PCD when undergoing atrophy.
List of gene perturbations that affect survival of larval abdominal muscles in metamorphosis
Stock ID/TRiP #
Chromator (Chro), control
DIOM cell death
Cysteine Proteinase 1 (Cp1)/Cathepsin L
DIOM cell death
Cysteine Proteinase 1 (Cp1)/Cathepsin L
DIOM cell death
DIOM cell death, delay of DEOM cell death
AMP-activated protein kinase α (AMPKα)/SNF1a
DIOM loss of morphology and degeneration
Inhibition of DEOM cell death
In our gene perturbation experiments, we crossed females of the reporter line MHC-tau-GFP/FM7-GFP; Mef2-GAL4, UAS-histone-mKO/TM6B Tb with males of the UAS-GeneX-cDNA or UAS-GeneX-shRNA lines. From the progeny, we selected non-Tubby prepupae expressing both fluorophores (e.g. MHC-tau-GFP/+; Mef2-GAL4, UAS-histone-mKO/UAS-GeneX-shRNA) for inspection of muscle phenotypes. For convenience, we will refer to these animals as Muscle-GO-GP (GO = Green + Orange live reporter, GP = Gene perturbation). The samples were examined using an Olympus MVX10 fluorescence macro zoom microscope (Olympus, Japan). The UAS-Chro-shRNA construct (TRiP#GL00503, B-36084) when crossed with Muscle-GO-GP displayed no abnormalities of muscles development, eclosion and ability to fly and was therefore used as control throughout this study.
Screening for muscle phenotypes using macro zoom microscopy
Usually, 20 Muscle-GO-GP prepupae were arranged dorsal side up in 4 groups of 5 samples on plastic Petri dishes. We recorded images at daily intervals for up to 6 days or until the adult flies eclosed using an Olympus MVX10 macro zoom microscope equipped with a DP73 digital CCD camera and cellSens acquisition software. Macro zoom microscopes are equipped with zoom objective lenses of high numerical aperture (N.A.) to resolve fine details in large samples at long working distances. We used the MVPLAPO 1× objective lens with an N.A. of 0.25. The fields of view were recorded twice with filters for green and orange fluorescence, at a zoom factor of 1.25, all of which resulted in digital color images (TIF or PNG format) of 2400 × 1800 pixels and a pixel size of 2.41 microns/pixel. Images were stored on a shared network drive to facilitate the visual inspection of phenotypes by multiple observers. To assess the effects of gene perturbations on muscle function we monitored eclosion and the ability to fly in adult flies.
Time-lapse confocal microscopy of metamorphosis
The protocol for sample preparation and time-lapse imaging of Drosophila pupae was previously described . Live samples were collected at the white pupal stage, rinsed with water to remove the fly food from their surface and inspected under a macro zoom fluorescence microscope to confirm expression of both reporter genes. Up to 30 prepupae were positioned dorsal side down on an uncoated 32 mm diameter glass bottom dish (MatTek, Ashland, Massachusetts). The live samples were mounted in CyGEL (Biostatus Ltd, Leicester, UK) to restrict their movement during imaging. Head eversion leads to compression and a posterior shift of the abdomen. As our goal was to view pupal abdominal segments 1–5 during live imaging, prepupae had to be placed in such a way that the anterior border of their 3rd abdominal segment was adjacent to one side of the field of view. A wet tissue was kept around the specimens to maintain humidity levels during imaging. We used the Zeiss LSM 5 Live (Carl Zeiss, Jena, Germany) inverted line scanning confocal microscope equipped with a motorized XY scanning stage to perform multi-location time-lapse imaging. 3D time-lapse image acquisition was carried out for 5 days at 30 minute intervals (240 time points per sample) using a 10×/0.3 EC-Plan-Neofluar M27 air objective, at a scan zoom of 0.5. The two colour channels were recorded sequentially; channel1 with an excitation laser of 488 nm, band path (BP) filter 500–525; channel 2 with 532 nm laser line, BP 560–675. Image stacks containing 35–40 optical slices were collected at 13.2 μm intervals. Each optical slice had a frame size of 1024 × 1024 pixels with a pixel size of 1.25 μm. The manufacturer’s Zen 2008 software was used for image acquisition, with the built-in multi-time series (MTS) macro controlling repetitive scanning of multiple locations. The MTS saved one LSM image file per time point and location. The temperature of the microscope room was set to 22 °C for most experiments. LSCM recordings of samples expressing Atro and AMPKα shRNA were performed at 25 °C. We used a data logger device to record temperature and humidity during imaging. The speed of metamorphosis depends on temperature. Samples imaged by LSCM at 22 °C eclosed on average 107 h after entering head eversion, while this period was reduced to around 75 h at 25 °C.
Image analysis workflow
We previously introduced a pipeline for the visualization and quantification of in vivo microscopy data . Most steps were carried out using custom software tools, unless otherwise indicated. Using the TLM-Converter custom software , we concatenated the image stacks stored in 8-bit LSM format to create one 3D time-lapse ICS file per sample with sizes ranging from 17 to 19 Gigabytes (GB). 3D stacks in ICS format were converted to maximum intensity projections (MIPs) to generate 2D time-lapse images which were saved as multi-page TIFF files. Uncompressed TIFF files of 240 time points had sizes of 737 MB. TIFF files could be compressed over 20 fold using JPEG compression without noticeable degradation in image quality.
Phenotypic analysis of time-lapse image data was performed using the TLM-Explorer tool . Besides experimental parameters such as genotype, time interval and resolution, the user has to define the onset of head eversion (HE) as the temporal reference point for comparing different datasets. HE marks the transition from the larval to the tripartite adult body plan comprising head, thorax and abdomen. The beginning of HE is recognized by rapid movements of abdominal muscles that lead to misalignments between optical slices during 3D imaging and blurring of image projections. Developmental time points are indicated in hours and minutes (h:m) after head eversion (aHE), where negative values refer to the prepupal and positive to the pupal stage.
The image analysis workflow was implemented as a custom tool in the C++. NET framework using the following libraries: FreeImage  handled the processing of multi-page TIFF files and libics the import of ICS files . To prepare videos, we exported annotated frames as JPEG files and built animations using the VirtualDub video processing software. Videos were exported as AVI files and converted to MP4 using the Freemake Video converter software. To create the figures in our manuscript, we used Photoshop CS3 (Adobe) and ACDSee Pro 5 (ACD Systems Int. Inc.).
Statistical data analysis was performed using Excel (Microsoft) and Minitab 16 (Minitab Inc.). Minitab was used to compute the confidence intervals of proportions and produce box-and-whisker plots (boxplots). Excel was applied to calculate proportions and plot bar charts.
Live imaging based screening for genes regulating survival of larval muscles in metamorphosis
Fluorescent microscopy reveals the contrasting fates of dorsal abdominal muscles in prepupae. Dorsal external oblique muscles (DEOMs) are eliminated by programmed cell death, while the dorsal internal oblique muscles (DIOMs) resist histolysis and persist into adulthood (Fig. 1a-c). According to nomenclature in the literature , the most dorsal muscles in the first five abdominal segments A1 to A5 are referred to as DEOM1 and DIOM1, more lateral ones as DEOM2/3 or DIOM2/3. To identify genes that promote cell death of doomed or survival of persistent larval muscles during metamorphosis, we developed a screen that is based on live imaging, targeted gene perturbation and time-lapse image analysis. Females of a master stock containing three transgenes, the muscle specific driver Mef2-GAL4, UAS-histone-mKO to label nuclei and MHC-tau-GFP to visualize the cytoplasm of muscles, were crossed to males carrying UAS effector constructs to drive the expression of transgenic proteins and small hairpin (sh) RNAs for RNA interference (RNAi). The phenotypic effects of muscle specific gene perturbation were assessed by in vivo microscopy in two steps as previously described . First, to screen for interesting phenotypes, we monitored live muscle development in a minimum of 20 specimens per genotype at daily intervals using a fluorescence macro zoom microscope. To evaluate effects on muscle function, we scored eclosion rates and the ability to fly. Second, gene perturbations resulting in interesting phenotypes were further examined by 3D time-lapse LSCM from the prepupal to pharate adult stage for 4–5 days at 30 min intervals. Time-lapse imaging was performed in multiple locations (up to 30 samples) using a x-y scanning stage. Using custom software, we created time-lapse maximum intensity projections (MIPs) of 3D image stacks and performed time-series image analysis of muscles development during metamorphosis. We had earlier performed a pilot screen with 119 publicly available UAS fly stocks, involving 98 unique genes, which were targeted by 100 RNAi and 19 protein-overexpression constructs  (Additional file 1: Table S1). Five of the gene perturbations specifically interfered with cell death or survival of larval muscles without affecting the development of newly formed adult muscles (Table 1). Macro zoom microscopy revealed that RNAi of the RNA helicase Rm62 (Fig. 1e) and Cysteine proteinase 1 (Cp1) (Fig. 1f), and overexpression of dominant negative TOR induced the removal of DIOMs (Fig. 1g). Atrophin RNAi prevented histolysis of a subset of DEOMs (Fig. 1h). RNAi of AMP-activated protein kinase α subunit (AMPKα) caused the loss of tubular morphology and degeneration of DIOMs (Fig. 1i). About half of the RNAi constructs displayed wildtype morphology and function of muscles, which were indistinguishable from flies without RNAi overexpression. One of these constructs, UAS-Chro-shRNA, was chosen as a control.
Live imaging of muscle histolysis by confocal microscopy
DIOMs are not required for eclosion
Validation of RNAi experiments
All UAS-shRNA constructs used were generated by the Harvard Transgenic RNAi Project (TRiP), which applied bioinformatics methods in the selection of target sequences to avoid off-target effects . The following lines of evidence support our notion that the shRNA constructs that caused cell death related phenotypes silenced the intended targets. Cell death by Cp1 silencing could be induced with two different shRNA constructs targeting different regions of the Cp1 mRNA (Table 1). Using a previously described fluorescent reporter gene expressing Cp1-mKO2 , we were able to show that both Cp1-shRNA were able to abolish muscular fluorescence in prepupae and pupae (Additional file 3: Figure S2). A previous study reported that the same AMPKα shRNA construct used in this study reduced AMPKα mRNA expression in embryos by 95 % when expressed using the maternal MTD-Gal4 driver . A previous genome-wide muscle-specific RNAi screen showed that Rm62 and Atro shRNAs constructs from the Vienna Drosphila RNAi Center (VDRC), which were expressed using the same Mef2-Gal4 driver used in this study, induced flightlessness, an indicator of adult muscle defects . The Rm62 and Atro RNAi constructs of the VDRC and TRiP collections target different regions of the corresponding mRNAs. In addition, consistent with our results, other studies have shown that Atro overexpression contributed to cell death in the nervous system  and was involved in embryonic myogenesis .
TOR and Rm62 protect persistent muscles from histolysis
The cathepsin-L homolog Cp1 promotes DIOM survival in mid-pupation
Silencing of Cp1, the gene encoding the ortholog of the lysosomal Cathepsin-L, led to decay of persistent muscles from +38 h onwards (mean TOD 55.3 ± 13.4 h) (Fig. 5b), around one day later than in the case of Rm62 RNAi (Fig. 4d). In contrast to TOR TED overexpression and Rm62 RNAi, removal of DIOMs was not associated with the fragmentation of muscles into GFP-labelled sarcolytes. Instead, we observed cell shrinkage and a gradual decay of the cytoplasmic tau-GFP fluorescence, indicating proteolysis or leakage of the fluorescent protein (Fig. 6c). Discrete nuclear histone-mKO labelling persisted slightly longer (3–6 h) than tau-GFP, before eventually changing to hazy fluorescence of lower brightness, supporting the notion that Cp1 RNAi, different from TOR TED overexpression and Rm62-RNAi, induced a type of cell death that included rapid protein degradation. In conclusion, Cp1 promotes survival of persistent muscles in mid-pupation. Compared to TOR and Rm62, Cp1 suppresses a different type of cell death that involves cell shrinkage and rapid protein degradation instead of fragmentation.
AMPKα RNAi leads to loss of tubular morphology and degeneration of DIOMs
Atrophin silencing inhibits histolysis of doomed muscles
In early pupation, TOR TED overexpression and Rm62-RNAi induced the breakdown of DIOMs into sarcolytes. The phenotypic similarity to normal histolysis of DEOMs suggests that TOR and Rm62 cell autonomously inhibit cell death of persistent muscles. Evidence in the literature indicates that mTOR can either promote or inhibit apoptosis depending on the biological context , supporting the idea that TOR may protect remodeled muscles from histolysis. However, since TOR TED overexpression may also create unphysiological conditions, we cannot rule out the possibility that the phenotype is an artefact. Circumstantial evidence links the RNA helicase Rm62 to TOR signaling since Rm62 protein was detected in the same complex as the protein Poly, which is a Drosophila homolog of a transcription elongation factor Elp6 and acts as a positive regulator of InR-TOR signaling . Time-lapse analysis uncovered phenotypic differences between the two gene perturbations. Compared to TOR TED, Rm62-RNAi induced histolysis occurred 15 h later and did not delay histolysis of DEOMs. In the context of abdominal muscles, TOR TED did not exactly behave like a dominant negative version of the endogenous TOR protein since TOR-RNAi mainly affected cell size and had negligible effects on cell survival . We can only speculate that the regulation of cell survival and cell growth are mediated by different protein complexes.
Cp1, the fly ortholog of lysosomal Cathepsin-L, has been proposed to mediate autophagic cell death since its gene expression is upregulated during salivary gland cell death . Contrary to this prediction, our results showed that Cp1 acts as a suppressor of cell death in DIOMs. In contrast to the phenotypes caused by TORTED overexpression and Rm62 RNAi, cell death occurred in a later stage of metamorphosis and was not associated with the formation of sarcolytes. Prior to DIOM disappearance, fluorescence of the tau-GFP and histone-mKO reporter decreased and became diffuse, suggesting that Cp1-RNAi induced protein and chromatin degradation inside muscles rather than their fragmentation into sarcolytes. A plausible explanation for the loss-of-function phenotype is that Cp1 may act to break down PCD-promoting proteins that start to accumulate in persistent muscles towards late metamorphosis, thus ensuring that DIOMs only degenerate after eclosion.
Previous studies have shown that autophagy promoted cell death in tissues like the midgut in Drosophila metamorphosis . Since TORTED is a potent inducer of autophagy in the fat body , activation of autophagy may trigger cell death of wildtype DEOMs and TOR TED overexpressing DIOMs. In a parallel study, we found that RNAi of 5 autophagy related genes (Atg5, Atg9, Atg12, Atg17, and Atg18) caused inhibition of atrophy during the remodeling of DIOMs . We examined 25 time-lapse datasets of pupae expressing shRNAs against Atg9 (7) and Atg5, Atg12 and Atg18 (6 each). However, we unable to find evidence that silencing of these genes inhibited or blocked histolysis of DEOMs, as we demonstrated for Atro-RNAi and TORTED overexpression. Hence, in agreement with a previous study , autophagy does not appear to play a role in mediating cell death of larval muscles.
Previous studies have shown that ecdysone receptor signaling cell-autonomously triggers apoptosis in DEOMs . Our results corroborate previous reports in rats and moths, which concluded that PCD in muscles did not display a variety of morphological features observed in classical apoptosis, such as phagocytosis or DNA fragmentation [10, 16]. Although we observed chromatin condensation prior to and during histolysis, we did not find evidence for nuclear fragmentation that can be observed in mono-nuclear non- muscle cells. In addition, the intensity and localization pattern of histone-mKO in sarcolytes showed remarkable stability, arguing against strong proteolytic activities which are commonly associated with apoptosis. Similarly, non-nuclear fusion proteins like tau-GFP, Cp1-mKO2 and Pros35-mOrange2 provided stable fluorescent labeling of sarcolytes. Our results are not necessarily in conflict with studies in Manduca which showed that PCD in ISMs involved the increased activation of the autophagy-lysosomal and ubiquitin-proteasome pathways . A tight control over proteolysis and, probably, lipolysis may ensure that sarcolytes continue to function as amino acid reservoirs that help to feed the developing fly during pupation.
So far, little is known about the genes that mediate DEOM PCD downstream of the ecdysone receptor. Our results suggest that Atrophin may play a role in promoting muscle degeneration. Consistent with previous reports that Atro overexpression leads to neuronal cell death [37, 43], our data show that Atro knockdown can inhibit histolysis of doomed larval muscles. In both neurons and muscles, Atro appears to show genetic interactions with TOR signaling. While TOR TED overexpression enhances neurodegeneration elicited by Atro gain-of-function , both Atro-RNAi and TOR TED overexpression delay histolysis of DEOM1s. The response of DEOMs to Atro silencing was cell-specific. Histolysis of all DEOM1s was delayed, while destruction of DEOM2s occurred at the normal time. Moreover, a subset of DEOM1s, particularly in the anterior abdominal segments was resistant to histolysis and survived until the end of metamorphosis. The reason for this specificity remains to be elucidated. Since Atro protein acts as a transcriptional co-repressor that recruits histone deacetylases , it may repress the transcription of pro-survival genes in doomed muscles. Given that loss of Atro is not sufficient to block PCD in all DEOMs, other redundant transcriptional regulators may repress yet to be discovered target genes.
Live microscopy of Drosophila metamorphosis is a powerful tool to study the processes and genes that promote and prevent cell death in muscles. Candidates of protective genes include Rm62, Cp1, TOR and AMPKα, all of which were discovered in a pilot screen involving 98 preselected unique genes. Time-lapse image analysis provides detailed insights into muscle degeneration that would be difficult to obtain from the study of fixed tissues. Cellular parameters like chromatin condensation, size, shape and fragmentation can be continuously monitored. For different genotypes, we can compare the time of death of different populations of muscles. Using a pilot screen, we identified 5 gene perturbations that could delay or block cell death of doomed muscles, or induce premature histolysis of persistent muscles. Atrophin is so far the only gene, whose silencing can rescue a subset of DEOMs until the end of metamorphosis. Remodeled muscles undergo atrophy. Interestingly, none of the perturbations that induced cell death of remodeled muscles showed discernible effects on the survival of newly formed adult muscles, suggesting that this approach may help find genes that protect muscles undergoing atrophy from PCD. Unbiased genome-wide reverse genetic screens in the future promise to identify more genes involved in myogenic PCD during insect metamorphosis, some of which may be evolutionarily conserved and relevant for human muscle wasting.
M.W. and W.C.P. were affiliated with the Bioinformatics Institute (BII) until 2013, YK until 2015. We thank the TRiP at Harvard Medical School (NIH/NIGMS R01-GM084947) for providing transgenic RNAi fly stocks used in this study. We thank the Bloomington Drosophila stock centre for providing fly stocks. We thank Hwee Kuan Lee and Lin Feng for providing academic advice to YK. We thank the IMCB (ACC and SC labs) for the use of Drosophila facilities.
The study and publication costs were funded by the BII, Agency for Science, Technology and Research (A*STAR), Singapore.
Premature cell death of persistent muscles. Overexpression of dominant negative TOR (TOR TED), and RNAi of Rm62 and Cp1 induce histolysis of DIOMs, which persist throughout metamorphosis until adulthood in controls (top left). The time-lapse frames show the development from the prepupal (−6 h) to the late pupal stage (+100 h). Muscles are labelled with tau-GFP (green) and histone-mKO (red). Pupae are oriented in an anterior–posterior direction from left to right. Views show part of the thorax and abdominal segments A1 to A4. Note the stability of fluorescently labelled sarcolytes derived from histolyzed muscles. None of the gene perturbations caused obvious cell death of newly formed adult muscles, such as IFMs, heart and abdominal muscles. The time-lapse images were recorded at 22 °C. (MP4 8845 kb)
Silencing of AMPKα leads to loss of tubular morphology and disintegration of DIOMs. Compared to control muscles (left panel), AMPKα shRNA DIOMs gradually lose tubular morphology from +15 h aHE onwards. From around +30 h, DIOMs turn into floating muscle spheroids, indicating a loss of cell attachment. Later, green fluorescence of spheroids rapidly disappears within 30 min intervals (see arrow at +56:30 h), while condensed nuclei stay behind, indicating rapid lysis without formation of sarcolytes. The time-lapse images were recorded at 25 °C. Muscles were labeled with Mhc-tau-GFP (green) and histone-mKO (magenta). (MP4 6582 kb)
Silencing of Atrophin inhibits muscle histolysis in metamorphosis. Atro-RNAi (right panel) delays histolysis of DEOM1s in all abdominal segments. While control DEOMs (left panels) are destroyed at the onset of HE, Atro shRNA muscles persist for up to 10 h into pupation. In abdominal segment A2, DEOMs (arrow) survive until adulthood. The time-lapse images were recorded at 25 °C. Muscles were labeled with Mhc-tau-GFP (green) and histone-mKO (magenta). The scale bars correspond to 200 μm. (MP4 8280 kb)
- 13.Finlayson LH. Normal and Induced Degeneration of Abdominal Muscles during Metamorphosis in the Lepidoptera. Q J Microsc Sci. 1956;s3-97:215–33.Google Scholar
- 30.Puah WC, Wasser M. Live imaging of muscles in Drosophila metamorphosis: Towards high-throughput gene identification and function analysis. Methods. 2015.Google Scholar
- 34.FreeImage [Internet]. Available from: http://freeimage.sourceforge.net/index.htmlGoogle Scholar
- 35.libics [Internet]. http://libics.sourceforge.net/. 2010 [cited 2010 Jul 8]. Available from: http://libics.sourceforge.net/Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.