DRP-1-mediated apoptosis induces muscle degeneration in dystrophin mutants

Mitochondria are double-membrane subcellular organelles with highly conserved metabolic functions including ATP production. Mitochondria shapes change continually through the combined actions of fission and fusion events rendering mitochondrial network very dynamic. Mitochondria are largely implicated in pathologies and mitochondrial dynamics is often disrupted upon muscle degeneration in various models. Currently, the exact roles of mitochondria in the molecular mechanisms that lead to muscle degeneration remain poorly understood. Here we report a role for DRP-1 in regulating apoptosis induced by dystrophin-dependent muscle degeneration. We found that: (i) dystrophin-dependent muscle degeneration was accompanied by a drastic increase in mitochondrial fragmentation that can be rescued by genetic manipulations of mitochondrial dynamics (ii) the loss of function of the fission gene drp-1 or the overexpression of the fusion genes eat-3 and fzo-1 provoked a reduction of muscle degeneration and an improved mobility of dystrophin mutant worms (iii) the functions of DRP-1 in apoptosis and of others apoptosis executors are important for dystrophin-dependent muscle cell death (iv) DRP-1-mediated apoptosis is also likely to induce age-dependent loss of muscle cell. Collectively, our findings point toward a mechanism involving mitochondrial dynamics to respond to trigger(s) of muscle degeneration via apoptosis in Caenorhabditis elegans.

Importantly, the main actors of mitochondrial fission and fusion have alternative functions separate from mitochondrial dynamics. It is well established that DRP-1 participates in peroximal division 27,28 and in the maintenance of the morphology and the function of the Endoplasmic Reticulum (ER) 29,30 . In C. elegans, DRP-1 can contribute to apoptotic processes independently from mitochondrial fission 31 . Furthermore, both DRP-1 and OPA1 allow for mitochondrial cristae remodeling separately from mitochondrial dynamics [32][33][34] . The alternative functions of OPA1 also include lipolysis regulation 35 . The pro-fusion protein MFN2 is an essential component of the contact sites between mitochondria and ER and contributes to mitochondrial Ca 2+ uptake and ER morphology maintenance 36 . Additionally, levels of MFN2 are regulated by ER stress and can affect the ER-Unfolded Protein Response (ER-UPR) and in turn, cell death 37,38 . Furthermore, axonal mitochondrial transport requires MFN2 but not mitochondrial fusion 39 .
Strength and function of muscle tissues are highly dependent on the energy produced by mitochondria. Compromised muscle tissues, due to deterioration and loss of muscle cells, and leading to muscle weakness, are hallmarks of muscle degeneration. The primary causes of muscle degeneration are various. For instance, it can be due to numerous genetic mutations causing muscular dystrophies. Muscle degeneration can also be caused by aging processes, and is called sarcopenia. To study muscle degeneration, we used the dys-1(cx18);hlh-1(cc561ts) C. elegans mutant, which exhibits progressive dystrophin-dependent muscle degeneration reflected by muscle weakness and dramatic loss of muscle cells 40 . DYS-1 is the homolog of dystrophin in mammals and mutations in this gene cause Duchenne Muscular Dystrophy in Human 41 . HLH-1 is a homolog of Myogenic Differentiation 1 (MyoD), a muscle transcription factor 42 . Despite the phylogenetic distance between mammals and C. elegans, the nematode presents striated body wall muscle similar to vertebrate skeletal muscle in terms of function and structural components; however, the overall architecture is less complex 43,44 . Mitochondrial fragmentation increases in muscle cells during dystrophin-dependent muscle degeneration in various animal models including nematode 45 , zebrafish 45 , mouse, which exhibits abnormalities of mitochondrial morphology and density as well as an up-regulation of DRP-1 levels 46,47 , dog 48 and in biopsies from Duchenne Muscular Dystrophy patients 49 . Fragmentation of mitochondria is also observed during muscle aging in nematode 50 , fly 51 and mouse 52 .
We hypothesized that genetic manipulations of the main actors of mitochondrial dynamics can improve muscle health. We found that decreasing mitochondrial fission or increasing mitochondrial fusion can partially rescue mitochondrial dynamics, reduce muscle degeneration and improve the mobility of dys-1(cx18);hlh-1(cc51ts) mutant worms. Conversely, increased mitochondrial fusion can enhance the muscle degeneration phenotype of dystrophic nematodes. Interestingly, we showed that DRP-1 functions mainly in the apoptosis pathway downstream of the caspase 3 to impact muscle degeneration and that others apoptotic executors are important for muscle health. Finally, apoptosis induced by DRP-1 seems also to be implicated in aging-dependent muscle loss. Collectively, our results point toward a novel mechanism by which, DRP-1-mediated apoptosis provokes muscle degeneration.

Results
Mitochondrial fragmentation increases upon dystrophin-deficiency. Mitochondrial shapes undergo continual changes through the combined actions of fission and fusion events rendering mitochondrial network very dynamic. In C. elegans, fission is mediated by DRP-1 7 , inner membrane fusion by EAT-3 (OPA1 homolog) 53 and outer membrane fusion by FZO-1 (MFN1 homolog) 8 (Fig. 1A). As C. elegans is a transparent organism, it is possible to visualize and evaluate mitochondrial network shape of muscle cells in lived worms 54 . In order to visualize mitochondrial morphologies, we used the ccIs4251 transgene, a genomic integrated GFP construction addressed specifically to the mitochondrial matrix and nuclei of the C. elegans body wall muscle cells (Fig. 1B-I) 55 . As previously published by Yang et al. 54 , the index of circularity of individual mitochondrion (a mathematical estimation of circular shapes) was calculated for each of the tested conditions to evaluate mitochondrial dynamics. A value of 1 for the index of mitochondria circularity means that mitochondria are perfectly circular and a value of 0 means that mitochondria are perfectly flat. In wild-type C. elegans muscle cells, tubular and circular mitochondria were in equal proportions (Fig. 1B) as reflected by a mitochondrial circularity index of around 0,5 (Fig. 1J) 54 . dys-1(cx18);hlh-1(cc561ts) mutant worms, which present progressive muscle degeneration, exhibited mitochondrial dynamics perturbations with more fragmented mitochondria in muscle cells than wild-type worms (Fig. 1B,F) as reflected by a higher index of mitochondria circularity compared to that of wildtype worms (0,75 +/− 0,014 in dys-1(cx18);hlh-1(cc561ts) mutant vs 0,52 +/− 0,009 in wild-type worms) (Fig. 1J). By comparing mitochondrial aspect ratio (major axis/minor axis of each mitochondrion), mitochondrial elongation (1-(minor axis/major axis of each mitochondrion) and mitochondrial size, we also found that dys-1(cx18);hlh-1(cc561ts) mutant worms presented shorter and smaller mitochondria than that of wild-type worms (Fig. S1A-D). Furthermore, dystrophic worms exhibited a less connected mitochondrial network than wild-type animals (Fig. S1D).

DRP-1 induces dystrophin-dependent muscle degeneration and locomotion defects.
To analyze the impact of mitochondrial dynamics manipulation on muscle health in dys-1(cx18);hlh-1(cc561ts) mutant worms, we further quantified muscle degeneration and estimated muscle activity. In C. elegans, the number of body-wall muscle cells is fixed from one individual to the other (95 muscle cells). By contrast to mammals, C. elegans muscle cells are mononuclear and do not fuse. Moreover, as C. elegans lacks satellite cells and therefore, regenerative capabilities, each muscle cell can be individually followed in vivo throughout its whole life 59,60 , and muscle degeneration can be quantified by actin staining allowing to visualize and to count abnormal or statistically significantly different from the mean obtained in the control condition. Representative confocal images of muscles in wild-type worms of: (E) actin network by phalloïdin staining; (F) mitochondrial network and nuclei (n) organization visualized with the ccIs4251 transgene in C. elegans body wall muscle cells; (G) apoptotic nuclei visualized by TUNEL staining and (H) the merge of the phalloidin, mitochondrial and TUNEL staining. (I-L"') Representative confocal images of body wall muscle cells of dys-1(cx18);hlh-1(cc561ts) mutant worms presenting different degrees of muscle degeneration. (I-I"') actin network by phalloïdin staining; (J-J"') mitochondrial network visualized with the ccIs4251 transgene in C. elegans body wall muscle cells; (K-K"') apoptotic nuclei visualized by TUNEL staining; (L-L"') merge of the phalloidin, mitochondrial and TUNEL staining. All the experiments were performed in L4 + 3 day-old worms. Yellow arrows indicate circular mitochondria. White arrow indicates apoptotic nuclei. Scale bar 10 μm. n indicates nuclei. (x630).
Alterations of the mitochondrial network are early events of dystrophin-dependent muscle degeneration. We next wondered whether mitochondrial dynamics perturbations or apoptosis appears first in dystrophin-dependent muscle degeneration. We performed actin and TUNEL staining on dys-1(cx18);hlh-1(cc561ts) mutant or wild-type worms carrying the ccIs4251 transgene. In wild-type worms, tubular and circular mitochondria were in equal proportions and organized along the actin network, which is arranged in filaments along the cell (Fig. 2E-H) and we could not detect any TUNEL positive cell (Fig. 2G). In dys-1(cx18);hlh-1(cc561ts) mutant worms, different stages of actin network perturbations were observed, likely reflecting various degrees of muscle degeneration. The actin staining can reveal an apparent totally preserved actin network or the apparition of wavelets or a more or less noticeable aggregation of actin. Muscle cells with mutated dystrophin and unperturbed actin network ( Fig. 2I-I') could be associated with either wild-type-like mitochondria (Fig. 2J) or mitochondria that appeared to be fragmented (Fig. 2J') and we could not detect any TUNEL staining under these conditions (Fig. 2K-K'). It is noteworthy that weak actin perturbations (Fig. 2I") were always associated with increased mitochondrial circularity in dys-1(cx18);hlh-1(cc561ts) mutant worms (Fig. 2J") but without any signs of apoptosis (Fig. 2K"). Finally, strong actin perturbations due to mutated dystrophin ( Fig. 2I"') were concomitant with circular mitochondria (Fig. 2J"') and TUNEL positive staining ( Fig. 2K"'). Together, these observations suggest that mitochondrial dynamics dysfunction is an early event in the molecular mechanisms leading to progressive muscle degeneration. Changes in mitochondrial morphologies are then followed by actin network perturbations and will eventually lead to apoptosis.
Cleavage of DRP-1 by CED-3 is required for dystrophin-dependent muscle degeneration but is dispensable for regulating mitochondrial fission. The links between mitochondrial dynamics and cell death have been extensively studied 16,31,56,61,62 . In C. elegans, the importance and the exact roles of DRP-1 in developmental apoptosis are still controversial 31,56 . A fraction of DRP-1 can be cleaved in vitro by the caspase CED-3, leading to the formation of DRP-1 complexes containing both, full-length DRP-1 and DRP-1 119-712 , which are suspected to play a role in apoptosis during worm development. This cleavage by CED-3 may redirect DRP-1 (n = 60 worms). (H) Quantification of worm trashing in each of the indicated strains (n = 45 worms at least). All the experiments were performed on L4 + 3 day-old worms. One-way ANOVA, Tukey's multiple comparisons test. (I) Number of abnormal C. elegans body wall muscle cells by two quadrants quantified by phalloïdin staining in wild-type worms; drp-1(tm1108) mutant worms; drp-1(tm1108) mutant worms expressing drp-1(+); drp-1(tm1108) mutant worms expressing drp-1(D118A) at day 2, 9, 17 and 19 of adulthood of the nematode (n = 63 worms at least). Student test. Data represent the mean obtained by pooling at least three independent assays. Errors bars represent SEM *p < 0.05 **p < 0.01 ***p < 0.001 ****p < 0,0001 ns. indicates that the mean is not statistically significantly different from the mean obtained in control worms.

Cleavage of DRP-1 by CED-3 is required for age-dependent muscle degeneration.
Physiological processes such as aging can also induce a type of muscle degeneration called sarcopenia. Sarcopenia, which results in the gradual loss of skeletal muscle mass in the elderly, is a major factor determining the decline in overall health and loss of autonomy of the aging population. In wild-type C. elegans, we revealed the spontaneous apparition of more than two abnormal muscle cells by muscle quadrant at days 17 and 19 of adulthood (Fig. 3I) accompanied by a gradual loss of locomotion as the animal ages (Fig. S5). The drp-1(tm1108) mutant worms showed impaired mobility early in adulthood life, which could be explained by energy production impairment 63,64 . Importantly, in absence of DRP-1, the apparition of abnormal muscle cells, over aging, was a very rare event (Fig. 3I). Yet, locomotion defects due to age cannot be improved by the absence of DRP-1. Our data demonstrate that DRP-1 is involved in both dystrophin-dependent and age-dependent muscle degeneration but is unlikely to be the key factor for locomotion defects that are observed over aging. To understand if the apoptotic functions of DRP-1 are important for age-dependent muscle degeneration as it is for dystrophin-dependent muscle degeneration, we analyzed drp-1(tm1108) mutant worms expressing either the wild-type drp-1 transgene (drp-1(+)) or the drp-1 transgene with the CED-3 cleavage site mutated (drp-1(D118A)) supposed to be inefficient in apoptosis (Fig. 3A). At day 17 and 19 of adulthood, drp-1(tm1108) mutant worms with drp-1(+) transgene exhibited apparition of abnormal muscle cells to the same extent as wild-type worms, demonstrating a role for DRP-1 in age-related muscle degeneration. At day 17 of adulthood, drp-1(tm1108) mutant worms expressing the drp-1(D118A) transgene presented very few abnormal muscle cells (Fig. 3I). By contrast, at day 19 of adulthood, drp-1(tm1108) mutant worms expressing the drp-1(D118A) transgene exhibited nearly the same level of muscle degeneration than wild-type worms (Fig. 3I). Collectively, our data suggest a role for DRP-1 through apoptotic-like processes in both dystrophin-dependent and age-related muscle degeneration. wild-type worms or (N) dys-1(cx18);hlh-1(cc561ts) mutant worms fed with ced-1 RNAi; of (O) wild-type worms or (P) dys-1(cx18);hlh-1(cc561ts) mutant worms fed with psr-1 RNAi; of (Q) wild-type worms or (R) dys-1(cx18);hlh-1(cc561ts) mutant worms fed with nuc-1 RNAi and of (S) wild-type worms or (T) dys-1(cx18);hlh-1(cc561ts) mutant worms fed with crn-6 RNAi. Scale bar 10 μm. (x630). n indicates nuclei. (U) Quantification of mitochondrial circularity in each of the indicated conditions (n = 27 worms at least). Note that quantitative PCR data showed that nuc-1 and crn-6 mRNA expression were reduced, in all strains, by ~50% and 70% upon nuc-1 and crn-6 RNAi treatments, respectively (Fig. S6). All the experiments were performed on L4 + 3 day-old worms. One-way ANOVA, Dunnett's multiple comparisons test. Data represent the mean obtained by pooling at least three independent assays. Errors bars represent SEM **p < 0.01 ***p < 0.001 ****p < 0.0001 ns. indicates that the mean is not statistically significantly different from the mean obtained in control conditions. www.nature.com/scientificreports/ The canonical apoptosis pathway mediates dystrophin-dependent muscle degeneration. Having established that the functions of DRP-1 in apoptosis play a key role in dystrophin-dependent muscle degeneration, we asked whether other executors of apoptosis could be important. For this, we performed RNAi experiments against several executors of cell death and tested the effects of each RNAi constructs on mitochondrial morphology and the number of abnormal cells in dys-1(cx18);hlh-1(cc561ts) mutant worms. We thus down-regulated CED-3, required for the execution of apoptosis via its activation by CED-4 (APAF-1 homolog in mammals) and for DRP-1 to act in apoptosis (Fig. 4A-A') 65,66 . We have also tested the effects of RNAi-mediated down-regulation of five actors implicated in the three sub-pathways involved in DNA degradation. First, WAH-1 (AIF homolog in mammals) and CPS-6 (EndoG homolog in mammals) which, upon cell death signal, are released from mitochondria and can interact together to translocate to the nucleus where they participate to DNA degradation 67 . CPS-6 can also interact with CRN-1, CRN-4, CRN-5 and CYP-13 to form the degradeosome 68 . Second, CRN-2 likely to encode an homolog of the TatD nuclease (TATDN1 homolog in mammals) and predicted to be mitochondrial 68 . CRN-2 acts in the same pathway as CRN-3 68 . Third, NUC-1 and CRN-6 (two DNase II homologs in mammals) that seem involved in the degradation of DNA debris from apoptotic cells in late stage of apoptosis 68,69 . Concerning corpse engulfment, important for the phagocytic process 70 , we have tested the effects of knocking-down CED-1, which is involved in cell-corpse recognition with CED-6 and CED-7 71 and PSR-1, which participates in the migration of engulfing cell 72,73 .

Discussion
Apoptotic muscle fibers have been frequently associated with Duchenne Muscular Dystrophy 76-79 and sarcopenia [80][81][82][83] in various species including Human. However, the exact molecular mechanisms leading to muscle apoptosis upon dystrophin deficiency or aging remain unclear. Here, we revealed a DRP-1-dependent apoptotic www.nature.com/scientificreports/ 11 Scientific RepoRts | (2018) 8:7354 | DOI:10.1038/s41598-018-25727-8 mechanism that participates in dystrophin-dependent and age-dependent muscle degeneration. DRP-1 is the principal and universal actor of mitochondrial fission. In mammals, mitochondrial fragmentation induced by DRP-1 can trigger apoptosis 84 . In C. elegans, a previous study suggested a marginal role for DRP-1 in apoptosis during development using the mutation drp-1(tm1108) 31 , whereas another study showed that the dominantly interfering mutation drp-1(K40A) caused mitochondria to form large blebs and induced excessive apoptosis in embryos 56 . Our study reveals that DRP-1 can play a key role in apoptosis that takes place during cellular stresses provoked either by the loss of dystrophin function or by aging. Our data also demonstrate that EAT-3 and FZO-1 overexpression in dystrophin deficient worms can reduce both mitochondrial fragmentation and muscle degeneration. A recent study showed that in mice, muscle-specific deletion of OPA1 (EAT-3 homolog in C. elegans) induces a precocious aging phenotype with muscle loss and weakness 85 and a reduced muscle mass has been also reported in mice ablated for both Mfn1 (FZO-1 homolog in C. elegans) and Mfn2 85 . These observations are nicely corroborating our findings and the fact that levels of mitochondrial fusion greatly impact muscle degeneration. Our data also suggest that manipulating mitochondrial fusion is sufficient to decrease apoptosis in dystrophic worms. In C. elegans, the direct links between mitochondrial dynamics and developmental apoptosis are highly controversial 25,26 ; our study suggests that, upon cellular stress such as dystrophin deficiency, the major actors of mitochondrial dynamics in worms, namely DRP-1, EAT-3 and FZO-1, can directly interplay with the apoptotic pathways.
We also showed that DRP-1 cleavage by the pro-caspase CED-3 is required for dystrophin-dependent and age-dependent muscle degeneration. Hence, our study demonstrates that DRP-1 sits downstream of CED-3 and strongly suggests that DRP-1 acts mostly via apoptosis rather than through mitochondrial fission to induce muscle cell death. In mammals, DRP-1-induced outer membrane fragmentation allows for release of mitochondrial pro-apoptotic factors such as cytochrome c, EndoG (CPS-6 in worms) or AIF (WAH-1 in worms) [13][14][15]86 . In C. elegans muscle cells, cytochrome c could be released from mitochondria to the cytosol during muscle degeneration 45,74 but its role in activating APAF-1 (CED-4 homolog in worms) to form the apoptosome is unlikely to be conserved 87,88 . Strong evidence suggests that both WAH-1 (AIF homolog in mammals) and CSP-6 (EndoG homolog in mammals) can be released from the mitochondria to the nucleus to degrade DNA upon apoptotic stimuli in C. elegans 67,89 . Our data emphasize a function for WAH-1/CSP-6 in dystrophin-dependent muscle degeneration potentially downstream of DRP-1. One possibility is that DRP-1, once activated by CED-3 induces mitochondrial outer membrane permeabilization that will, in turn, allows for WAH-1/CSP-6 release from the mitochondria and translocation into the nucleus. However, we cannot exclude that DRP-1 acts in a parallel pathway of WAH-1/CSP-6. Our data strongly suggest that DRP-1 does not participate in corpse engulfment during apoptosis and is likely to act upstream of NUC-1 and CRN-6. Collectively, our data converge toward a mechanism that underlies DRP-1-dependent mitochondrial pathways activated downstream of CED-3 to impact dystrophin-dependent muscle degeneration. The cleavage of DRP-1 by caspase CED-3 seems to be restricted to C. elegans 31 suggesting that it could be an ancestral role of CED-3 that was lost over evolution. However, we also found that diminution of CED-3 expression can decrease muscle degeneration of a dystrophin C. elegans mutant in a DRP-1-independent manner. Our data are consistent with the fact that maintenance of C. elegans muscle integrity during aging depends on CED-3 90 . Interestingly, skeletal muscles of aged rats exhibit a high level of activated caspase-3 91 and muscles of Duchenne Muscular Dystrophy patients present an increase of both caspase-3 expression 92 and activity 93 . Collectively, the role of the caspase-3 in muscle degeneration appears to be likely conserved among species.
Having established that DRP-1 acts mostly in apoptosis to impact muscle degeneration upon dystrophin deficiency, it would be interesting to investigate whether others molecular mechanisms than cleavage by caspase-3 can regulate DRP-1 activity 94 . Control of intracellular concentration of Ca 2+ ([Ca 2+ ] i ) homeostasis is critical for the maintenance of muscle contraction and relaxation. Elevated [Ca 2+ ] i increases production of reactive oxygen species (ROS) 95 , activates calpain 96 , impairs autophagy 97 , increases mitochondrial Ca 2+ accumulation 96 and induces cell death 98,99 . Intriguingly, a drastic increase in the [Ca 2+ ] i in myofibers and myotubes deficient in dystrophin was observed in several studies 100,101 . In mdx mice, abnormal elevation of [Ca 2+ ] i is due to SERCA (sarco(endo)plasmic reticulum Ca 2+ ATPase) inhibition suggesting an importance of the ER-dependent regulation of Ca 2+ in Duchenne Muscular Dystrophy 102,103 . Moreover, muscle degeneration of the dys-1(cx18);hlh-1(cc561ts) mutants worms is likely to be a calcium-dependent 104,105 . Calcium is also a well-known activator of DRP-1. For instance, ER-calcium release induces uptake of calcium by mitochondria and DRP-1 activation allowing for calcium-dependent mitochondrial fission 106 . Calcineurin, a calcium-dependent phosphatase, dephosphorylates DRP1 and promotes mitochondrial fragmentation and cell vulnerability to apoptosis 107,108 . In neurons of C. elegans, phosphorylation of DRP-1 by Ca/calmodulin-dependent kinase II (CaMKII) inhibits DRP-1 activity 109 . One possibility is that the loss of dystrophin function could induce changes in intracellular calcium that would activate DRP-1 and cell death.
In conclusion, our findings point to mitochondrial dynamics as an early signaling hub that controls cell death in dystrophin-dependent muscle degeneration. A better understanding of the role of DRP-1 in muscle degeneration could allow in the long term to identify new targets for treatments of muscular dystrophies or to improve muscle degeneration in the elderly.

Materials and Methods
Strains. Most of the strains used were obtained from the Caenorhabditis Genetic Center with the exception of the drp-1(tm1108) strain provided by the National Bioresource Project for the Nematode to Sylvia Lee's lab (Cornell University-Ithaca, NY-USA). N2 was used as the wild-type (WT) strain. The dys-1(cx18);hlh-1(cc561ts) 40 , dys-1(cx18);hlh-1(cc561ts);drp-1(tm1108) and the dys-1(cx18);hlh-1(cc561ts);ccIs4251 strains were constructed using standard genetic methods. The drp-1(tm1108) strain, which has no detectable DRP-1 protein expression, is likely to represent a strong loss-of-function or null mutant 31 . The integrated ccIs4251 transgene contains Pmyo-3::Ngfp-lacZ (nuclear localization of GFP) and Pmyo-3::Mtgfp (mitochondrial localization of GFP) 55  All strains were maintained at 15 °C on standard NGM (Nematode Growth Medium) agar plates seeded with Escherichia coli strain OP50 or with HT115 bacteria for RNAi experiments. fzo-1(+) transgene under control of endogenous fzo-1 promoter. A first 3.8 Kbp genomic fragment containing the 163 bp sequence upstream, the 2.8 Kbp coding region, and the 762 bp sequence downstream of the fzo-1 coding region was amplified from genomic DNA by PCR. A second 3.6 Kbp genomic fragment containing the 113 bp sequence upstream, the 2.8 Kbp coding region, and the 702 bp sequence downstream of the fzo-1 coding region was amplified from the first fragment described above.

Generation of drp-1, eat-3, fzo-1 constructs and transgenic lines.
For generation of transgenic animals: 10 ng/μl of transgene (drp-1(+) or drp-1(D118A) or eat-3(+) or fzo-1(+)); 2,5 ng/μl of the co-injection marker pCFJ90 (Pmyo-2::MCherry::unc-54utr) and 116,5 ng/μl Bluescript plasmid filler DNA were microinjected into the gonads of adult wild-type, dys-1(cx18);hlh-1(cc561ts);drp-1(tm1108) or dys-1(cx18);hlh-1(cc561ts) mutant worms using standard methods 110 . F1 progeny were selected on the basis of MCherry fluorescence. Individual F2 animals were isolated to establish independent lines. RNA interference. RNAi bacteria clones from the commercial C. elegans RNAi collection (Ahringher laboratory-Gene Service Inc) were cultured overnight in LB medium supplemented with 100 mg/ml ampicillin and 12.5 mg/ml tetracycline (LBAT) at 37 C. The day after, 0.01 volume of this pre-culture was added in 1 volume of LBAT and incubated at 37 C. At 0.6-0.8 DO, bacteria were concentrated 5 times and then 300 μL of the bacteria culture was seeded on NGM plates containing ampicillin (100 ug/mL) and tetracycline (12,5 μg/mL). Plates were allowed to dry at room temperature for at least 3 days. Then, plates were stored at 4 C during 2 months utmost. The day before the experiment, 4 mM IPTG was added to induce bacteria overnight at room temperature. Gravid worms were allowed to lay eggs during 6 hours on NGM plates seeded with RNAi bacteria carrying specific dsRNA-expressing bacteria and progeny grew on RNAi plates until the day of experiment. Unless indicated differently, experiments were performed at the age of 3 days after the L4 larval stage (L4 + 3 days). As RNAi efficiency control, dpy-13 RNAi was performed in parallel and resulted in nearly 100% dumpy. As RNAi-negative control, plates seeded with HT115 carrying the empty vector RNAi L4440 were used. Phalloidin staining. Phalloidin staining was carried out as described previously 111 . Synchronized adult (day 2, 9, 17 and 19 of adulthood) animals were fixed in 1 ml of PBS supplemented with 20 μl of 37% formaldehyde, extracted with acetone at −20 °C, and incubated in 3U of phalloidinAlexa Fluor ® 488 (Molecular Probes) for 2 h. Imager Z1 microscope was used for microscopic observations. Only the two most visible quadrants of each animal were counted for the quantification of muscle degeneration.
Thrashing assay. L4 worms were transferred to a plate during 3 days. In the case of transgenic lines, only the MCherry-fluorescent L4 worms (expressing the co-injection marker Pmyo-2::mCherry::unc-54utr) were transferred. Synchronized adult animals (day 2, 9, 17 and 21 of adulthood), were moved onto a fresh 2% agarose 12 wells-plates containing M9 bufferpre-equilibrated at a temperature of 15 °C. Thrashing frequency was measured during the swimming sessions in the liquid. Videos of animal movements were acquired using a Macro Zoom microscope. The videos were recorded for 30 seconds at 30 frames per second. The thrashing frequency of the worms, measured in body bends per 30 seconds, was quantified using the open-source wrMTrck plugin for Fiji software 112 (http://www.phage.dk/plugins/wrmtrck.html). One body bend was defined as a change indirection of www.nature.com/scientificreports/ Mitochondrial morphology analysis. Gravid adults with integrated ccIs4251 transgene were allowed to lay eggs on NGM plates for 6 hours before being removed from the plate. L4 + 3 day-old progeny worms were immobilized in 3.3 mM levamisole on a 2% agarose pad for acquisition. Confocal images of the vulva area of lived paralyzed worms were taken on a Zeiss LSM 510 Meta or Zeiss LSM 800 using a 63X oil objective. The percentage of muscle cells exhibiting interconnecting filaments in between mitochondria blebs was calculated after observation, with the unaided human eye of the experimenter, of each muscle cells with an increase of brightness of the confocal images. To quantify mitochondrial circularity, Fiji software and a homemade macro were used to process and to analyze captured images. To fix threshold, the inverse Fourier transformation (Inverse FFT) was used in order to minimize the noise and to sharpen the contrast between mitochondria and the background. Nucleus of each image was removed by hand (with the unaided human eye). « Analyze particles » command was used to obtain values for circularity (calculated by 4πArea/perimeter²) for each mitochondrion in an image, ignoring mitochondria <5 square-pixels or on the edge of the image. Mitochondrial circularity at 0 refers to a straight line whereas mitochondrial circularity at 1 refers to a perfect circle.
In order to determine the accuracy and consistency of the circularity measurements obtained with Fiji, we have compared the circularity indexes of 96 confocal images measured by visual counting and the circularity indexes measured by the macro processed. No statistical differences were found between the two methods. TUNEL assay. Synchronized L4 + 3 day-old worms were fixed in 1 ml of PBS supplemented with 20 μl of 37% formaldehyde, extracted with acetone at −20 °C, and incubated in 50 μL of TUNEL solution (In Situ Cell Death Detection Kit, Fluorescein Roche) for 1.5 hours at 37 C. After incubation, worms were washed 3 times in PBST for 10 min. Then, phalloïdin staining with phalloïdin Alexa Fluor ® 633 (Molecular Probes) was performed like previously described above. Worms were then mounted in Vectorshield mounting medium with DAPI (Vector Laboratories) and visualized using either a Zeiss Axiophot microscope or Zeiss LSM 710 confocal microscope. Only the two most visible quadrants of each animal were counted for the quantification of TUNEL positives muscle cells.
Statistical analyses. One-way ANOVA was used to test statistical differences between independent groups within the same experiment. Statistical significance was tested with Tukey and Dunnett post tests. Two-tailed unpaired Student's t-test was used to examine direct differences between two independent groups.
No data sets were generated or analyzed during the current study.