Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Overexpression of an ALS-associated FUS mutation in C. elegans disrupts NMJ morphology and leads to defective neuromuscular transmission

Overexpression of an ALS-associated FUS mutation in C. elegans disrupts NMJ morphology and leads... © 2020. Published by The Company of Biologists Ltd | Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 RESEARCH ARTICLE Overexpression of an ALS-associated FUS mutation in C. elegans disrupts NMJ morphology and leads to defective neuromuscular transmission 1, 2 3 4 4,5 3 Sebastian M. Markert *, Michael Skoruppa , Bin Yu , Ben Mulcahy , Mei Zhen , Shangbang Gao , 2 1, Michael Sendtner and Christian Stigloher * ABSTRACT muscles atrophy (Hardiman et al., 2017). ALS is hypothesized to be a synaptopathy (Fogarty, 2019), because the disease usually starts with The amyotrophic lateral sclerosis (ALS) neurodegenerative disorder dysfunction and degeneration of synapses, before axons and dendrites has been associated with multiple genetic lesions, including become dystrophic and the neurons undergo cell death (Chou, 1992). mutations in the gene for fused in sarcoma (FUS), a nuclear- Most cases of ALS (∼90%) are considered ‘spontaneous’– due to a localized RNA/DNA-binding protein. Neuronal expression of the combination of de novo genetic and environmental factors. The pathological form of FUS proteins in Caenorhabditis elegans results remaining∼10% of ALS cases are hereditary and are caused by defined in mislocalization and aggregation of FUS in the cytoplasm, and leads mutations in different genes. A number of cellular defects have been to impairment of motility. However, the mechanisms by which the implicated in the pathophysiology of ALS, including defective mutant FUS disrupts neuronal health and function remain unclear. endosomal and receptor trafficking, as well as changes in autophagy Here we investigated the impact of ALS-associated FUS on motor and axonal transport (Burk and Pasterkamp, 2019). neuron health using correlative light and electron microscopy, Mutations in the gene for FUS (fused in sarcoma) make up electron tomography, and electrophysiology. We show that ectopic approximately 5% of hereditary ALS cases (Kwiatkowski et al., expression of wild-type or ALS-associated human FUS impairs 2009; López-Erauskin et al., 2018; Vance et al., 2009). The FUS synaptic vesicle docking at neuromuscular junctions. ALS-associated protein has a RNA/DNA recognition motif, a putative nuclear export FUS led to the emergence of a population of large, electron-dense, signal, several disorganized domains, and a nuclear localization and filament-filled endosomes. Electrophysiological recording signal at the C-terminus (Kino et al., 2011; Lorenzo-Betancor et al., revealed reduced transmission from motor neurons to muscles. 2014; Murakami et al., 2012). FUS has been associated with a Together, these results suggest a pathological effect of ALS-causing plethora of cellular functions, including translation, splicing, RNA FUS at synaptic structure and function organization. transport, and DNA damage response (Andersson et al., 2008; Lagier-Tourenne et al., 2010; Ratti and Buratti, 2016; Therrien et al., This article has an associated First Person interview with the first 2016; Wang et al., 2018). Under physiological conditions, FUS is author of the paper. predominantly located in the nucleus, but it is able to shuttle between KEY WORDS: C. elegans, Fused in sarcoma, Amyotrophic lateral nucleus and cytoplasm in response to different stimuli (Gal et al., sclerosis, Super-resolution array tomography, Electron tomography, 2011; Vance et al., 2013). Pathological mutations are suggested to Neuromuscular junction result in toxic gain of function, as ALS-associated mutant FUS is prone to cytoplasmic accumulation. The potential role of loss of INTRODUCTION function due to FUS sequestration in the cytoplasm is still being Amyotrophic lateral sclerosis (ALS) is a severe disease of the debated (An et al., 2019). Currently, mechanisms by which mutant locomotor system where motor neurons progressively degenerate and FUS proteins lead to motor neuron degeneration are still largely enigmatic. A prevailing hypothesis is that mutated FUS proteins form irreversible hydrogels that impair ribonucleoprotein (RNP) granule University of Wü rzburg, Biocenter, Imaging Core Facility, Am Hubland, Wü rzburg 2 function (Murakami et al., 2015). 97074, Germany. University Hospital Wü rzburg, Institute of Clinical Neurobiology, Versbacherstraße 5, 97080 Wü rzburg, Germany. Huazhong University of Science In a previous study, several variants of human ALS-associated and Technology, Key Laboratory of Molecular Biophysics of the Ministry of FUS were ectopically expressed in the Caenorhabditis elegans Education, College of Life Science and Technology, Wuhan 430074, China. (MAUPAS 1900) nervous system (Murakami et al., 2012). These Lunenfeld-Tanenbaum Research Institute, Mount Sinai Hospital, 600 University Avenue, Toronto, Ontario M5G 1X5, Canada. University of Toronto, Department of worms died prematurely, and had impaired motility, suggesting Molecular Genetics, Physiology and Institute of Medical Science, 1 King’s College degeneration of the locomotor system through a gain-of-function Circle, Toronto, Ontario M5S 1A8, Canada. mechanism. Strength of these phenotypes was positively correlated *Authors for correspondence: (smarker5@jhmi.edu; christian.stigloher@ with the severity of human ALS caused by each variant, supporting uni-wuerzburg.de) C. elegans as a promising experimental paradigm to interrogate S.M.M., 0000-0001-9069-156X; B.M., 0000-0002-3336-245X; M.Z., 0000-0003- mechanisms by which mutant FUS proteins lead to structural and 0086-9622; S.G., 0000-0001-5431-4628; M.S., 0000-0002-4737-2974; C.S., 0000- functional disruption of the nervous system. Here, we used 0001-6941-2669 behavioral, ultrastructural, and electrophysiological approaches to This is an Open Access article distributed under the terms of the Creative Commons Attribution investigate how ALS-associated FUS impacts the C. elegans License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, neuromuscular system. We show that the organization of the distribution and reproduction in any medium provided that the original work is properly attributed. neuromuscular junction (NMJ) is impacted by human FUS, and Received 16 July 2020; Accepted 27 October 2020 ALS-associated FUS severely impairs functional communication Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 between motor neurons and muscles. These disease mechanisms previous results (Murakami et al., 2012, 2015), we observed a may contribute insights to other forms of ALS, as well as other shortened lifespan of FUS501 animals compared to wild-type (N2) neurodegenerative diseases. animals, as well as a genetically wild-type sibling strain derived from outcrossing FUS501 animals with N2 (referred to as ‘wild- RESULTS type’; Fig. 1A). FUS501 animals were also more likely to die from Expression of ALS-associated FUS in the C. elegans nervous internal hatching of progeny (a phenotype referred to as bagging), system results in reduced lifespan and impaired motility likely due to impairment of the neural circuit that modulates To determine the time window for analysis of the C. elegans ALS- egg-laying (Fig. 1B). Consistent with motor dysfunction, FUS501 model, we examined the hermaphrodite worms that stably and pan- animals showed reduced motility in swimming assays (Fig. 1C). neuronally express a pathological form of FUS, FUS C-terminal Animals expressing wild-type human FUS (referred to as FUSwt deletion (referred as FUS501 henceforth) (Murakami et al., 2012, henceforth) displayed a milder reduction of lifespan, but did not 2015) for effects on lifespan and motor function. Consistent with display a motor defect at the ages tested (Fig. 1A–C). Fig. 1. C. elegans expressing pan-neuronal human FUS501 exhibit lifespan and motility defects. (A) Animals expressing human FUS501 have a shorter median and maximum lifespan compared to wild-type animals, as well as animals expressing wild-type human FUS (n≥102 deaths per genotype; log-rank test, P<0.0001). Animals expressing wild-type human FUS also displayed a mild lifespan defect (compared to N2, n≥160 deaths per genotype; log-rank test, P=0.0014). (B) Internal hatching of progeny occurred more frequently in FUS501 animals than in non-transgenic animals and animals expressing wild-type human FUS. (C) C. elegans expressing FUS501 have defective swimming compared to wild-type, whereas animals expressing wild-type human FUS are able to swim similarly to N2 controls (n≥14 per genotype; one-way ANOVA followed by Tukey’s multiple comparisons, *P=0.0112; ****P<0.0001). (D) The alleles of FUS used in this study. Wild-type human FUS contains regions enriched with certain amino acids (one-letter code given), an RNA recognition motif (RRM) and a nuclear targeting signal at the C-terminus (NLS). Mutations in this region are often linked to ALS. Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 Based on these results, we chose to proceed with comparative 0.0049; Fig. 6A; Table 2). For dense core vesicles (DCVs), the ultrastructural and functional analyses for FUS501 and wild-type diameter was similar between wild-type and FUSwt (wild-type: animals on day 3 of adulthood. We reasoned that a pathological 45.6±10.6 nm; FUSwt: 47±10.6 nm; Table 2), but was significantly synaptic phenotype should already be pronounced prior to smaller in FUS501 (35.6±5.6; P-values: FUS501 versus FUSwt: widespread degeneration, before causing secondary phenotypes 0.0074; FUS501 versus wild-type: 4.0e-08; Table 2). DCVs made that could make analysis more difficult. At this age, pathological up 4.5% of all vesicles in wild-type, 4.9% in FUSwt, and 11.4% in features, including the FUS protein localization, aggregation, and FUS501 (P-values of two-sample t-tests: FUSwt versus wild-type: motility defects were prominent, while animal survival was still 0.83; FUS501 versus FUSwt: 0.012; FUS501 versus wild-type: high (Fig. 1A,C). <0.00001). Thus, in FUS501 worms, vesicle pools in NMJs contain a higher proportion of DCVs that are smaller in size (Fig. 6A). Large endosomes with electron-dense inclusions are enriched at the neuromuscular junctions (NMJs) of FUS501 FUS disrupts the organization and docking of vesicles adults at the NMJ In order to assess whether FUS501 motility defects reflect synaptic We further compared the distance between vesicles and the center of structural defects, we used electron tomography to map the the presynaptic dense projection. In animals expressing FUSwt and organization of excitatory and inhibitory NMJs in animals FUS501, both clear- and dense-core vesicles were closer to the expressing wild-type human FUS (FUSwt) and FUS501, as well presynaptic dense projection than wild-type controls (Fig. 6B; as the non-transgenic, wild-type animals that share otherwise the Table 2). For both classes of vesicles, the effect of FUSwt was more same genetic background. pronounced than FUS501 (Fig. 6B; Table 2). To assess how altered Wild-type and FUSwt NMJs were composed of a presynaptic distribution of vesicles may affect their release properties, we dense projection and associated pool of clear and dense core quantified the number of docked vesicles at each synapse. Both vesicles, and rarely contained large endosomes. However, large FUSwt and FUS501 had significantly fewer docked vesicles than endosome-like organelles were enriched at the NMJs of animals wild-type controls (P-values: FUSwt versus wild-type: 0.025; expressing FUS501, and a population of large endosomes FUS501 versus FUSwt: 0.44; FUS501 versus wild-type: 0.011; (>100 nm) was exclusively present in FUS501 worms (Figs 2–4). Fig. 6C; Table 2). Docked vesicles in FUSwt and FUS501 were This was accompanied by a reduction in the smaller (∼50 nm) closer to the center of the presynaptic dense projection than wild- diameter endosomes in FUS501 (Fig. 4A). These endosome-like type controls, similar to the global vesicle distribution described organelles were mostly electron-light, and present in both earlier (Fig. 6D; Table 2). Thus, in FUSwt and FUS501, fewer cholinergic and GABAergic NMJs (Figs 2 and 3). Most vesicles are docked, but those that are docked located closer to the endosomes were spherical, although more complex networks also presynaptic dense projection. occurred (see Fig. 5; Fig. S1). In FUS501 mutants, approximately 41% of endosome-like FUS501 animals exhibit reduced endogenous postsynaptic structures had filamentous, electron-dense inclusions (35/85, from currents 26 tomograms). Wild-type and FUSwt animals also had a similar Given the defects in the structural organization of NMJ in animals percentage of filled endosomes (36% and 41% respectively; 5/12 expressing human FUS, we recorded the functional output of the from 5 tomograms, and 14/39 from 17 tomograms). While ‘empty’ NMJ by patch clamping muscle cells in animals on day 3 of endosomes in each genotype were of a similar diameter (P-values: adulthood. The endogenous postsynaptic currents (enPSCs) reflect FUS501 versus FUSwt: 0.089; FUS501 versus wild-type: 0.24; synaptic vesicle fusion events that are due to either spontaneous FUSwt versus wild-type: 1.0), endosomes in FUS501 worms release or release evoked by endogenous activity in the motor containing electron-dense inclusions were larger than similar filled neurons or upstream circuits, and subsequent detection of endosomes in wild-type and FUSwt animals (Fig. 4B, Table 1). neurotransmitter (acetylcholine or GABA) by receptors on the The statistical difference in endosome diameter between FUSwt body wall muscles (Richmond and Jorgensen, 1999). The frequency and FUS501 was highly significant (P-value=0.00074), but only a of enPSCs was not different between the wild-type controls and tendency between wild-type and FUS501 (P-value=0.098) FUSwt, whereas FUS501 animals exhibited a >50% reduction in (Fig. 4B). There was no significant difference between wild-type enPSC frequency (Fig. 7A,B). There was no statistical difference in and FUSwt (P-value=0.40). The low number of endosomes the enPSC amplitude among the three genotypes (Fig. 7A,C). The analyzed for wild-type might explain the lack of statistical presence of a frequency defect in the absence of amplitude defect significance. Nevertheless, these data suggest that FUS501 suggests a reduction in presynaptic vesicle release from motor expression in the C. elegans nervous system results in the neurons onto muscle cells. This could be due to defective vesicle emergence of a population of large endosome-like structures at release at individual NMJs, a reduction in the number of NMJs, or a the NMJ, likely through enlargement of endosomes that contain combination of both. Because there was no change in enPSC electron-dense filaments. amplitude, vesicle loading, and postsynaptic reception are likely unaffected at this timepoint. The size of vesicles at the NMJ is modified by human FUS Given the presence of large endosomes in FUS501 terminals, we FUS501 proteins aggregate in the nuclei and cytoplasm used automated classification tools (Kaltdorf et al., 2017, 2018) to of motor neurons reconstruct the vesicle pools from tomograms of cholinergic NMJs. Given that FUS501 NMJs harbor a pool of large, filament-filled We found that clear synaptic vesicles (CCVs) had similar diameters endosome-like structures, we tested if the FUS501 protein was between the three genotypes, although the small differences were physically present at NMJs, thus could be directly and locally statistically significant (wild-type: 26.4±7 nm; FUSwt: 22.6± responsible for defects in neurotransmission. 8.8 nm; FUS501: 25.2±7 nm; P-values: FUSwt versus wild-type: Light microscopy studies have suggested that FUS501 forms 8.9e-06; FUS501 versus FUSwt: 0.0024; FUS501 versus wild-type: aggregates in motor neuron cell bodies and neurites that expand the Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 Fig. 2. FUS501 affects the synaptic ultrastructure of cholinergic motor neurons. Shown are virtual ∼1 nm slices from electron tomograms of cholinergic motor neurons, segmentations of these slices, and the 3D models of the whole tomograms. Segmented structures: plasma membrane (green), mitochondria (orange), dense projections (dark blue), microtubules (cyan), endoplasmic reticulum (lavender), dense core vesicles (yellow), clear core vesicles (white), docked clear core vesicles (red), and endosome-like structures (pink). Large, endosome-like structures appear in synapses affected by mutated FUS501, but not in FUSwt and wild-type controls. FUSwt controls show smaller structures that presumably represent normal endosomes. Scale bars: 200 nm. ventral nerve cords (Murakami et al., 2012, 2015; see also Figs. S2 DISCUSSION and S3). We used super-resolution array tomography (srAT; Markert FUS501-induced formation of large endosomal-like et al., 2017) to localize FUS proteins in their ultrastructural context. structures at NMJs FUS501 was detected as aggregated clusters in the nucleus and the FUS501 worms had a population of unusually large vesicles cytoplasm of motor neuron soma, whereas FUSwt signals were at NMJs. Their cellular nature remains to be better defined, diffuse and limited to the nucleus (data not shown). FUS501 was not but their morphology and location are consistent with in the nerve cords (Fig. 8). The lack of detection of FUS in motor endosomes described at C. elegans NMJs: They are roughly neuron processes may indicate its absence, alternatively, there was spherical and located within the synaptic vesicle pool (Watanabe insufficient preservation of small amounts of epitope. et al., 2013a), and they do not resemble other intracellular Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 Fig. 3. FUS501 affects the synaptic ultrastructure of GABAergic motor neurons. Shown are virtual ∼1 nm slices from electron tomograms of GABAergic motor neurons, segmentations of these slices, and the 3D models of the whole tomograms. Segmented structures: plasma membrane (green), mitochondria (orange), dense projections (dark blue), microtubules (cyan), endoplasmic reticulum (lavender), dense core vesicles (yellow), clear core vesicles (white), docked clear core vesicles (red), and endosome-like structures (pink). Large, endosome-like structures appear in synapses affected by mutated FUS501, but not in FUSwt and wild-type controls. Scale bars: 200 nm. structures, such as autophagosomes (Meléndez et al., 2003) and Putative endosomes in each genotype, FUS501, as well as FUSwt the endoplasmic reticulum (ER). Autophagosomes feature and wild-type control animals, were either ‘empty’ or contained double membranes, while our electron tomograms clearly show electron-dense filamentous inclusions. The larger population of that these large vesicles have single membranes (Fig. 4). ER in putative endosomes in FUS501 animals selectively contained our tomograms appeared as irregularly shaped tubes located at inclusions. Interestingly, in Alzheimer’s disease and Down the synaptic periphery, whereas these vesicles appeared syndrome, endosomes have also been reported to be enlarged due throughout synapses including regions close to the active zone. to acceleration of endocytosis (Cataldo et al., 2008; Colacurcio Thus, the population of large vesicles most likely represents et al., 2018). However, our data suggest a decrease in endocytosis, endosomes. which is discussed further in the next section. Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 some ultrastructural features of the nervous system, as well as in some aging phenotypes. But the degree of abnormality was substantially less that of FUS501, and was not consistent across all physiological phenotypes, for example, the frequency of endogenous postsynaptic currents. Such an effect, the mild phenotype upon overexpression of non-pathological form of a protein, has been observed for other neurodegenerative, ALS- included, associated genes (Cuvelier et al., 2018; Mirra et al., 2017; Mitchell et al., 2013; Moloney et al., 2018). Expression of human FUS, whether FUSwt or FUS501, resulted in changes to vesicle size and distribution at the NMJ. It also resulted in fewer docked vesicles. Although this was observed at both FUSwt and FUS501 terminals, reduced enPSCs were only observed in FUS501. If changes in vesicle docking contribute to defects in neurotransmission at FUS501 NMJs, additional factors must be at play. These may be within the synaptic terminal (e.g. changes in coupling of synaptic vesicles to voltage-gated calcium channels) (Chang and Martin, 2016), within the motor neuron (e.g. reduced excitability) (Guo et al., 2017; Liu et al., 2016; Naujock et al., 2016), or in upstream circuits that modulate endogenous motor neuron activity. Alternatively, a reduction of NMJ number might account for this finding. C. elegans possesses a FUS ortholog, FUST-1, with about 50% identity on the protein level. It was reported previously that deletions in fust-1 caused neuronal degradation and paralysis, whereas overexpression did not have any obvious effects (Therrien et al., 2016). It is unknown if mutations in FUST-1 can cause similar phenotypes as FUS501. If not, it would not be surprising for two Fig. 4. Larger endosomes that contain electron-dense filaments are reasons. First, FUS501 (and other FUS pathological mutations) causes caused by FUS501. Electron tomograms of hermaphrodite worms on day 3 a gain-of-function phenotype (Murakami et al., 2012). Second, there of adulthood (5 for wild-type, 17 for FUSwt, and 26 for FUS501) were used are an unusually large number of RRM family proteins in C. elegans to manually measure endosome diameters. For each endosome, the and many of them function redundantly (Thompson et al., 2019). average diameter calculated from the longest and shortest measured In ALS, excitotoxicity is a topic of concern with regards to neuron diameter was used for subsequent analysis. (A) Density plot of endosome diameters. FUS501 worms show populations of especially large endosomes degeneration (Fogarty, 2019). It is thought that neurons are driven to not present in controls (arrows). (B) Comparison of endosome diameters in decay by hyper-excitability for several forms of ALS (Fogarty, 2018). relation to the presence of electron-dense filaments. Statistical analysis via However, for FUS-mediated mouse model, motor neuron Mann–Whitney–Wilcoxon test. Data are depicted as violin plots. Median degeneration was reported to be preceded by hypo-excitability (closed circles) and mean (open circles) are given on each plot. For details (Martínez-Silva et al., 2018), echoing the high heterogeneity of see Materials and Methods and Table 1. ALS (Hardiman et al., 2017). Our results are more consistent with the hypo-excitability model and the conclusion from the mouse Electron-dense filamentous aggregates observed in these FUS models (Kong et al., 2009; Martínez-Silva et al., 2018; endosomes remain enigmatic. It is unknown whether these Ruiz et al., 2010). filaments are native structures, or if filaments in FUSwt or FUS501 contain FUS. It is also possible that such inclusions are a srAT reveals the ultrastructural context of FUS localization general aging phenotype, as they were also observed in our age Consistent with the implication of fluorescent microscopy studies matched wild-type controls. FUS501 aggregation may also promote (Murakami et al., 2012, 2015) we observed aggregated FUS in aggregation of other proteins. It is well established that the C. motor neuron nucleus and cytoplasm by srAT. But we did not detect elegans intracellular environment becomes more prone to FUS accumulations in axons, most likely due to the low abundance aggregation across ageing, and this is exacerbated by the of preserved epitopes, or insufficient sampling of our serial sections. overexpression of aggregation-prone proteins (David et al., 2010; A recent study in rodents reported FUS accumulation in synapses Huang et al., 2019; Walther et al., 2015). (Deshpande et al., 2019). Interestingly, in early development, FUS Thus, FUS501 appears to selectively increase the diameter of was predominantly found in postsynapses, but in mature neurons it endosome-like structures that contain inclusions. These endosome- was found in axon terminals. In C. elegans, FUS might only localize like structures may be involved in bulk endocytosis, consistent with to synapses in certain developmental stages or not at all. a possible defect in synaptic vesicle cycling in FUS501 nerve A previous study (Murakami et al., 2015) showed strong light- terminals. microscopy evidence that irreversible FUS hydrogels are associated with RNP granules, which have been described by electron FUS501 interferes with neurotransmission microscopy using samples preserved by a different protocol For both ultrastructural and functional analyses, we used two lines (Biggiogera et al., 1997; Jokhi et al., 2013; Souquere et al., as controls in all experiments, wild-type animals derived from 2009). We could not identify RNA granules in our EM sections; outcrossing the FUS501 strain with N2, and FUSwt. visualizing ultrastructural features of FUS aggregates or the RNP Overexpression of wild-type human FUS led to mild effects on granules they attach to may require different preservation protocols. Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 Fig. 5. Morphology of the large endosomes in FUS501. Examples of large endosomes (pink) in synapses of NMJs. Only endosomes in tomograms containing a dense projection are shown here. They appear large, with diameters of >80 nm, and often contain some electron-dense material (dark blue). (A) Typical example. The electron-dense content is often branched, as shown. (B) Electron-dense content appears partially in distinct dots. (C) Instance of a complete large endosome. (D) In one instance, large endosomes formed a group and ‘network’ as shown. Scale bars: (A, B, D): 100 nm, (C): 50 nm. For details see Materials and Methods and Table 1. A reduction of protein translation might account for perinatal death in mice (Hicks et al., 2000). Irreversible hydrogels endosome and vesicle docking defects formed by mutated FUS impair RNP granule function. This reduces FUS has many functions related to DNA and RNA processing and the rate of new protein synthesis (Murakami et al., 2015). This has a maintenance (Ratti and Buratti, 2016), and knock-out leads to systemic effect on neurons, however, the effect on axons and Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 Table 1. Numbers and sizes of endosomes Endosomes per µm Number of endosomes Mean endosome diameter [nm] nerve cord Number of tomograms Clear Filled Clear Filled Clear Filled Wild-type 5 7 5 53.0±5.2 74.5±17.0 5.0 7.0 FUSwt 17 25 14 52.0±4.4 59.3±14.1 7.4 4.1 FUS501 26 50 35 53.8±4.8 88.5±35.6 9.6 6.7 synapses is likely particularly detrimental, since such intracellular Swimming assay In swimming experiments, animals cultured one day after the L4 larval stage compartments have been shown to be heavily dependent on local were scored as day 1 adults. Ten animals at a time were transferred into a translation regulation by RNP granules in mouse (Akbalik and PDMS chamber on a glass slide containing 0.5 ml of M9 buffer. Swimming Schuman, 2014; Holt and Schuman, 2013; Jung et al., 2014) and in was recorded for 30 s using a Basler acA2500-60 μm camera mounted on a C. elegans (Yan et al., 2009). Furthermore, the replacement of dissection microscope. A single swim cycle was defined as a complete murine FUS with a mutated human form activates an integrated sinusoidal movement through the head and tail. stress response, and inhibits local intra-axonal protein synthesis in hippocampal neurons and sciatic nerves, resulting in synaptic High-pressure freezing dysfunction (López-Erauskin et al., 2018). A recent study found that The samples are subjected to >2100 bars of pressure and cooling rates −1 transcription of an acetylcholine receptor is compromised in FUS- of >20,000 K s . All samples used in this study were cryo-immobilized mediated ALS, thus supporting the idea that FUS affects using an EM HPM100 (Leica Microsystems) high-pressure freezing transcription and transcript processing (Picchiarelli et al., 2019). machine. The procedure was to use freezing platelets (Leica Microsystems) with 100 µm recesses. They were slightly overfilled with It is plausible that the ultrastructural defects that we found are OP50 paste (see below) and then worms were transferred into the platelet. A caused by reduced protein synthesis. Under physiological conditions, second platelet with a flat surface was placed on top as a lid. The samples vesicles are recycled via the ultrafast endocytosis pathway (Watanabe were processed and then stored in liquid nitrogen until freeze-substitution. et al., 2013a,b). After endocytosis, vesicles are regenerated in a clathrin-dependent manner (Watanabe et al., 2014). Shortage of E. coli OP50 bacteria paste clathrin or other components of this pathway could cause an A 100 ml volume of E. coli OP50 overnight culture was pelleted at 1500× g, accumulation of endocytosed membrane represented by large washed with 400 µl 20% bovine serum albumin (BSA) in M9 (Stiernagle, endosomes. Our observation that median distance of vesicles to the 2006) (3.0 g KH PO , 6.0 g Na HPO , 0.5 g NaCl, 1 ml 1 M MgSO ,H O 2 4 2 4 4 2 active zone is reduced in FUS worms and a reduction of vesicle pool to 1 l; sterilize by autoclaving), spun down again, and carefully size is consistent with this hypothesis. re-suspended in 20 µl 20% BSA in M9. In conclusion, we have shown synaptic architecture and functional changes in worms expressing FUS501. Our results implicate a direct Freeze-substitution and resin embedding for structural analyses or indirect role of human FUS in the organization of synaptic vesicles The protocol is based on (Weimer, 2006). A description of the individual steps of the freeze-substitution and resin embedding can be found in and synaptic transmission from motor neurons to muscles. These (Stigloher et al., 2011). We used Epon instead of Araldite in this study. phenotypic analyses of the C. elegans ALS model can aid the elucidation of cellular mechanisms that contribute to the ALS disease. Freeze-substitution and resin embedding for srAT analyses The protocol is also based on (Weimer, 2006). A detailed description of the MATERIALS AND METHODS individual steps of the freeze-substitution and resin embedding can be found Worm strains in (Markert et al., 2017). All C. elegans worms were maintained according to standard methods (Brenner, 1974). The transgenic animals used in this study were ZM9566 {hpIs239[Prgef-1::GFP::fus(del501)]; plasmid pJH2392}, which ectopically Ultramicrotomy for srAT and panneuronally expresses FUS501, and ZM5838 {hpIs223[Prgef-1::GFP:: A detailed description of the individual steps can be found in (Markert et al., fus(wt)]; plasmid pJH2382}, which ectopically and panneuronally expresses 2017). Briefly, 100 nm sections were produced using a special diamond wild-type FUS, both under the control of the Prgef-1 promoter. The wild-type knife with a boat large enough to accommodate glass slides (histo Jumbo control was ZM9569, derived by selecting wild-type siblings from the final diamond knife, DiATOME). Slides were submerged in the boat before outcross of the parent FUS501 strain with N2 Bristol. Since generating sectioning. Then the desired number of sections was cut without transgenic animals inevitably creates background mutations, and aging interruption. If necessary, a long ribbon was carefully divided into smaller might be sensitive to accumulative effect of silent mutations, ZM9569 ribbons using two mounted eyelashes. represents a more appropriate control to compare with the recovered ZM9566 FUS501 strain from the same set of crosses. ZM9566 and ZM5838 Ultramicrotomy for electron tomography both feature multiple copy insertions. ZM9566 and ZM5838 were both For imaging with a 200 kV TEM, we used sections up to 250 nm with good outcrossed six times to remove any potential background mutations results. Sections between 150 and 200 nm in thickness were favored. (Murakami et al., 2012). In addition, we included the N2 Bristol strain as an additional control in the lifespan, bagging, and swimming assays. Immunostaining Ultrathin sections of LR White-embedded tissue were immunostained for Lifespan srAT. Ultramicrotomy exposed epitopes at the section surface. Thus, C. elegans were synchronized at the L4 stage and manually picked to fresh sections could be stained, even though antibodies do generally not penetrate plates every 1–2 days to prevent contamination by progeny. Animals that the resin. were immotile and did not respond to gentle prodding with a platinum wire A detailed protocol can be found in (Markert et al., 2017). In brief, were scored as dead. Animals that bagged (died due to internal hatching of sections were placed in a humid chamber, and blocking buffer was applied progeny) or crawled up the walls of the plate and desiccated were censored to the sections prior to staining with primary and secondary antibodies. They from the lifespan analysis. were washed with buffer and then stained with a DNA stain, where Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 Fig. 6. Expression of human FUS modifies the size, position, and docking of vesicles at NMJs. Electron tomograms of hermaphrodite worms on day 3 of adulthood (wild-type, FUSwt, or FUS501) were used for automated and manual analysis. Only cholinergic synapses were included. Statistical analysis via Mann–Whitney–Wilcoxon test. Data are depicted as violin plots. Median (closed circles) and mean (open circles) are given on each plot. (A,B) Vesicle reconstruction and classification via the 3D ART VeSElecT and automated classification Fiji macros (Kaltdorf et al., 2017, 2018). In total, 2053 vesicles at cholinergic NMJs were reconstructed for wild-type, 164 for FUSwt, and 1030 for FUS501. (A) Linear distances of the center points of all vesicles to the center of the active zone (AZ) as given by the classification macro. (B) Vesicle radii as given by the classification macro. (C,D) Analysis of vesicles docked to the plasma membrane via manual analysis. (C) Numbers of docked vesicles per tomogram normalized to approximate volume of the dense projection in a given tomogram. (D) Linear distances of docked vesicles to the center of the AZ. For details see Materials and Methods and Table 2. Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 Table 2. Vesicle quantification Vesicle distance to Vesicle diameter [nm] active zone [nm] Docked vesicle distance CCV DCV CCV DCV Number of docked vesicles to active zone [nm] Wild-type 26.3±6.9 45.6±10.7 373±211 505±127 10.5±3.4 136±102 FUSwt 22.5±8.8 47.1±10.6 208±116 261±80.7 6.2±5.9 81.0±54.9 FUS501 25.2±7.0 35.5±5.6 249±141 339±85.3 2.7±2.4 73.5±37.8 applicable. Lastly, sections were mounted with Mowiol and stored at 4°C for Contrasting up to a week before fluorescence imaging. Contrasting was achieved by floating the grids sections down on drops of We used a polyclonal antibody against GFP (Abcam, catalog number: 2.5% uranyl acetate in ethanol for 15 min and 50% Reynolds’ lead citrate ab13970) at a dilution of 1:500 and a polyclonal antibody against FUS (Reynolds, 1963) in ddH O for 10 min. During incubation, samples were (Bethyl, A300-293A) at a dilution of 1:1000. covered to minimize evaporation. During incubation in lead citrate, sodium hydroxide pellets were placed around the samples to decrease local carbon dioxide concentration. Carbon dioxide forms precipitates with lead citrate. Imaging for srAT In between contrasting steps, the grids were washed first in ethanol, then in The workflow of srAT imaging has been published in detail (Markert et al., 50% ethanol in ddH O, and finally in ddH O. After contrasting, they were 2016, 2017). 2 2 thoroughly washed in ddH O and blotted dry with filter paper. Preparation of sections and imaging for electron tomography Carbon coating and placement of gold fiducials Imaging was performed with the SerialEM (Mastronarde, 2005) and IMOD Grids used for electron tomography were coated with a thin layer of carbon (Kremer et al., 1996) software packages. A 200 kV JEM-2100 (JEOL) to prevent charging during imaging at high tilt angles. The carbon layer had electron microscope equipped with a TemCam F416 4k×4k camera (Tietz an approximate thickness of 3 nm. Video and Imaging Processing Systems) was used for all TEM imaging and Gold fiducials were used to facilitate tomogram reconstruction. To electron tomography. achieve fiducial placement, a non-specific antibody conjugated with 10 nm gold particles was used. The antibody was diluted 1:10 with ddH O and 50 µl of this dilution were pipetted on a piece of clean parafilm. The carbon- coated grids were then floated on the drop for 10 min on each side, with a single wash in ddH O in between and at the end. A single wash meant that the grid was submerged in water for 1 second and then immediately dried with a filter paper. The gold fiducial placement was always performed right after carbon coating or at most a few hours later. For unknown reasons, longer delays caused very pronounced electron-dense precipitation on the sections, making them unsuitable for imaging in extreme cases. Acquisition of tilt series Tilt series for this thesis were acquired either from 60° to -60° or from 70° to -70°. Double tilts were performed where appropriate and possible, i.e., tilt series from a region of interest were acquired in two orthogonal tilt axes. This was achieved by manually rotating the grid by about 90° in the sample holder. Double tilts improved tomogram quality significantly. They were not performed when the tomogram of a single axis was sufficient to answer the specific questions. Tomogram reconstruction All tomograms were reconstructed with the eTomo software from the IMOD package (Kremer et al., 1996). Gold fiducials were always included to improve the alignment of the tilt series. For the step of tomogram positioning, the option ‘find boundary model automatically’ was used. Manual adjustments of the boundary model were almost never necessary. Tomograms were always created using the ‘Back Projection’ algorithm. Segmentation and 3D reconstruction Segmentation and 3D reconstruction were performed with the 3Dmod software from the IMOD package (Kremer et al., 1996). Investigators were blinded regarding the genotype. All structures except for the vesicles were segmented as closed objects using the ‘sculpt’ tool. Clear core and dense core vesicles were annotated as perfect spheres by creating a point in the center of a vesicle using the ‘normal’ drawing tool. This point was then Fig. 7. The frequency of endogenous postsynaptic currents is resized with the mouse wheel to match the outer dimensions of the given decreased in FUS501 transgenic animals. (A) Representative traces of vesicle. Global quality of points was set to 4 to achieve smooth spheres and endogenous postsynaptic currents in wild-type, FUSwt, and FUS501 the ‘drawing style’ of points was set to ‘fill’ to obtain closed surfaces. All animals. Muscles were held at -60 mV. (B) The frequency of endogenous other objects except the dense projections were meshed to obtain closed postsynaptic currents was significantly decreased in FUS501 compared to wild-type or FUSwt animals. (C) The amplitude of endogenous postsynaptic surfaces here as well. Dense projections were left with the default drawing currents showed no significant difference between strains. style ‘lines’. The ‘interpolator’ tool was used whenever appropriate. For Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 Fig. 8. FUS501 is present in nucleus and cytoplasm of motor neurons but not in the nerve cords. (A) Scanning electron micrograph of the ventral nerve cord (VNC) of a FUS501 worm. (B) Immunofluorescence staining against mutated FUS acquired by SIM of the same region as in A. (C) Overlay of A and B localizes FUS501 signals to the nucleus (Nu) and cytoplasm of a motor neuron but does not show any signals on the synapses of the VNC. (D) Scanning electron micrographs of the VNC region of an adult hermaphrodite worm expressing mutated FUS. Consecutive 200 nm sections are shown. (E) Aligned SIM channels shown for reference. (F) Overlay of D and E. Hoechst was used for correlation. FUS501 was stained via direct antibody (red) and via its GFP-tag (green). Both stainings overlap significantly (arrowheads). Scale bar: 1 µm. large structures like the plasma membranes, gaps of 20 virtual sections or Average endosome diameters were calculated from the longest and more were linearly interpolated. For mitochondria and microtubules shortest diameter of a given endosome measured manually on the virtual typically gaps of ten sections were interpolated. Larger spherical tomogram slice where the endosome appeared largest. structures like endosomes were interpolated with the ‘spherical’ option. Dense projections were not interpolated. Statistical analyses For vesicles and endosomes statistical analyses and their representations Quantitative analyses were performed with R (R Core Team, 2017). Kruskal–Wallis tests (one- Automatic vesicle reconstruction from electron tomograms was performed way ANOVA on ranks; normality of the data is not assumed) determined if via macros for the open source image processing software Fiji (Schindelin samples originated from the same distributions. The Mann–Whitney– et al., 2012) as described in (Kaltdorf et al., 2017). They were then Wilcoxon test was then used to determine statistical significance of automatically classified into clear core and dense core vesicles according to differences between pairs in the groups. For the survival assay, log-rank tests (Kaltdorf et al., 2018). Manual adjustments of the outcomes were not were used. For the swimming assay we performed unpaired two-tailed performed. However, if overall classification results for a given tomogram Student’s t-tests. The following significance levels were applied: *P<0.05, were not satisfactory, this tomogram was excluded from analysis. **P<0.01, ***P<0.001. The active zone was determined manually for the classification Variability of quantitative data samples was measured via median macro. A point on the plasma membrane that is closest to the center absolute deviation (MAD). The MAD is more robust against outliers and of gravity of the dense projection seen in a given tomogram was set as suitable for non-parametric data, i.e., data that does not show normal the center of the active zone. The center of gravity of the dense distribution (Pham-Gia and Hung, 2001). It is defined as the median of the projection was chosen by visual judgment of the user during the macro absolute deviations of the data’s median: workflow. Manual vesicle reconstruction from electron tomograms was performed MAD ¼ medianðjx  xjÞ: via 3Dmod from the software package IMOD (Kremer et al., 1996). The for a univariate dataset x , x , …, x where x is the median of the data: center points of vesicles were set by the user’s judgment and set as centers of 1 2 n x ¼ medianðxÞ. spheres with the approximate outer diameter of the vesicles. The dense projections were segmented manually, and their center of gravity was determined with the ‘imodinfo’ function of IMOD. The center Electrophysiology of the active zone was defined as the intersection of the inner plasma The dissection of the C. elegans was described previously (Richmond and membrane and an orthogonal line through the center of gravity of the dense Jorgensen, 1999). Briefly, hermaphrodites on day 3 of adulthood were glued projection. Linear distances of vesicles to the active zone were measured to a PDMS-coated cover glass covered with bath solution. The integrity of with the ‘measure’ tool in 3Dmod from the centers of the vesicles to the the ventral body muscle and the ventral nerve cord were visually examined center of the active zone. via DIC microscopy, and muscle cells were patched using fire-polished Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 accelerated endocytosis promotes late endocytic defects. Am. J. Pathol. 173, 4–6MΩ resistant borosilicate pipettes (World Precision Instruments, USA). 370-384. doi:10.2353/ajpath.2008.071053 Membrane currents were recorded in the whole-cell configuration by a Chang, Q. and Martin, L. J. (2016). Voltage-gated calcium channels are abnormal Digidata 1550B and a MultiClamp 700B amplifier, using the Clampex 10 in cultured spinal motoneurons in the G93A-SOD1 transgenic mouse model of software and processed with Clampfit 10 (Axon Instruments, Molecular ALS. Neurobiol. Dis. 93, 78-95. doi:10.1016/j.nbd.2016.04.009 Devices, USA). Data were digitized at 10–20 kHz and filtered at 2.6 kHz. Chou SM (1992) Pathology-light microscopy of amyotrophic lateral sclerosis. The recording solutions were as described in our previous studies (Gao and In Handbook of amyotrophic lateral sclerosis (ed Smith RA), pp. 133-181. Marcel Dekker, New York. Zhen, 2011). Specifically, the pipette solution contains (in mM): K-gluconate Colacurcio, D. J., Pensalfini, A., Jiang, Y. and Nixon, R. A. (2018). Dysfunction of 115; KCl 25; CaCl 0.1; MgCl 5; BAPTA 1; HEPES 10; Na ATP 5; 2 2 2 autophagy and endosomal-lysosomal pathways: roles in pathogenesis of down Na GTP 0.5;cAMP0.5;cGMP0.5,pH7.2 with KOH, ∼320 mOsm. The syndrome and Alzheimer’s disease. Free Radic. Biol. Med. 114, 40-51. doi:10. bath solution consists of (in mM): NaCl 150; KCl 5; CaCl 5; MgCl 1; 2 2 1016/j.freeradbiomed.2017.10.001 glucose 10; sucrose 5; HEPES 15, pH7.3 with NaOH, ∼330 mOsm. Leak Cuvelier, E., Méquinion, M., Leghay, C., Sibran, W., Stievenard, A., Sarchione, currents were not subtracted. All chemicals were from Sigma-Aldrich. A., Bonte, M.-A., Vanbesien-Mailliot, C., Viltart, O., Saitoski, K. et al. (2018). Overexpression of wild-type human alpha-synuclein causes metabolism Experiments were performed at room temperatures (20–22°C) abnormalities in Thy1-aSYN transgenic mice. Front. Mol. Neurosci. 11, 321. doi:10.3389/fnmol.2018.00321 Acknowledgements David, D. C., Ollikainen, N., Trinidad, J. C., Cary, M. P., Burlingame, A. L. and The authors cordially thank Veronika Perschin for help with data analysis, Philip Kenyon, C. (2010). Widespread protein aggregation as an inherent part of aging Kollmannsberger and the Center for Computational and Theoretical Biology (CCTB) in C. elegans. PLoS Biol. 8, e1000450. doi:10.1371/journal.pbio.1000450 Wü rzburg for fruitful discussions throughout the project and advice on statistical Deshpande, D., Higelin, J., Schoen, M., Vomhof, T., Boeckers, T. M., Demestre, analysis, Yi Li for help with the behavioral study, Daniela Bunsen, Glaudia Gehrig- M. and Michaelis, J. (2019). Synaptic FUS localization during motoneuron Hohn, and Brigitte Trost of the Imaging Core Facility of the Biocenter of the University development and its accumulation in human ALS synapses. Front. Cell. Neurosci. of Wü rzburg for technical support, and Georg Krohne for advice and fruitful 13, 256. doi:10.3389/fncel.2019.00256 discussions throughout the project. Fogarty, M. J. (2018). Driven to decay: excitability and synaptic abnormalities in amyotrophic lateral sclerosis. Brain Res. Bull. 140, 318-333. doi:10.1016/j. brainresbull.2018.05.023 Competing interests Fogarty, M. J. (2019). Amyotrophic lateral sclerosis as a synaptopathy. Neural The authors declare no competing or financial interests. Regen. Res. 14, 189-192. doi:10.4103/1673-5374.244782 Gal, J., Zhang, J., Kwinter, D. M., Zhai, J., Jia, H., Jia, J. and Zhu, H. (2011). Author contributions Nuclear localization sequence of FUS and induction of stress granules by ALS Conceptualization: S.M.M., M.P.S., M.Z., M.S., C.S.; Methodology: S.M.M., M.P.S.; mutants. Neurobiol. Aging 32, 2323.e27-2323.e40. doi:10.1016/j.neurobiolaging. Validation: S.M.M., M.P.S.; Formal analysis: S.M.M., M.P.S., B.M., S.G.; 2010.06.010 Investigation: S.M.M., M.P.S., B.Y., B.M.; Resources: M.Z., M.S., C.S.; Data Gao, S. and Zhen, M. (2011). Action potentials drive body wall muscle contractions curation: S.M.M.; Writing - original draft: S.M.M.; Writing - review & editing: S.M.M., in Caenorhabditis elegans. Proc. Natl Acad. Sci. USA 108, 2557-2562. doi:10. M.P.S., B.M., M.Z., S.G., M.S., C.S.; Visualization: S.M.M., M.P.S., B.M.; 1073/pnas.1012346108 Supervision: M.Z., S.G., M.S., C.S.; Project administration: C.S.; Funding Guo, W., Naujock, M., Fumagalli, L., Vandoorne, T., Baatsen, P., Boon, R., acquisition: S.M.M., M.Z., C.S. Ordovás, L., Patel, A., Welters, M., Vanwelden, T. et al. (2017). HDAC6 inhibition reverses axonal transport defects in motor neurons derived from FUS- ALS patients. Nat. Commun. 8, 861. doi:10.1038/s41467-017-00911-y Funding Hardiman, O., Al-Chalabi, A., Chio, A., Corr, E. M., Logroscino, G., Robberecht, This work was supported by the Deutsche Forschungsgemeinschaft (DFG) W., Shaw, P. J., Simmons, Z. and van den Berg, L. H. (2017). Amyotrophic [GRK2581-P06, STI700/1-1 to C.S.]; by The Brain Canada Foundation [to M.Z.]; and lateral sclerosis. Nat. Rev. Dis. Primers 3, 17071. doi:10.1038/nrdp.2017.71 by the Studienstiftung des Deutschen Volkes [to S.M.M.]. The publication was Hicks, G. G., Singh, N., Nashabi, A., Mai, S., Bozek, G., Klewes, L., Arapovic, D., supported by the Open Access Publication Fund of the University of Wü rzburg. White, E. K., Koury, M. J., Oltz, E. M. et al. (2000). Fus deficiency in mice results in defective B-lymphocyte development and activation, high levels of chromosomal Data availability instability and perinatal death. Nat. Genet. 24, 175. doi:10.1038/72842 The raw data of our quantifications as well as example tomograms including 3D Holt, C. E. and Schuman, E. M. (2013). The central dogma decentralized: new models are available via the dryad depository at https://doi.org/10.5061/dryad. perspectives on RNA function and local translation in Neurons. Neuron 80, 9ghx3ffg0. 648-657. doi:10.1016/j.neuron.2013.10.036 Huang, C., Wagner-Valladolid, S., Stephens, A. D., Jung, R., Poudel, C., Sinnige, T., Lechler, M. C., Schlö rit, N., Lu, M., Laine, R. F. et al. (2019). Supplementary information Intrinsically aggregation-prone proteins form amyloid-like aggregates and Supplementary information available online at contribute to tissue aging in Caenorhabditis elegans. eLife 8, e43059. doi:10. https://bio.biologists.org/lookup/doi/10.1242/bio.055129.supplemental 7554/eLife.43059 Jokhi, V., Ashley, J., Nunnari, J., Noma, A., Ito, N., Wakabayashi-Ito, N., Moore, M. J. and Budnik, V. (2013). Torsin mediates primary envelopment of large References ribonucleoprotein granules at the nuclear envelope. Cell Rep. 3, 988-995. doi:10. Akbalik, G. and Schuman, E. M. (2014). mRNA, live and unmasked. Science 343, 1016/j.celrep.2013.03.015 375-376. doi:10.1126/science.1249623 Jung, H., Gkogkas, C. G., Sonenberg, N. and Holt, C. E. (2014). Remote control of An, H., Skelt, L., Notaro, A., Highley, J. R., Fox, A. H., La Bella, V., Buchman, V. L. gene function by local translation. Cell 157, 26-40. doi:10.1016/j.cell.2014.03.005 and Shelkovnikova, T. A. (2019). ALS-linked FUS mutations confer loss and gain Kaltdorf, K. V., Schulze, K., Helmprobst, F., Kollmannsberger, P., Dandekar, T. of function in the nucleus by promoting excessive formation of dysfunctional and Stigloher, C. (2017). FIJI macro 3D ART VeSElecT: 3D automated paraspeckles. Acta Neuropathol. Commun. 7, 7. doi:10.1186/s40478-019-0658-x reconstruction tool for vesicle structures of electron tomograms. PLoS Comput. Andersson, M. K., Ståhlberg, A., Arvidsson, Y., Olofsson, A., Semb, H., Biol. 13, e1005317. doi:10.1371/journal.pcbi.1005317 Stenman, G., Nilsson, O. and Åman, P. (2008). The multifunctional FUS, EWS Kaltdorf, K. V., Theiss, M., Markert, S. M., Zhen, M., Dandekar, T., Stigloher, C. and TAF15 proto-oncoproteins show cell type-specific expression patterns and and Kollmannsberger, P. (2018). Automated classification of synaptic vesicles in involvement in cell spreading and stress response. BMC Cell Biol. 9, 37. doi:10. electron tomograms of C. elegans using machine learning. PLoS ONE 13, 1186/1471-2121-9-37 e0205348. doi:10.1371/journal.pone.0205348 Biggiogera, M., Bottone, M. G. and Pellicciari, C. (1997). Nuclear Kino, Y., Washizu, C., Aquilanti, E., Okuno, M., Kurosawa, M., Yamada, M., Doi, ribonucleoprotein-containing structures undergo severe rearrangement during H. and Nukina, N. (2011). Intracellular localization and splicing regulation of FUS/ spontaneous thymocyte apoptosis. A morphological study by electron TLS are variably affected by amyotrophic lateral sclerosis-linked mutations. microscopy. Histochem. Cell Biol. 107, 331-336. doi:10.1007/s004180050118 Nucleic Acids Res. 39, 2781-2798. doi:10.1093/nar/gkq1162 Brenner, S. (1974). The genetics of caenorhabditis elegans. Genetics 77, 71-94. Kong, L., Wang, X., Choe, D. W., Polley, M., Burnett, B. G., Bosch-Marcé, M., Burk, K. and Pasterkamp, R. J. (2019). Disrupted neuronal trafficking in Griffin, J. W., Rich, M. M. and Sumner, C. J. (2009). Impaired synaptic vesicle amyotrophic lateral sclerosis. Acta Neuropathol. 137, 859-877. doi:10.1007/ release and immaturity of neuromuscular junctions in spinal muscular atrophy s00401-019-01964-7 mice. J. Neurosci. 29, 842-851. doi:10.1523/JNEUROSCI.4434-08.2009 Cataldo, A. M., Mathews, P. M., Boiteau, A. B., Hassinger, L. C., Peterhoff, C. M., Kremer, J. R., Mastronarde, D. N. and McIntosh, J. R. (1996). Computer Jiang, Y., Mullaney, K., Neve, R. L., Gruenberg, J. and Nixon, R. A. (2008). visualization of three-dimensional image data using IMOD. J. Struct. Biol. 116, Down syndrome fibroblast model of Alzheimer-related endosome pathology: 71-76. doi:10.1006/jsbi.1996.0013 Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 Kwiatkowski, T. J., Bosco, D. A., LeClerc, A. L., Tamrazian, E., Vanderburg, regulation of acetylcholine receptor transcription at neuromuscular junctions is C. R., Russ, C., Davis, A., Gilchrist, J., Kasarskis, E. J., Munsat, T. et al. compromised in amyotrophic lateral sclerosis. Nat. Neurosci. 22, 1793-1805. (2009). Mutations in the FUS/TLS gene on chromosome 16 cause familial doi:10.1038/s41593-019-0498-9 amyotrophic lateral sclerosis. Science 323, 1205-1208. doi:10.1126/science. Ratti, A. and Buratti, E. (2016). Physiological functions and pathobiology of TDP-43 1166066 and FUS/TLS proteins. J. Neurochem. 138, 95-111. doi:10.1111/jnc.13625 Lagier-Tourenne, C., Polymenidou, M. and Cleveland, D. W. (2010). TDP-43 and Reynolds, E. S. (1963). The use of lead citrate at high pH as an electron-opaque FUS/TLS: emerging roles in RNA processing and neurodegeneration. Hum. Mol. stain in electron microscopy. J. Cell Biol. 17, 208-212. doi:10.1083/jcb.17.1.208 Genet. 19, R46-R64. doi:10.1093/hmg/ddq137 Richmond, J. E. and Jorgensen, E. M. (1999). One GABA and two acetylcholine Liu, M.-L., Zang, T. and Zhang, C.-L. (2016). Direct lineage reprogramming reveals receptors function at the C. elegans neuromuscular junction. Nat. Neurosci. 2, disease-specific phenotypes of motor neurons from human ALS patients. Cell 791-797. doi:10.1038/12160 Rep. 14, 115-128. doi:10.1016/j.celrep.2015.12.018 R Core Team (2017). R: A language and environment for statistical computing. Lopez-Erauskin, J., Tadokoro, T., Baughn, M. W., Myers, B., McAlonis-Downes, R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project. M., Chillon-Marinas, C., Asiaban, J. N., Artates, J., Bui, A. T., Vetto, A. P. et al. org/. (2018). ALS/FTD-linked mutation in FUS suppresses intra-axonal protein Ruiz, R., Casanas, J. J., Torres-Benito, L., Cano, R. and Tabares, L. (2010). synthesis and drives disease without nuclear loss-of-function of FUS. Neuron Altered intracellular Ca2 Homeostasis in nerve terminals of severe spinal 100, 816-830.e7. doi:10.1016/j.neuron.2018.09.044 muscular atrophy mice. J. Neurosci. 30, 849-857. doi:10.1523/JNEUROSCI. Lorenzo-Betancor, O., Ogaki, K., Soto-Ortolaza, A., Labbé, C., Vilarino-Gü ell, 4496-09.2010 C., Rajput, A., Rajput, A. H., Pastor, P., Ortega, S., Lorenzo, E. et al. (2014). Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Longair, M., Pietzsch, Analysis of nuclear export sequence regions of FUS-related RNA-binding proteins T., Preibisch, S., Rueden, C., Saalfeld, S., Schmid, B. et al. (2012). Fiji: an in essential tremor. PLoS ONE 9, e111989. doi:10.1371/journal.pone.0111989 open-source platform for biological-image analysis. Nat. Meth. 9, 676-682. doi:10. Markert, S. M., Britz, S., Proppert, S., Lang, M., Witvliet, D., Mulcahy, B., Sauer, 1038/nmeth.2019 M., Zhen, M., Bessereau, J.-L. and Stigloher, C. (2016). Filling the gap: adding Souquere, S., Mollet, S., Kress, M., Dautry, F., Pierron, G. and Weil, D. (2009). super-resolution to array tomography for correlated ultrastructural and molecular Unravelling the ultrastructure of stress granules and associated P-bodies in identification of electrical synapses at the C. elegans connectome. human cells. J. Cell Sci. 122, 3619-3626. doi:10.1242/jcs.054437 Neurophotonics 3, 041802. doi:10.1117/1.NPh.3.4.041802 Stiernagle, T. (2006). Maintenance of C. elegans (February 11, 2006), WormBook, Markert, S. M., Bauer, V., Muenz, T. S., Jones, N. G., Helmprobst, F., Britz, S., ed. The C. elegans Research Community, WormBook. doi:10.1895/wormbook.1. Sauer, M., Rö ssler, W., Engstler, M. and Stigloher, C. (2017). Chapter 2 - 3D 101.1 subcellular localization with superresolution array tomography on ultrathin Stigloher, C., Zhan, H., Zhen, M., Richmond, J. and Bessereau, J.-L. (2011). The sections of various species. In Methods in Cell Biology (T. Mü ller-Reichert and presynaptic dense projection of the Caenorhabiditis elegans cholinergic P. Verkade ed.), pp. 21-47. Academic Press. neuromuscular junction localizes synaptic vesicles at the active zone through Martınez-Silva ́ , M. D. L., Imhoff-Manuel, R. D., Sharma, A., Heckman, C. J., SYD-2/liprin and UNC-10/RIM-dependent interactions. J. Neurosci. 31, Shneider, N. A., Roselli, F., Zytnicki, D. and Manuel, M. (2018). Hypoexcitability 4388-4396. doi:10.1523/JNEUROSCI.6164-10.2011 precedes denervation in the large fast-contracting motor units in two unrelated Therrien, M., Rouleau, G. A., Dion, P. A. and Parker, J. A. (2016). FET proteins mouse models of ALS. eLife 7, e30955. doi:10.7554/eLife.30955 regulate lifespan and neuronal integrity. Sci. Rep. 6, 25159. doi:10.1038/ Mastronarde, D. N. (2005). Automated electron microscope tomography using srep25159 robust prediction of specimen movements. J. Struct. Biol. 152, 36-51. doi:10.1016/ Thompson, M., Bixby, R., Dalton, R., Vandenburg, A., Calarco, J. A. and Norris, j.jsb.2005.07.007 A. D. (2019). Splicing in a single neuron is coordinately controlled by RNA binding Melendez, A., Talloczy, Z., Seaman, M., Eskelinen, E.-L., Hall, D. H. and Levine, proteins and transcription factors. eLife 8, e46726. doi:10.7554/eLife.46726 ́ ́ B. (2003). Autophagy genes are essential for dauer development and life-span Vance, C., Rogelj, B., Hortobágyi, T., De Vos, K. J., Nishimura, A. L., extension in C. elegans. Science 301, 1387-1391. doi:10.1126/science.1087782 Sreedharan, J., Hu, X., Smith, B., Ruddy, D., Wright, P. et al. (2009). Mirra, A., Rossi, S., Scaricamazza, S., Di Salvio, M., Salvatori, I., Valle, C., Mutations in FUS, an RNA processing protein, cause familial amyotrophic lateral Rusmini, P., Poletti, A., Cestra, G., Carrì, M. T. et al. (2017). Functional sclerosis Type 6. Science 323, 1208-1211. doi:10.1126/science.1165942 interaction between FUS and SMN underlies SMA-like splicing changes in wild- Vance, C., Scotter, E. L., Nishimura, A. L., Troakes, C., Mitchell, J. C., Kathe, C., type hFUS mice. Sci. Rep. 7, 2033. doi:10.1038/s41598-017-02195-0 Urwin, H., Manser, C., Miller, C. C., Hortobagyi, T. et al. (2013). ALS mutant Mitchell, J. C., McGoldrick, P., Vance, C., Hortobagyi, T., Sreedharan, J., Rogelj, FUS disrupts nuclear localization and sequesters wild-type FUS within B., Tudor, E. L., Smith, B. N., Klasen, C., Miller, C. C. J. et al. (2013). cytoplasmic stress granules. Hum. Mol. Genet. 22, 2676-2688. doi:10.1093/ Overexpression of human wild-type FUS causes progressive motor neuron hmg/ddt117 degeneration in an age- and dose-dependent fashion. Acta Neuropathol. 125, Walther, D. M., Kasturi, P., Zheng, M., Pinkert, S., Vecchi, G., Ciryam, P., 273-288. doi:10.1007/s00401-012-1043-z Morimoto, R. I., Dobson, C. M., Vendruscolo, M., Mann, M. et al. (2015). Moloney, C., Rayaprolu, S., Howard, J., Fromholt, S., Brown, H., Collins, M., Widespread proteome remodeling and aggregation in aging C. elegans. Cell 161, Cabrera, M., Duffy, C., Siemienski, Z., Miller, D. et al. (2018). Analysis of spinal 919-932. doi:10.1016/j.cell.2015.03.032 and muscle pathology in transgenic mice overexpressing wild-type and ALS- Wang, H., Guo, W., Mitra, J., Hegde, P. M., Vandoorne, T., Eckelmann, B. J., linked mutant MATR3. Acta Neuropathol. Commun. 6, 137. doi:10.1186/s40478- Mitra, S., Tomkinson, A. E., Van Den Bosch, L. and Hegde, M. L. (2018). 018-0631-0 Mutant FUS causes DNA ligation defects to inhibit oxidative damage repair in Murakami, T., Yang, S.-P., Xie, L., Kawano, T., Fu, D., Mukai, A., Bohm, C., Chen, Amyotrophic Lateral Sclerosis. Nat. Commun. 9, 3683. doi:10.1038/s41467-018- F., Robertson, J., Suzuki, H. et al. (2012). ALS mutations in FUS cause neuronal 06111-6 dysfunction and death in Caenorhabditis elegans by a dominant gain-of-function Watanabe, S., Liu, Q., Davis, M. W., Hollopeter, G., Thomas, N., Jorgensen, mechanism. Hum. Mol. Genet. 21, 1-9. doi:10.1093/hmg/ddr417 N. B. and Jorgensen, E. M. (2013a). Ultrafast endocytosis at Caenorhabditis Murakami, T., Qamar, S., Lin, J. Q., Schierle, G. S. K., Rees, E., Miyashita, A., elegans neuromuscular junctions. eLife 2, e00723. doi:10.7554/eLife.00723 Costa, A. R., Dodd, R. B., Chan, F. T. S., Michel, C. H. et al. (2015). ALS/FTD Watanabe, S., Rost, B. R., Camacho-Perez, M., Davis, M. W., Sohl-Kielczynski, ́ ̈ mutation-induced phase transition of FUS liquid droplets and reversible hydrogels B., Rosenmund, C. and Jorgensen, E. M. (2013b). Ultrafast endocytosis at into irreversible hydrogels impairs RNP granule function. Neuron 88, 678-690. mouse hippocampal synapses. Nature 504, 242-247. doi:10.1038/nature12809 doi:10.1016/j.neuron.2015.10.030 Watanabe, S., Trimbuch, T., Camacho-Pérez, M., Rost, B. R., Brokowski, B., Naujock, M., Stanslowsky, N., Bufler, S., Naumann, M., Reinhardt, P., Sohl-Kielczynski, B., Felies, A., Davis, M. W., Rosenmund, C. and Jorgensen, Sterneckert, J., Kefalakes, E., Kassebaum, C., Bursch, F., Lojewski, X. E. M. (2014). Clathrin regenerates synaptic vesicles from endosomes. Nature 515, et al. (2016). 4-Aminopyridine induced activity rescues hypoexcitable motor 228-233. doi:10.1038/nature13846 neurons from amyotrophic lateral sclerosis patient-derived induced pluripotent Weimer, R. M. (2006). Preservation of C. elegans tissue via high-pressure freezing stem cells. Stem Cells 34, 1563-1575. doi:10.1002/stem.2354 and freeze-substitution for ultrastructural analysis and immunocytochemistry. In Pham-Gia, T. and Hung, T. L. (2001). The mean and median absolute deviations. C. elegans (ed. K. Strange), pp. 203-221. Humana Press. Math. Comput. Model. 34, 921-936. doi:10.1016/S0895-7177(01)00109-1 Yan, D., Wu, Z., Chisholm, A. D. and Jin, Y. (2009). The DLK-1 kinase promotes Picchiarelli, G., Demestre, M., Zuko, A., Been, M., Higelin, J., Dieterlé, S., Goy, mRNA stability and local translation in C. elegans synapses and axon M.-A., Mallik, M., Sellier, C., Scekic-Zahirovic, J. et al. (2019). FUS-mediated regeneration. Cell 138, 1005-1018. doi:10.1016/j.cell.2009.06.023 Biology Open http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Biology Open The Company of Biologists

Overexpression of an ALS-associated FUS mutation in C. elegans disrupts NMJ morphology and leads to defective neuromuscular transmission

Loading next page...
 
/lp/the-company-of-biologists/overexpression-of-an-als-associated-fus-mutation-in-c-elegans-disrupts-SZB2yyvMHR

References (122)

Publisher
The Company of Biologists
Copyright
© 2021 The Company of Biologists. All rights reserved.
eISSN
2046-6390
DOI
10.1242/bio.055129
Publisher site
See Article on Publisher Site

Abstract

© 2020. Published by The Company of Biologists Ltd | Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 RESEARCH ARTICLE Overexpression of an ALS-associated FUS mutation in C. elegans disrupts NMJ morphology and leads to defective neuromuscular transmission 1, 2 3 4 4,5 3 Sebastian M. Markert *, Michael Skoruppa , Bin Yu , Ben Mulcahy , Mei Zhen , Shangbang Gao , 2 1, Michael Sendtner and Christian Stigloher * ABSTRACT muscles atrophy (Hardiman et al., 2017). ALS is hypothesized to be a synaptopathy (Fogarty, 2019), because the disease usually starts with The amyotrophic lateral sclerosis (ALS) neurodegenerative disorder dysfunction and degeneration of synapses, before axons and dendrites has been associated with multiple genetic lesions, including become dystrophic and the neurons undergo cell death (Chou, 1992). mutations in the gene for fused in sarcoma (FUS), a nuclear- Most cases of ALS (∼90%) are considered ‘spontaneous’– due to a localized RNA/DNA-binding protein. Neuronal expression of the combination of de novo genetic and environmental factors. The pathological form of FUS proteins in Caenorhabditis elegans results remaining∼10% of ALS cases are hereditary and are caused by defined in mislocalization and aggregation of FUS in the cytoplasm, and leads mutations in different genes. A number of cellular defects have been to impairment of motility. However, the mechanisms by which the implicated in the pathophysiology of ALS, including defective mutant FUS disrupts neuronal health and function remain unclear. endosomal and receptor trafficking, as well as changes in autophagy Here we investigated the impact of ALS-associated FUS on motor and axonal transport (Burk and Pasterkamp, 2019). neuron health using correlative light and electron microscopy, Mutations in the gene for FUS (fused in sarcoma) make up electron tomography, and electrophysiology. We show that ectopic approximately 5% of hereditary ALS cases (Kwiatkowski et al., expression of wild-type or ALS-associated human FUS impairs 2009; López-Erauskin et al., 2018; Vance et al., 2009). The FUS synaptic vesicle docking at neuromuscular junctions. ALS-associated protein has a RNA/DNA recognition motif, a putative nuclear export FUS led to the emergence of a population of large, electron-dense, signal, several disorganized domains, and a nuclear localization and filament-filled endosomes. Electrophysiological recording signal at the C-terminus (Kino et al., 2011; Lorenzo-Betancor et al., revealed reduced transmission from motor neurons to muscles. 2014; Murakami et al., 2012). FUS has been associated with a Together, these results suggest a pathological effect of ALS-causing plethora of cellular functions, including translation, splicing, RNA FUS at synaptic structure and function organization. transport, and DNA damage response (Andersson et al., 2008; Lagier-Tourenne et al., 2010; Ratti and Buratti, 2016; Therrien et al., This article has an associated First Person interview with the first 2016; Wang et al., 2018). Under physiological conditions, FUS is author of the paper. predominantly located in the nucleus, but it is able to shuttle between KEY WORDS: C. elegans, Fused in sarcoma, Amyotrophic lateral nucleus and cytoplasm in response to different stimuli (Gal et al., sclerosis, Super-resolution array tomography, Electron tomography, 2011; Vance et al., 2013). Pathological mutations are suggested to Neuromuscular junction result in toxic gain of function, as ALS-associated mutant FUS is prone to cytoplasmic accumulation. The potential role of loss of INTRODUCTION function due to FUS sequestration in the cytoplasm is still being Amyotrophic lateral sclerosis (ALS) is a severe disease of the debated (An et al., 2019). Currently, mechanisms by which mutant locomotor system where motor neurons progressively degenerate and FUS proteins lead to motor neuron degeneration are still largely enigmatic. A prevailing hypothesis is that mutated FUS proteins form irreversible hydrogels that impair ribonucleoprotein (RNP) granule University of Wü rzburg, Biocenter, Imaging Core Facility, Am Hubland, Wü rzburg 2 function (Murakami et al., 2015). 97074, Germany. University Hospital Wü rzburg, Institute of Clinical Neurobiology, Versbacherstraße 5, 97080 Wü rzburg, Germany. Huazhong University of Science In a previous study, several variants of human ALS-associated and Technology, Key Laboratory of Molecular Biophysics of the Ministry of FUS were ectopically expressed in the Caenorhabditis elegans Education, College of Life Science and Technology, Wuhan 430074, China. (MAUPAS 1900) nervous system (Murakami et al., 2012). These Lunenfeld-Tanenbaum Research Institute, Mount Sinai Hospital, 600 University Avenue, Toronto, Ontario M5G 1X5, Canada. University of Toronto, Department of worms died prematurely, and had impaired motility, suggesting Molecular Genetics, Physiology and Institute of Medical Science, 1 King’s College degeneration of the locomotor system through a gain-of-function Circle, Toronto, Ontario M5S 1A8, Canada. mechanism. Strength of these phenotypes was positively correlated *Authors for correspondence: (smarker5@jhmi.edu; christian.stigloher@ with the severity of human ALS caused by each variant, supporting uni-wuerzburg.de) C. elegans as a promising experimental paradigm to interrogate S.M.M., 0000-0001-9069-156X; B.M., 0000-0002-3336-245X; M.Z., 0000-0003- mechanisms by which mutant FUS proteins lead to structural and 0086-9622; S.G., 0000-0001-5431-4628; M.S., 0000-0002-4737-2974; C.S., 0000- functional disruption of the nervous system. Here, we used 0001-6941-2669 behavioral, ultrastructural, and electrophysiological approaches to This is an Open Access article distributed under the terms of the Creative Commons Attribution investigate how ALS-associated FUS impacts the C. elegans License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, neuromuscular system. We show that the organization of the distribution and reproduction in any medium provided that the original work is properly attributed. neuromuscular junction (NMJ) is impacted by human FUS, and Received 16 July 2020; Accepted 27 October 2020 ALS-associated FUS severely impairs functional communication Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 between motor neurons and muscles. These disease mechanisms previous results (Murakami et al., 2012, 2015), we observed a may contribute insights to other forms of ALS, as well as other shortened lifespan of FUS501 animals compared to wild-type (N2) neurodegenerative diseases. animals, as well as a genetically wild-type sibling strain derived from outcrossing FUS501 animals with N2 (referred to as ‘wild- RESULTS type’; Fig. 1A). FUS501 animals were also more likely to die from Expression of ALS-associated FUS in the C. elegans nervous internal hatching of progeny (a phenotype referred to as bagging), system results in reduced lifespan and impaired motility likely due to impairment of the neural circuit that modulates To determine the time window for analysis of the C. elegans ALS- egg-laying (Fig. 1B). Consistent with motor dysfunction, FUS501 model, we examined the hermaphrodite worms that stably and pan- animals showed reduced motility in swimming assays (Fig. 1C). neuronally express a pathological form of FUS, FUS C-terminal Animals expressing wild-type human FUS (referred to as FUSwt deletion (referred as FUS501 henceforth) (Murakami et al., 2012, henceforth) displayed a milder reduction of lifespan, but did not 2015) for effects on lifespan and motor function. Consistent with display a motor defect at the ages tested (Fig. 1A–C). Fig. 1. C. elegans expressing pan-neuronal human FUS501 exhibit lifespan and motility defects. (A) Animals expressing human FUS501 have a shorter median and maximum lifespan compared to wild-type animals, as well as animals expressing wild-type human FUS (n≥102 deaths per genotype; log-rank test, P<0.0001). Animals expressing wild-type human FUS also displayed a mild lifespan defect (compared to N2, n≥160 deaths per genotype; log-rank test, P=0.0014). (B) Internal hatching of progeny occurred more frequently in FUS501 animals than in non-transgenic animals and animals expressing wild-type human FUS. (C) C. elegans expressing FUS501 have defective swimming compared to wild-type, whereas animals expressing wild-type human FUS are able to swim similarly to N2 controls (n≥14 per genotype; one-way ANOVA followed by Tukey’s multiple comparisons, *P=0.0112; ****P<0.0001). (D) The alleles of FUS used in this study. Wild-type human FUS contains regions enriched with certain amino acids (one-letter code given), an RNA recognition motif (RRM) and a nuclear targeting signal at the C-terminus (NLS). Mutations in this region are often linked to ALS. Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 Based on these results, we chose to proceed with comparative 0.0049; Fig. 6A; Table 2). For dense core vesicles (DCVs), the ultrastructural and functional analyses for FUS501 and wild-type diameter was similar between wild-type and FUSwt (wild-type: animals on day 3 of adulthood. We reasoned that a pathological 45.6±10.6 nm; FUSwt: 47±10.6 nm; Table 2), but was significantly synaptic phenotype should already be pronounced prior to smaller in FUS501 (35.6±5.6; P-values: FUS501 versus FUSwt: widespread degeneration, before causing secondary phenotypes 0.0074; FUS501 versus wild-type: 4.0e-08; Table 2). DCVs made that could make analysis more difficult. At this age, pathological up 4.5% of all vesicles in wild-type, 4.9% in FUSwt, and 11.4% in features, including the FUS protein localization, aggregation, and FUS501 (P-values of two-sample t-tests: FUSwt versus wild-type: motility defects were prominent, while animal survival was still 0.83; FUS501 versus FUSwt: 0.012; FUS501 versus wild-type: high (Fig. 1A,C). <0.00001). Thus, in FUS501 worms, vesicle pools in NMJs contain a higher proportion of DCVs that are smaller in size (Fig. 6A). Large endosomes with electron-dense inclusions are enriched at the neuromuscular junctions (NMJs) of FUS501 FUS disrupts the organization and docking of vesicles adults at the NMJ In order to assess whether FUS501 motility defects reflect synaptic We further compared the distance between vesicles and the center of structural defects, we used electron tomography to map the the presynaptic dense projection. In animals expressing FUSwt and organization of excitatory and inhibitory NMJs in animals FUS501, both clear- and dense-core vesicles were closer to the expressing wild-type human FUS (FUSwt) and FUS501, as well presynaptic dense projection than wild-type controls (Fig. 6B; as the non-transgenic, wild-type animals that share otherwise the Table 2). For both classes of vesicles, the effect of FUSwt was more same genetic background. pronounced than FUS501 (Fig. 6B; Table 2). To assess how altered Wild-type and FUSwt NMJs were composed of a presynaptic distribution of vesicles may affect their release properties, we dense projection and associated pool of clear and dense core quantified the number of docked vesicles at each synapse. Both vesicles, and rarely contained large endosomes. However, large FUSwt and FUS501 had significantly fewer docked vesicles than endosome-like organelles were enriched at the NMJs of animals wild-type controls (P-values: FUSwt versus wild-type: 0.025; expressing FUS501, and a population of large endosomes FUS501 versus FUSwt: 0.44; FUS501 versus wild-type: 0.011; (>100 nm) was exclusively present in FUS501 worms (Figs 2–4). Fig. 6C; Table 2). Docked vesicles in FUSwt and FUS501 were This was accompanied by a reduction in the smaller (∼50 nm) closer to the center of the presynaptic dense projection than wild- diameter endosomes in FUS501 (Fig. 4A). These endosome-like type controls, similar to the global vesicle distribution described organelles were mostly electron-light, and present in both earlier (Fig. 6D; Table 2). Thus, in FUSwt and FUS501, fewer cholinergic and GABAergic NMJs (Figs 2 and 3). Most vesicles are docked, but those that are docked located closer to the endosomes were spherical, although more complex networks also presynaptic dense projection. occurred (see Fig. 5; Fig. S1). In FUS501 mutants, approximately 41% of endosome-like FUS501 animals exhibit reduced endogenous postsynaptic structures had filamentous, electron-dense inclusions (35/85, from currents 26 tomograms). Wild-type and FUSwt animals also had a similar Given the defects in the structural organization of NMJ in animals percentage of filled endosomes (36% and 41% respectively; 5/12 expressing human FUS, we recorded the functional output of the from 5 tomograms, and 14/39 from 17 tomograms). While ‘empty’ NMJ by patch clamping muscle cells in animals on day 3 of endosomes in each genotype were of a similar diameter (P-values: adulthood. The endogenous postsynaptic currents (enPSCs) reflect FUS501 versus FUSwt: 0.089; FUS501 versus wild-type: 0.24; synaptic vesicle fusion events that are due to either spontaneous FUSwt versus wild-type: 1.0), endosomes in FUS501 worms release or release evoked by endogenous activity in the motor containing electron-dense inclusions were larger than similar filled neurons or upstream circuits, and subsequent detection of endosomes in wild-type and FUSwt animals (Fig. 4B, Table 1). neurotransmitter (acetylcholine or GABA) by receptors on the The statistical difference in endosome diameter between FUSwt body wall muscles (Richmond and Jorgensen, 1999). The frequency and FUS501 was highly significant (P-value=0.00074), but only a of enPSCs was not different between the wild-type controls and tendency between wild-type and FUS501 (P-value=0.098) FUSwt, whereas FUS501 animals exhibited a >50% reduction in (Fig. 4B). There was no significant difference between wild-type enPSC frequency (Fig. 7A,B). There was no statistical difference in and FUSwt (P-value=0.40). The low number of endosomes the enPSC amplitude among the three genotypes (Fig. 7A,C). The analyzed for wild-type might explain the lack of statistical presence of a frequency defect in the absence of amplitude defect significance. Nevertheless, these data suggest that FUS501 suggests a reduction in presynaptic vesicle release from motor expression in the C. elegans nervous system results in the neurons onto muscle cells. This could be due to defective vesicle emergence of a population of large endosome-like structures at release at individual NMJs, a reduction in the number of NMJs, or a the NMJ, likely through enlargement of endosomes that contain combination of both. Because there was no change in enPSC electron-dense filaments. amplitude, vesicle loading, and postsynaptic reception are likely unaffected at this timepoint. The size of vesicles at the NMJ is modified by human FUS Given the presence of large endosomes in FUS501 terminals, we FUS501 proteins aggregate in the nuclei and cytoplasm used automated classification tools (Kaltdorf et al., 2017, 2018) to of motor neurons reconstruct the vesicle pools from tomograms of cholinergic NMJs. Given that FUS501 NMJs harbor a pool of large, filament-filled We found that clear synaptic vesicles (CCVs) had similar diameters endosome-like structures, we tested if the FUS501 protein was between the three genotypes, although the small differences were physically present at NMJs, thus could be directly and locally statistically significant (wild-type: 26.4±7 nm; FUSwt: 22.6± responsible for defects in neurotransmission. 8.8 nm; FUS501: 25.2±7 nm; P-values: FUSwt versus wild-type: Light microscopy studies have suggested that FUS501 forms 8.9e-06; FUS501 versus FUSwt: 0.0024; FUS501 versus wild-type: aggregates in motor neuron cell bodies and neurites that expand the Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 Fig. 2. FUS501 affects the synaptic ultrastructure of cholinergic motor neurons. Shown are virtual ∼1 nm slices from electron tomograms of cholinergic motor neurons, segmentations of these slices, and the 3D models of the whole tomograms. Segmented structures: plasma membrane (green), mitochondria (orange), dense projections (dark blue), microtubules (cyan), endoplasmic reticulum (lavender), dense core vesicles (yellow), clear core vesicles (white), docked clear core vesicles (red), and endosome-like structures (pink). Large, endosome-like structures appear in synapses affected by mutated FUS501, but not in FUSwt and wild-type controls. FUSwt controls show smaller structures that presumably represent normal endosomes. Scale bars: 200 nm. ventral nerve cords (Murakami et al., 2012, 2015; see also Figs. S2 DISCUSSION and S3). We used super-resolution array tomography (srAT; Markert FUS501-induced formation of large endosomal-like et al., 2017) to localize FUS proteins in their ultrastructural context. structures at NMJs FUS501 was detected as aggregated clusters in the nucleus and the FUS501 worms had a population of unusually large vesicles cytoplasm of motor neuron soma, whereas FUSwt signals were at NMJs. Their cellular nature remains to be better defined, diffuse and limited to the nucleus (data not shown). FUS501 was not but their morphology and location are consistent with in the nerve cords (Fig. 8). The lack of detection of FUS in motor endosomes described at C. elegans NMJs: They are roughly neuron processes may indicate its absence, alternatively, there was spherical and located within the synaptic vesicle pool (Watanabe insufficient preservation of small amounts of epitope. et al., 2013a), and they do not resemble other intracellular Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 Fig. 3. FUS501 affects the synaptic ultrastructure of GABAergic motor neurons. Shown are virtual ∼1 nm slices from electron tomograms of GABAergic motor neurons, segmentations of these slices, and the 3D models of the whole tomograms. Segmented structures: plasma membrane (green), mitochondria (orange), dense projections (dark blue), microtubules (cyan), endoplasmic reticulum (lavender), dense core vesicles (yellow), clear core vesicles (white), docked clear core vesicles (red), and endosome-like structures (pink). Large, endosome-like structures appear in synapses affected by mutated FUS501, but not in FUSwt and wild-type controls. Scale bars: 200 nm. structures, such as autophagosomes (Meléndez et al., 2003) and Putative endosomes in each genotype, FUS501, as well as FUSwt the endoplasmic reticulum (ER). Autophagosomes feature and wild-type control animals, were either ‘empty’ or contained double membranes, while our electron tomograms clearly show electron-dense filamentous inclusions. The larger population of that these large vesicles have single membranes (Fig. 4). ER in putative endosomes in FUS501 animals selectively contained our tomograms appeared as irregularly shaped tubes located at inclusions. Interestingly, in Alzheimer’s disease and Down the synaptic periphery, whereas these vesicles appeared syndrome, endosomes have also been reported to be enlarged due throughout synapses including regions close to the active zone. to acceleration of endocytosis (Cataldo et al., 2008; Colacurcio Thus, the population of large vesicles most likely represents et al., 2018). However, our data suggest a decrease in endocytosis, endosomes. which is discussed further in the next section. Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 some ultrastructural features of the nervous system, as well as in some aging phenotypes. But the degree of abnormality was substantially less that of FUS501, and was not consistent across all physiological phenotypes, for example, the frequency of endogenous postsynaptic currents. Such an effect, the mild phenotype upon overexpression of non-pathological form of a protein, has been observed for other neurodegenerative, ALS- included, associated genes (Cuvelier et al., 2018; Mirra et al., 2017; Mitchell et al., 2013; Moloney et al., 2018). Expression of human FUS, whether FUSwt or FUS501, resulted in changes to vesicle size and distribution at the NMJ. It also resulted in fewer docked vesicles. Although this was observed at both FUSwt and FUS501 terminals, reduced enPSCs were only observed in FUS501. If changes in vesicle docking contribute to defects in neurotransmission at FUS501 NMJs, additional factors must be at play. These may be within the synaptic terminal (e.g. changes in coupling of synaptic vesicles to voltage-gated calcium channels) (Chang and Martin, 2016), within the motor neuron (e.g. reduced excitability) (Guo et al., 2017; Liu et al., 2016; Naujock et al., 2016), or in upstream circuits that modulate endogenous motor neuron activity. Alternatively, a reduction of NMJ number might account for this finding. C. elegans possesses a FUS ortholog, FUST-1, with about 50% identity on the protein level. It was reported previously that deletions in fust-1 caused neuronal degradation and paralysis, whereas overexpression did not have any obvious effects (Therrien et al., 2016). It is unknown if mutations in FUST-1 can cause similar phenotypes as FUS501. If not, it would not be surprising for two Fig. 4. Larger endosomes that contain electron-dense filaments are reasons. First, FUS501 (and other FUS pathological mutations) causes caused by FUS501. Electron tomograms of hermaphrodite worms on day 3 a gain-of-function phenotype (Murakami et al., 2012). Second, there of adulthood (5 for wild-type, 17 for FUSwt, and 26 for FUS501) were used are an unusually large number of RRM family proteins in C. elegans to manually measure endosome diameters. For each endosome, the and many of them function redundantly (Thompson et al., 2019). average diameter calculated from the longest and shortest measured In ALS, excitotoxicity is a topic of concern with regards to neuron diameter was used for subsequent analysis. (A) Density plot of endosome diameters. FUS501 worms show populations of especially large endosomes degeneration (Fogarty, 2019). It is thought that neurons are driven to not present in controls (arrows). (B) Comparison of endosome diameters in decay by hyper-excitability for several forms of ALS (Fogarty, 2018). relation to the presence of electron-dense filaments. Statistical analysis via However, for FUS-mediated mouse model, motor neuron Mann–Whitney–Wilcoxon test. Data are depicted as violin plots. Median degeneration was reported to be preceded by hypo-excitability (closed circles) and mean (open circles) are given on each plot. For details (Martínez-Silva et al., 2018), echoing the high heterogeneity of see Materials and Methods and Table 1. ALS (Hardiman et al., 2017). Our results are more consistent with the hypo-excitability model and the conclusion from the mouse Electron-dense filamentous aggregates observed in these FUS models (Kong et al., 2009; Martínez-Silva et al., 2018; endosomes remain enigmatic. It is unknown whether these Ruiz et al., 2010). filaments are native structures, or if filaments in FUSwt or FUS501 contain FUS. It is also possible that such inclusions are a srAT reveals the ultrastructural context of FUS localization general aging phenotype, as they were also observed in our age Consistent with the implication of fluorescent microscopy studies matched wild-type controls. FUS501 aggregation may also promote (Murakami et al., 2012, 2015) we observed aggregated FUS in aggregation of other proteins. It is well established that the C. motor neuron nucleus and cytoplasm by srAT. But we did not detect elegans intracellular environment becomes more prone to FUS accumulations in axons, most likely due to the low abundance aggregation across ageing, and this is exacerbated by the of preserved epitopes, or insufficient sampling of our serial sections. overexpression of aggregation-prone proteins (David et al., 2010; A recent study in rodents reported FUS accumulation in synapses Huang et al., 2019; Walther et al., 2015). (Deshpande et al., 2019). Interestingly, in early development, FUS Thus, FUS501 appears to selectively increase the diameter of was predominantly found in postsynapses, but in mature neurons it endosome-like structures that contain inclusions. These endosome- was found in axon terminals. In C. elegans, FUS might only localize like structures may be involved in bulk endocytosis, consistent with to synapses in certain developmental stages or not at all. a possible defect in synaptic vesicle cycling in FUS501 nerve A previous study (Murakami et al., 2015) showed strong light- terminals. microscopy evidence that irreversible FUS hydrogels are associated with RNP granules, which have been described by electron FUS501 interferes with neurotransmission microscopy using samples preserved by a different protocol For both ultrastructural and functional analyses, we used two lines (Biggiogera et al., 1997; Jokhi et al., 2013; Souquere et al., as controls in all experiments, wild-type animals derived from 2009). We could not identify RNA granules in our EM sections; outcrossing the FUS501 strain with N2, and FUSwt. visualizing ultrastructural features of FUS aggregates or the RNP Overexpression of wild-type human FUS led to mild effects on granules they attach to may require different preservation protocols. Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 Fig. 5. Morphology of the large endosomes in FUS501. Examples of large endosomes (pink) in synapses of NMJs. Only endosomes in tomograms containing a dense projection are shown here. They appear large, with diameters of >80 nm, and often contain some electron-dense material (dark blue). (A) Typical example. The electron-dense content is often branched, as shown. (B) Electron-dense content appears partially in distinct dots. (C) Instance of a complete large endosome. (D) In one instance, large endosomes formed a group and ‘network’ as shown. Scale bars: (A, B, D): 100 nm, (C): 50 nm. For details see Materials and Methods and Table 1. A reduction of protein translation might account for perinatal death in mice (Hicks et al., 2000). Irreversible hydrogels endosome and vesicle docking defects formed by mutated FUS impair RNP granule function. This reduces FUS has many functions related to DNA and RNA processing and the rate of new protein synthesis (Murakami et al., 2015). This has a maintenance (Ratti and Buratti, 2016), and knock-out leads to systemic effect on neurons, however, the effect on axons and Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 Table 1. Numbers and sizes of endosomes Endosomes per µm Number of endosomes Mean endosome diameter [nm] nerve cord Number of tomograms Clear Filled Clear Filled Clear Filled Wild-type 5 7 5 53.0±5.2 74.5±17.0 5.0 7.0 FUSwt 17 25 14 52.0±4.4 59.3±14.1 7.4 4.1 FUS501 26 50 35 53.8±4.8 88.5±35.6 9.6 6.7 synapses is likely particularly detrimental, since such intracellular Swimming assay In swimming experiments, animals cultured one day after the L4 larval stage compartments have been shown to be heavily dependent on local were scored as day 1 adults. Ten animals at a time were transferred into a translation regulation by RNP granules in mouse (Akbalik and PDMS chamber on a glass slide containing 0.5 ml of M9 buffer. Swimming Schuman, 2014; Holt and Schuman, 2013; Jung et al., 2014) and in was recorded for 30 s using a Basler acA2500-60 μm camera mounted on a C. elegans (Yan et al., 2009). Furthermore, the replacement of dissection microscope. A single swim cycle was defined as a complete murine FUS with a mutated human form activates an integrated sinusoidal movement through the head and tail. stress response, and inhibits local intra-axonal protein synthesis in hippocampal neurons and sciatic nerves, resulting in synaptic High-pressure freezing dysfunction (López-Erauskin et al., 2018). A recent study found that The samples are subjected to >2100 bars of pressure and cooling rates −1 transcription of an acetylcholine receptor is compromised in FUS- of >20,000 K s . All samples used in this study were cryo-immobilized mediated ALS, thus supporting the idea that FUS affects using an EM HPM100 (Leica Microsystems) high-pressure freezing transcription and transcript processing (Picchiarelli et al., 2019). machine. The procedure was to use freezing platelets (Leica Microsystems) with 100 µm recesses. They were slightly overfilled with It is plausible that the ultrastructural defects that we found are OP50 paste (see below) and then worms were transferred into the platelet. A caused by reduced protein synthesis. Under physiological conditions, second platelet with a flat surface was placed on top as a lid. The samples vesicles are recycled via the ultrafast endocytosis pathway (Watanabe were processed and then stored in liquid nitrogen until freeze-substitution. et al., 2013a,b). After endocytosis, vesicles are regenerated in a clathrin-dependent manner (Watanabe et al., 2014). Shortage of E. coli OP50 bacteria paste clathrin or other components of this pathway could cause an A 100 ml volume of E. coli OP50 overnight culture was pelleted at 1500× g, accumulation of endocytosed membrane represented by large washed with 400 µl 20% bovine serum albumin (BSA) in M9 (Stiernagle, endosomes. Our observation that median distance of vesicles to the 2006) (3.0 g KH PO , 6.0 g Na HPO , 0.5 g NaCl, 1 ml 1 M MgSO ,H O 2 4 2 4 4 2 active zone is reduced in FUS worms and a reduction of vesicle pool to 1 l; sterilize by autoclaving), spun down again, and carefully size is consistent with this hypothesis. re-suspended in 20 µl 20% BSA in M9. In conclusion, we have shown synaptic architecture and functional changes in worms expressing FUS501. Our results implicate a direct Freeze-substitution and resin embedding for structural analyses or indirect role of human FUS in the organization of synaptic vesicles The protocol is based on (Weimer, 2006). A description of the individual steps of the freeze-substitution and resin embedding can be found in and synaptic transmission from motor neurons to muscles. These (Stigloher et al., 2011). We used Epon instead of Araldite in this study. phenotypic analyses of the C. elegans ALS model can aid the elucidation of cellular mechanisms that contribute to the ALS disease. Freeze-substitution and resin embedding for srAT analyses The protocol is also based on (Weimer, 2006). A detailed description of the MATERIALS AND METHODS individual steps of the freeze-substitution and resin embedding can be found Worm strains in (Markert et al., 2017). All C. elegans worms were maintained according to standard methods (Brenner, 1974). The transgenic animals used in this study were ZM9566 {hpIs239[Prgef-1::GFP::fus(del501)]; plasmid pJH2392}, which ectopically Ultramicrotomy for srAT and panneuronally expresses FUS501, and ZM5838 {hpIs223[Prgef-1::GFP:: A detailed description of the individual steps can be found in (Markert et al., fus(wt)]; plasmid pJH2382}, which ectopically and panneuronally expresses 2017). Briefly, 100 nm sections were produced using a special diamond wild-type FUS, both under the control of the Prgef-1 promoter. The wild-type knife with a boat large enough to accommodate glass slides (histo Jumbo control was ZM9569, derived by selecting wild-type siblings from the final diamond knife, DiATOME). Slides were submerged in the boat before outcross of the parent FUS501 strain with N2 Bristol. Since generating sectioning. Then the desired number of sections was cut without transgenic animals inevitably creates background mutations, and aging interruption. If necessary, a long ribbon was carefully divided into smaller might be sensitive to accumulative effect of silent mutations, ZM9569 ribbons using two mounted eyelashes. represents a more appropriate control to compare with the recovered ZM9566 FUS501 strain from the same set of crosses. ZM9566 and ZM5838 Ultramicrotomy for electron tomography both feature multiple copy insertions. ZM9566 and ZM5838 were both For imaging with a 200 kV TEM, we used sections up to 250 nm with good outcrossed six times to remove any potential background mutations results. Sections between 150 and 200 nm in thickness were favored. (Murakami et al., 2012). In addition, we included the N2 Bristol strain as an additional control in the lifespan, bagging, and swimming assays. Immunostaining Ultrathin sections of LR White-embedded tissue were immunostained for Lifespan srAT. Ultramicrotomy exposed epitopes at the section surface. Thus, C. elegans were synchronized at the L4 stage and manually picked to fresh sections could be stained, even though antibodies do generally not penetrate plates every 1–2 days to prevent contamination by progeny. Animals that the resin. were immotile and did not respond to gentle prodding with a platinum wire A detailed protocol can be found in (Markert et al., 2017). In brief, were scored as dead. Animals that bagged (died due to internal hatching of sections were placed in a humid chamber, and blocking buffer was applied progeny) or crawled up the walls of the plate and desiccated were censored to the sections prior to staining with primary and secondary antibodies. They from the lifespan analysis. were washed with buffer and then stained with a DNA stain, where Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 Fig. 6. Expression of human FUS modifies the size, position, and docking of vesicles at NMJs. Electron tomograms of hermaphrodite worms on day 3 of adulthood (wild-type, FUSwt, or FUS501) were used for automated and manual analysis. Only cholinergic synapses were included. Statistical analysis via Mann–Whitney–Wilcoxon test. Data are depicted as violin plots. Median (closed circles) and mean (open circles) are given on each plot. (A,B) Vesicle reconstruction and classification via the 3D ART VeSElecT and automated classification Fiji macros (Kaltdorf et al., 2017, 2018). In total, 2053 vesicles at cholinergic NMJs were reconstructed for wild-type, 164 for FUSwt, and 1030 for FUS501. (A) Linear distances of the center points of all vesicles to the center of the active zone (AZ) as given by the classification macro. (B) Vesicle radii as given by the classification macro. (C,D) Analysis of vesicles docked to the plasma membrane via manual analysis. (C) Numbers of docked vesicles per tomogram normalized to approximate volume of the dense projection in a given tomogram. (D) Linear distances of docked vesicles to the center of the AZ. For details see Materials and Methods and Table 2. Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 Table 2. Vesicle quantification Vesicle distance to Vesicle diameter [nm] active zone [nm] Docked vesicle distance CCV DCV CCV DCV Number of docked vesicles to active zone [nm] Wild-type 26.3±6.9 45.6±10.7 373±211 505±127 10.5±3.4 136±102 FUSwt 22.5±8.8 47.1±10.6 208±116 261±80.7 6.2±5.9 81.0±54.9 FUS501 25.2±7.0 35.5±5.6 249±141 339±85.3 2.7±2.4 73.5±37.8 applicable. Lastly, sections were mounted with Mowiol and stored at 4°C for Contrasting up to a week before fluorescence imaging. Contrasting was achieved by floating the grids sections down on drops of We used a polyclonal antibody against GFP (Abcam, catalog number: 2.5% uranyl acetate in ethanol for 15 min and 50% Reynolds’ lead citrate ab13970) at a dilution of 1:500 and a polyclonal antibody against FUS (Reynolds, 1963) in ddH O for 10 min. During incubation, samples were (Bethyl, A300-293A) at a dilution of 1:1000. covered to minimize evaporation. During incubation in lead citrate, sodium hydroxide pellets were placed around the samples to decrease local carbon dioxide concentration. Carbon dioxide forms precipitates with lead citrate. Imaging for srAT In between contrasting steps, the grids were washed first in ethanol, then in The workflow of srAT imaging has been published in detail (Markert et al., 50% ethanol in ddH O, and finally in ddH O. After contrasting, they were 2016, 2017). 2 2 thoroughly washed in ddH O and blotted dry with filter paper. Preparation of sections and imaging for electron tomography Carbon coating and placement of gold fiducials Imaging was performed with the SerialEM (Mastronarde, 2005) and IMOD Grids used for electron tomography were coated with a thin layer of carbon (Kremer et al., 1996) software packages. A 200 kV JEM-2100 (JEOL) to prevent charging during imaging at high tilt angles. The carbon layer had electron microscope equipped with a TemCam F416 4k×4k camera (Tietz an approximate thickness of 3 nm. Video and Imaging Processing Systems) was used for all TEM imaging and Gold fiducials were used to facilitate tomogram reconstruction. To electron tomography. achieve fiducial placement, a non-specific antibody conjugated with 10 nm gold particles was used. The antibody was diluted 1:10 with ddH O and 50 µl of this dilution were pipetted on a piece of clean parafilm. The carbon- coated grids were then floated on the drop for 10 min on each side, with a single wash in ddH O in between and at the end. A single wash meant that the grid was submerged in water for 1 second and then immediately dried with a filter paper. The gold fiducial placement was always performed right after carbon coating or at most a few hours later. For unknown reasons, longer delays caused very pronounced electron-dense precipitation on the sections, making them unsuitable for imaging in extreme cases. Acquisition of tilt series Tilt series for this thesis were acquired either from 60° to -60° or from 70° to -70°. Double tilts were performed where appropriate and possible, i.e., tilt series from a region of interest were acquired in two orthogonal tilt axes. This was achieved by manually rotating the grid by about 90° in the sample holder. Double tilts improved tomogram quality significantly. They were not performed when the tomogram of a single axis was sufficient to answer the specific questions. Tomogram reconstruction All tomograms were reconstructed with the eTomo software from the IMOD package (Kremer et al., 1996). Gold fiducials were always included to improve the alignment of the tilt series. For the step of tomogram positioning, the option ‘find boundary model automatically’ was used. Manual adjustments of the boundary model were almost never necessary. Tomograms were always created using the ‘Back Projection’ algorithm. Segmentation and 3D reconstruction Segmentation and 3D reconstruction were performed with the 3Dmod software from the IMOD package (Kremer et al., 1996). Investigators were blinded regarding the genotype. All structures except for the vesicles were segmented as closed objects using the ‘sculpt’ tool. Clear core and dense core vesicles were annotated as perfect spheres by creating a point in the center of a vesicle using the ‘normal’ drawing tool. This point was then Fig. 7. The frequency of endogenous postsynaptic currents is resized with the mouse wheel to match the outer dimensions of the given decreased in FUS501 transgenic animals. (A) Representative traces of vesicle. Global quality of points was set to 4 to achieve smooth spheres and endogenous postsynaptic currents in wild-type, FUSwt, and FUS501 the ‘drawing style’ of points was set to ‘fill’ to obtain closed surfaces. All animals. Muscles were held at -60 mV. (B) The frequency of endogenous other objects except the dense projections were meshed to obtain closed postsynaptic currents was significantly decreased in FUS501 compared to wild-type or FUSwt animals. (C) The amplitude of endogenous postsynaptic surfaces here as well. Dense projections were left with the default drawing currents showed no significant difference between strains. style ‘lines’. The ‘interpolator’ tool was used whenever appropriate. For Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 Fig. 8. FUS501 is present in nucleus and cytoplasm of motor neurons but not in the nerve cords. (A) Scanning electron micrograph of the ventral nerve cord (VNC) of a FUS501 worm. (B) Immunofluorescence staining against mutated FUS acquired by SIM of the same region as in A. (C) Overlay of A and B localizes FUS501 signals to the nucleus (Nu) and cytoplasm of a motor neuron but does not show any signals on the synapses of the VNC. (D) Scanning electron micrographs of the VNC region of an adult hermaphrodite worm expressing mutated FUS. Consecutive 200 nm sections are shown. (E) Aligned SIM channels shown for reference. (F) Overlay of D and E. Hoechst was used for correlation. FUS501 was stained via direct antibody (red) and via its GFP-tag (green). Both stainings overlap significantly (arrowheads). Scale bar: 1 µm. large structures like the plasma membranes, gaps of 20 virtual sections or Average endosome diameters were calculated from the longest and more were linearly interpolated. For mitochondria and microtubules shortest diameter of a given endosome measured manually on the virtual typically gaps of ten sections were interpolated. Larger spherical tomogram slice where the endosome appeared largest. structures like endosomes were interpolated with the ‘spherical’ option. Dense projections were not interpolated. Statistical analyses For vesicles and endosomes statistical analyses and their representations Quantitative analyses were performed with R (R Core Team, 2017). Kruskal–Wallis tests (one- Automatic vesicle reconstruction from electron tomograms was performed way ANOVA on ranks; normality of the data is not assumed) determined if via macros for the open source image processing software Fiji (Schindelin samples originated from the same distributions. The Mann–Whitney– et al., 2012) as described in (Kaltdorf et al., 2017). They were then Wilcoxon test was then used to determine statistical significance of automatically classified into clear core and dense core vesicles according to differences between pairs in the groups. For the survival assay, log-rank tests (Kaltdorf et al., 2018). Manual adjustments of the outcomes were not were used. For the swimming assay we performed unpaired two-tailed performed. However, if overall classification results for a given tomogram Student’s t-tests. The following significance levels were applied: *P<0.05, were not satisfactory, this tomogram was excluded from analysis. **P<0.01, ***P<0.001. The active zone was determined manually for the classification Variability of quantitative data samples was measured via median macro. A point on the plasma membrane that is closest to the center absolute deviation (MAD). The MAD is more robust against outliers and of gravity of the dense projection seen in a given tomogram was set as suitable for non-parametric data, i.e., data that does not show normal the center of the active zone. The center of gravity of the dense distribution (Pham-Gia and Hung, 2001). It is defined as the median of the projection was chosen by visual judgment of the user during the macro absolute deviations of the data’s median: workflow. Manual vesicle reconstruction from electron tomograms was performed MAD ¼ medianðjx  xjÞ: via 3Dmod from the software package IMOD (Kremer et al., 1996). The for a univariate dataset x , x , …, x where x is the median of the data: center points of vesicles were set by the user’s judgment and set as centers of 1 2 n x ¼ medianðxÞ. spheres with the approximate outer diameter of the vesicles. The dense projections were segmented manually, and their center of gravity was determined with the ‘imodinfo’ function of IMOD. The center Electrophysiology of the active zone was defined as the intersection of the inner plasma The dissection of the C. elegans was described previously (Richmond and membrane and an orthogonal line through the center of gravity of the dense Jorgensen, 1999). Briefly, hermaphrodites on day 3 of adulthood were glued projection. Linear distances of vesicles to the active zone were measured to a PDMS-coated cover glass covered with bath solution. The integrity of with the ‘measure’ tool in 3Dmod from the centers of the vesicles to the the ventral body muscle and the ventral nerve cord were visually examined center of the active zone. via DIC microscopy, and muscle cells were patched using fire-polished Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 accelerated endocytosis promotes late endocytic defects. Am. J. Pathol. 173, 4–6MΩ resistant borosilicate pipettes (World Precision Instruments, USA). 370-384. doi:10.2353/ajpath.2008.071053 Membrane currents were recorded in the whole-cell configuration by a Chang, Q. and Martin, L. J. (2016). Voltage-gated calcium channels are abnormal Digidata 1550B and a MultiClamp 700B amplifier, using the Clampex 10 in cultured spinal motoneurons in the G93A-SOD1 transgenic mouse model of software and processed with Clampfit 10 (Axon Instruments, Molecular ALS. Neurobiol. Dis. 93, 78-95. doi:10.1016/j.nbd.2016.04.009 Devices, USA). Data were digitized at 10–20 kHz and filtered at 2.6 kHz. Chou SM (1992) Pathology-light microscopy of amyotrophic lateral sclerosis. The recording solutions were as described in our previous studies (Gao and In Handbook of amyotrophic lateral sclerosis (ed Smith RA), pp. 133-181. Marcel Dekker, New York. Zhen, 2011). Specifically, the pipette solution contains (in mM): K-gluconate Colacurcio, D. J., Pensalfini, A., Jiang, Y. and Nixon, R. A. (2018). Dysfunction of 115; KCl 25; CaCl 0.1; MgCl 5; BAPTA 1; HEPES 10; Na ATP 5; 2 2 2 autophagy and endosomal-lysosomal pathways: roles in pathogenesis of down Na GTP 0.5;cAMP0.5;cGMP0.5,pH7.2 with KOH, ∼320 mOsm. The syndrome and Alzheimer’s disease. Free Radic. Biol. Med. 114, 40-51. doi:10. bath solution consists of (in mM): NaCl 150; KCl 5; CaCl 5; MgCl 1; 2 2 1016/j.freeradbiomed.2017.10.001 glucose 10; sucrose 5; HEPES 15, pH7.3 with NaOH, ∼330 mOsm. Leak Cuvelier, E., Méquinion, M., Leghay, C., Sibran, W., Stievenard, A., Sarchione, currents were not subtracted. All chemicals were from Sigma-Aldrich. A., Bonte, M.-A., Vanbesien-Mailliot, C., Viltart, O., Saitoski, K. et al. (2018). Overexpression of wild-type human alpha-synuclein causes metabolism Experiments were performed at room temperatures (20–22°C) abnormalities in Thy1-aSYN transgenic mice. Front. Mol. Neurosci. 11, 321. doi:10.3389/fnmol.2018.00321 Acknowledgements David, D. C., Ollikainen, N., Trinidad, J. C., Cary, M. P., Burlingame, A. L. and The authors cordially thank Veronika Perschin for help with data analysis, Philip Kenyon, C. (2010). Widespread protein aggregation as an inherent part of aging Kollmannsberger and the Center for Computational and Theoretical Biology (CCTB) in C. elegans. PLoS Biol. 8, e1000450. doi:10.1371/journal.pbio.1000450 Wü rzburg for fruitful discussions throughout the project and advice on statistical Deshpande, D., Higelin, J., Schoen, M., Vomhof, T., Boeckers, T. M., Demestre, analysis, Yi Li for help with the behavioral study, Daniela Bunsen, Glaudia Gehrig- M. and Michaelis, J. (2019). Synaptic FUS localization during motoneuron Hohn, and Brigitte Trost of the Imaging Core Facility of the Biocenter of the University development and its accumulation in human ALS synapses. Front. Cell. Neurosci. of Wü rzburg for technical support, and Georg Krohne for advice and fruitful 13, 256. doi:10.3389/fncel.2019.00256 discussions throughout the project. Fogarty, M. J. (2018). Driven to decay: excitability and synaptic abnormalities in amyotrophic lateral sclerosis. Brain Res. Bull. 140, 318-333. doi:10.1016/j. brainresbull.2018.05.023 Competing interests Fogarty, M. J. (2019). Amyotrophic lateral sclerosis as a synaptopathy. Neural The authors declare no competing or financial interests. Regen. Res. 14, 189-192. doi:10.4103/1673-5374.244782 Gal, J., Zhang, J., Kwinter, D. M., Zhai, J., Jia, H., Jia, J. and Zhu, H. (2011). Author contributions Nuclear localization sequence of FUS and induction of stress granules by ALS Conceptualization: S.M.M., M.P.S., M.Z., M.S., C.S.; Methodology: S.M.M., M.P.S.; mutants. Neurobiol. Aging 32, 2323.e27-2323.e40. doi:10.1016/j.neurobiolaging. Validation: S.M.M., M.P.S.; Formal analysis: S.M.M., M.P.S., B.M., S.G.; 2010.06.010 Investigation: S.M.M., M.P.S., B.Y., B.M.; Resources: M.Z., M.S., C.S.; Data Gao, S. and Zhen, M. (2011). Action potentials drive body wall muscle contractions curation: S.M.M.; Writing - original draft: S.M.M.; Writing - review & editing: S.M.M., in Caenorhabditis elegans. Proc. Natl Acad. Sci. USA 108, 2557-2562. doi:10. M.P.S., B.M., M.Z., S.G., M.S., C.S.; Visualization: S.M.M., M.P.S., B.M.; 1073/pnas.1012346108 Supervision: M.Z., S.G., M.S., C.S.; Project administration: C.S.; Funding Guo, W., Naujock, M., Fumagalli, L., Vandoorne, T., Baatsen, P., Boon, R., acquisition: S.M.M., M.Z., C.S. Ordovás, L., Patel, A., Welters, M., Vanwelden, T. et al. (2017). HDAC6 inhibition reverses axonal transport defects in motor neurons derived from FUS- ALS patients. Nat. Commun. 8, 861. doi:10.1038/s41467-017-00911-y Funding Hardiman, O., Al-Chalabi, A., Chio, A., Corr, E. M., Logroscino, G., Robberecht, This work was supported by the Deutsche Forschungsgemeinschaft (DFG) W., Shaw, P. J., Simmons, Z. and van den Berg, L. H. (2017). Amyotrophic [GRK2581-P06, STI700/1-1 to C.S.]; by The Brain Canada Foundation [to M.Z.]; and lateral sclerosis. Nat. Rev. Dis. Primers 3, 17071. doi:10.1038/nrdp.2017.71 by the Studienstiftung des Deutschen Volkes [to S.M.M.]. The publication was Hicks, G. G., Singh, N., Nashabi, A., Mai, S., Bozek, G., Klewes, L., Arapovic, D., supported by the Open Access Publication Fund of the University of Wü rzburg. White, E. K., Koury, M. J., Oltz, E. M. et al. (2000). Fus deficiency in mice results in defective B-lymphocyte development and activation, high levels of chromosomal Data availability instability and perinatal death. Nat. Genet. 24, 175. doi:10.1038/72842 The raw data of our quantifications as well as example tomograms including 3D Holt, C. E. and Schuman, E. M. (2013). The central dogma decentralized: new models are available via the dryad depository at https://doi.org/10.5061/dryad. perspectives on RNA function and local translation in Neurons. Neuron 80, 9ghx3ffg0. 648-657. doi:10.1016/j.neuron.2013.10.036 Huang, C., Wagner-Valladolid, S., Stephens, A. D., Jung, R., Poudel, C., Sinnige, T., Lechler, M. C., Schlö rit, N., Lu, M., Laine, R. F. et al. (2019). Supplementary information Intrinsically aggregation-prone proteins form amyloid-like aggregates and Supplementary information available online at contribute to tissue aging in Caenorhabditis elegans. eLife 8, e43059. doi:10. https://bio.biologists.org/lookup/doi/10.1242/bio.055129.supplemental 7554/eLife.43059 Jokhi, V., Ashley, J., Nunnari, J., Noma, A., Ito, N., Wakabayashi-Ito, N., Moore, M. J. and Budnik, V. (2013). Torsin mediates primary envelopment of large References ribonucleoprotein granules at the nuclear envelope. Cell Rep. 3, 988-995. doi:10. Akbalik, G. and Schuman, E. M. (2014). mRNA, live and unmasked. Science 343, 1016/j.celrep.2013.03.015 375-376. doi:10.1126/science.1249623 Jung, H., Gkogkas, C. G., Sonenberg, N. and Holt, C. E. (2014). Remote control of An, H., Skelt, L., Notaro, A., Highley, J. R., Fox, A. H., La Bella, V., Buchman, V. L. gene function by local translation. Cell 157, 26-40. doi:10.1016/j.cell.2014.03.005 and Shelkovnikova, T. A. (2019). ALS-linked FUS mutations confer loss and gain Kaltdorf, K. V., Schulze, K., Helmprobst, F., Kollmannsberger, P., Dandekar, T. of function in the nucleus by promoting excessive formation of dysfunctional and Stigloher, C. (2017). FIJI macro 3D ART VeSElecT: 3D automated paraspeckles. Acta Neuropathol. Commun. 7, 7. doi:10.1186/s40478-019-0658-x reconstruction tool for vesicle structures of electron tomograms. PLoS Comput. Andersson, M. K., Ståhlberg, A., Arvidsson, Y., Olofsson, A., Semb, H., Biol. 13, e1005317. doi:10.1371/journal.pcbi.1005317 Stenman, G., Nilsson, O. and Åman, P. (2008). The multifunctional FUS, EWS Kaltdorf, K. V., Theiss, M., Markert, S. M., Zhen, M., Dandekar, T., Stigloher, C. and TAF15 proto-oncoproteins show cell type-specific expression patterns and and Kollmannsberger, P. (2018). Automated classification of synaptic vesicles in involvement in cell spreading and stress response. BMC Cell Biol. 9, 37. doi:10. electron tomograms of C. elegans using machine learning. PLoS ONE 13, 1186/1471-2121-9-37 e0205348. doi:10.1371/journal.pone.0205348 Biggiogera, M., Bottone, M. G. and Pellicciari, C. (1997). Nuclear Kino, Y., Washizu, C., Aquilanti, E., Okuno, M., Kurosawa, M., Yamada, M., Doi, ribonucleoprotein-containing structures undergo severe rearrangement during H. and Nukina, N. (2011). Intracellular localization and splicing regulation of FUS/ spontaneous thymocyte apoptosis. A morphological study by electron TLS are variably affected by amyotrophic lateral sclerosis-linked mutations. microscopy. Histochem. Cell Biol. 107, 331-336. doi:10.1007/s004180050118 Nucleic Acids Res. 39, 2781-2798. doi:10.1093/nar/gkq1162 Brenner, S. (1974). The genetics of caenorhabditis elegans. Genetics 77, 71-94. Kong, L., Wang, X., Choe, D. W., Polley, M., Burnett, B. G., Bosch-Marcé, M., Burk, K. and Pasterkamp, R. J. (2019). Disrupted neuronal trafficking in Griffin, J. W., Rich, M. M. and Sumner, C. J. (2009). Impaired synaptic vesicle amyotrophic lateral sclerosis. Acta Neuropathol. 137, 859-877. doi:10.1007/ release and immaturity of neuromuscular junctions in spinal muscular atrophy s00401-019-01964-7 mice. J. Neurosci. 29, 842-851. doi:10.1523/JNEUROSCI.4434-08.2009 Cataldo, A. M., Mathews, P. M., Boiteau, A. B., Hassinger, L. C., Peterhoff, C. M., Kremer, J. R., Mastronarde, D. N. and McIntosh, J. R. (1996). Computer Jiang, Y., Mullaney, K., Neve, R. L., Gruenberg, J. and Nixon, R. A. (2008). visualization of three-dimensional image data using IMOD. J. Struct. Biol. 116, Down syndrome fibroblast model of Alzheimer-related endosome pathology: 71-76. doi:10.1006/jsbi.1996.0013 Biology Open RESEARCH ARTICLE Biology Open (2020) 9, bio055129. doi:10.1242/bio.055129 Kwiatkowski, T. J., Bosco, D. A., LeClerc, A. L., Tamrazian, E., Vanderburg, regulation of acetylcholine receptor transcription at neuromuscular junctions is C. R., Russ, C., Davis, A., Gilchrist, J., Kasarskis, E. J., Munsat, T. et al. compromised in amyotrophic lateral sclerosis. Nat. Neurosci. 22, 1793-1805. (2009). Mutations in the FUS/TLS gene on chromosome 16 cause familial doi:10.1038/s41593-019-0498-9 amyotrophic lateral sclerosis. Science 323, 1205-1208. doi:10.1126/science. Ratti, A. and Buratti, E. (2016). Physiological functions and pathobiology of TDP-43 1166066 and FUS/TLS proteins. J. Neurochem. 138, 95-111. doi:10.1111/jnc.13625 Lagier-Tourenne, C., Polymenidou, M. and Cleveland, D. W. (2010). TDP-43 and Reynolds, E. S. (1963). The use of lead citrate at high pH as an electron-opaque FUS/TLS: emerging roles in RNA processing and neurodegeneration. Hum. Mol. stain in electron microscopy. J. Cell Biol. 17, 208-212. doi:10.1083/jcb.17.1.208 Genet. 19, R46-R64. doi:10.1093/hmg/ddq137 Richmond, J. E. and Jorgensen, E. M. (1999). One GABA and two acetylcholine Liu, M.-L., Zang, T. and Zhang, C.-L. (2016). Direct lineage reprogramming reveals receptors function at the C. elegans neuromuscular junction. Nat. Neurosci. 2, disease-specific phenotypes of motor neurons from human ALS patients. Cell 791-797. doi:10.1038/12160 Rep. 14, 115-128. doi:10.1016/j.celrep.2015.12.018 R Core Team (2017). R: A language and environment for statistical computing. Lopez-Erauskin, J., Tadokoro, T., Baughn, M. W., Myers, B., McAlonis-Downes, R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project. M., Chillon-Marinas, C., Asiaban, J. N., Artates, J., Bui, A. T., Vetto, A. P. et al. org/. (2018). ALS/FTD-linked mutation in FUS suppresses intra-axonal protein Ruiz, R., Casanas, J. J., Torres-Benito, L., Cano, R. and Tabares, L. (2010). synthesis and drives disease without nuclear loss-of-function of FUS. Neuron Altered intracellular Ca2 Homeostasis in nerve terminals of severe spinal 100, 816-830.e7. doi:10.1016/j.neuron.2018.09.044 muscular atrophy mice. J. Neurosci. 30, 849-857. doi:10.1523/JNEUROSCI. Lorenzo-Betancor, O., Ogaki, K., Soto-Ortolaza, A., Labbé, C., Vilarino-Gü ell, 4496-09.2010 C., Rajput, A., Rajput, A. H., Pastor, P., Ortega, S., Lorenzo, E. et al. (2014). Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Longair, M., Pietzsch, Analysis of nuclear export sequence regions of FUS-related RNA-binding proteins T., Preibisch, S., Rueden, C., Saalfeld, S., Schmid, B. et al. (2012). Fiji: an in essential tremor. PLoS ONE 9, e111989. doi:10.1371/journal.pone.0111989 open-source platform for biological-image analysis. Nat. Meth. 9, 676-682. doi:10. Markert, S. M., Britz, S., Proppert, S., Lang, M., Witvliet, D., Mulcahy, B., Sauer, 1038/nmeth.2019 M., Zhen, M., Bessereau, J.-L. and Stigloher, C. (2016). Filling the gap: adding Souquere, S., Mollet, S., Kress, M., Dautry, F., Pierron, G. and Weil, D. (2009). super-resolution to array tomography for correlated ultrastructural and molecular Unravelling the ultrastructure of stress granules and associated P-bodies in identification of electrical synapses at the C. elegans connectome. human cells. J. Cell Sci. 122, 3619-3626. doi:10.1242/jcs.054437 Neurophotonics 3, 041802. doi:10.1117/1.NPh.3.4.041802 Stiernagle, T. (2006). Maintenance of C. elegans (February 11, 2006), WormBook, Markert, S. M., Bauer, V., Muenz, T. S., Jones, N. G., Helmprobst, F., Britz, S., ed. The C. elegans Research Community, WormBook. doi:10.1895/wormbook.1. Sauer, M., Rö ssler, W., Engstler, M. and Stigloher, C. (2017). Chapter 2 - 3D 101.1 subcellular localization with superresolution array tomography on ultrathin Stigloher, C., Zhan, H., Zhen, M., Richmond, J. and Bessereau, J.-L. (2011). The sections of various species. In Methods in Cell Biology (T. Mü ller-Reichert and presynaptic dense projection of the Caenorhabiditis elegans cholinergic P. Verkade ed.), pp. 21-47. Academic Press. neuromuscular junction localizes synaptic vesicles at the active zone through Martınez-Silva ́ , M. D. L., Imhoff-Manuel, R. D., Sharma, A., Heckman, C. J., SYD-2/liprin and UNC-10/RIM-dependent interactions. J. Neurosci. 31, Shneider, N. A., Roselli, F., Zytnicki, D. and Manuel, M. (2018). Hypoexcitability 4388-4396. doi:10.1523/JNEUROSCI.6164-10.2011 precedes denervation in the large fast-contracting motor units in two unrelated Therrien, M., Rouleau, G. A., Dion, P. A. and Parker, J. A. (2016). FET proteins mouse models of ALS. eLife 7, e30955. doi:10.7554/eLife.30955 regulate lifespan and neuronal integrity. Sci. Rep. 6, 25159. doi:10.1038/ Mastronarde, D. N. (2005). Automated electron microscope tomography using srep25159 robust prediction of specimen movements. J. Struct. Biol. 152, 36-51. doi:10.1016/ Thompson, M., Bixby, R., Dalton, R., Vandenburg, A., Calarco, J. A. and Norris, j.jsb.2005.07.007 A. D. (2019). Splicing in a single neuron is coordinately controlled by RNA binding Melendez, A., Talloczy, Z., Seaman, M., Eskelinen, E.-L., Hall, D. H. and Levine, proteins and transcription factors. eLife 8, e46726. doi:10.7554/eLife.46726 ́ ́ B. (2003). Autophagy genes are essential for dauer development and life-span Vance, C., Rogelj, B., Hortobágyi, T., De Vos, K. J., Nishimura, A. L., extension in C. elegans. Science 301, 1387-1391. doi:10.1126/science.1087782 Sreedharan, J., Hu, X., Smith, B., Ruddy, D., Wright, P. et al. (2009). Mirra, A., Rossi, S., Scaricamazza, S., Di Salvio, M., Salvatori, I., Valle, C., Mutations in FUS, an RNA processing protein, cause familial amyotrophic lateral Rusmini, P., Poletti, A., Cestra, G., Carrì, M. T. et al. (2017). Functional sclerosis Type 6. Science 323, 1208-1211. doi:10.1126/science.1165942 interaction between FUS and SMN underlies SMA-like splicing changes in wild- Vance, C., Scotter, E. L., Nishimura, A. L., Troakes, C., Mitchell, J. C., Kathe, C., type hFUS mice. Sci. Rep. 7, 2033. doi:10.1038/s41598-017-02195-0 Urwin, H., Manser, C., Miller, C. C., Hortobagyi, T. et al. (2013). ALS mutant Mitchell, J. C., McGoldrick, P., Vance, C., Hortobagyi, T., Sreedharan, J., Rogelj, FUS disrupts nuclear localization and sequesters wild-type FUS within B., Tudor, E. L., Smith, B. N., Klasen, C., Miller, C. C. J. et al. (2013). cytoplasmic stress granules. Hum. Mol. Genet. 22, 2676-2688. doi:10.1093/ Overexpression of human wild-type FUS causes progressive motor neuron hmg/ddt117 degeneration in an age- and dose-dependent fashion. Acta Neuropathol. 125, Walther, D. M., Kasturi, P., Zheng, M., Pinkert, S., Vecchi, G., Ciryam, P., 273-288. doi:10.1007/s00401-012-1043-z Morimoto, R. I., Dobson, C. M., Vendruscolo, M., Mann, M. et al. (2015). Moloney, C., Rayaprolu, S., Howard, J., Fromholt, S., Brown, H., Collins, M., Widespread proteome remodeling and aggregation in aging C. elegans. Cell 161, Cabrera, M., Duffy, C., Siemienski, Z., Miller, D. et al. (2018). Analysis of spinal 919-932. doi:10.1016/j.cell.2015.03.032 and muscle pathology in transgenic mice overexpressing wild-type and ALS- Wang, H., Guo, W., Mitra, J., Hegde, P. M., Vandoorne, T., Eckelmann, B. J., linked mutant MATR3. Acta Neuropathol. Commun. 6, 137. doi:10.1186/s40478- Mitra, S., Tomkinson, A. E., Van Den Bosch, L. and Hegde, M. L. (2018). 018-0631-0 Mutant FUS causes DNA ligation defects to inhibit oxidative damage repair in Murakami, T., Yang, S.-P., Xie, L., Kawano, T., Fu, D., Mukai, A., Bohm, C., Chen, Amyotrophic Lateral Sclerosis. Nat. Commun. 9, 3683. doi:10.1038/s41467-018- F., Robertson, J., Suzuki, H. et al. (2012). ALS mutations in FUS cause neuronal 06111-6 dysfunction and death in Caenorhabditis elegans by a dominant gain-of-function Watanabe, S., Liu, Q., Davis, M. W., Hollopeter, G., Thomas, N., Jorgensen, mechanism. Hum. Mol. Genet. 21, 1-9. doi:10.1093/hmg/ddr417 N. B. and Jorgensen, E. M. (2013a). Ultrafast endocytosis at Caenorhabditis Murakami, T., Qamar, S., Lin, J. Q., Schierle, G. S. K., Rees, E., Miyashita, A., elegans neuromuscular junctions. eLife 2, e00723. doi:10.7554/eLife.00723 Costa, A. R., Dodd, R. B., Chan, F. T. S., Michel, C. H. et al. (2015). ALS/FTD Watanabe, S., Rost, B. R., Camacho-Perez, M., Davis, M. W., Sohl-Kielczynski, ́ ̈ mutation-induced phase transition of FUS liquid droplets and reversible hydrogels B., Rosenmund, C. and Jorgensen, E. M. (2013b). Ultrafast endocytosis at into irreversible hydrogels impairs RNP granule function. Neuron 88, 678-690. mouse hippocampal synapses. Nature 504, 242-247. doi:10.1038/nature12809 doi:10.1016/j.neuron.2015.10.030 Watanabe, S., Trimbuch, T., Camacho-Pérez, M., Rost, B. R., Brokowski, B., Naujock, M., Stanslowsky, N., Bufler, S., Naumann, M., Reinhardt, P., Sohl-Kielczynski, B., Felies, A., Davis, M. W., Rosenmund, C. and Jorgensen, Sterneckert, J., Kefalakes, E., Kassebaum, C., Bursch, F., Lojewski, X. E. M. (2014). Clathrin regenerates synaptic vesicles from endosomes. Nature 515, et al. (2016). 4-Aminopyridine induced activity rescues hypoexcitable motor 228-233. doi:10.1038/nature13846 neurons from amyotrophic lateral sclerosis patient-derived induced pluripotent Weimer, R. M. (2006). Preservation of C. elegans tissue via high-pressure freezing stem cells. Stem Cells 34, 1563-1575. doi:10.1002/stem.2354 and freeze-substitution for ultrastructural analysis and immunocytochemistry. In Pham-Gia, T. and Hung, T. L. (2001). The mean and median absolute deviations. C. elegans (ed. K. Strange), pp. 203-221. Humana Press. Math. Comput. Model. 34, 921-936. doi:10.1016/S0895-7177(01)00109-1 Yan, D., Wu, Z., Chisholm, A. D. and Jin, Y. (2009). The DLK-1 kinase promotes Picchiarelli, G., Demestre, M., Zuko, A., Been, M., Higelin, J., Dieterlé, S., Goy, mRNA stability and local translation in C. elegans synapses and axon M.-A., Mallik, M., Sellier, C., Scekic-Zahirovic, J. et al. (2019). FUS-mediated regeneration. Cell 138, 1005-1018. doi:10.1016/j.cell.2009.06.023 Biology Open

Journal

Biology OpenThe Company of Biologists

Published: Dec 15, 2020

There are no references for this article.