Nicolas et al. Acta Neuropathologica Communications (2022) 10:20 https://doi.org/10.1186/s40478-022-01314-x RESEARCH Open Access A postzygotic de novo NCDN mutation identified in a sporadic FTLD patient results in neurochondrin haploinsufficiency and altered FUS granule dynamics Gaël Nicolas1*† , Myriam Sévigny2,3*†, François Lecoquierre1, Florent Marguet4, Andréanne Deschênes2,3,6, Maria Carment del Pelaez2,3, Sébastien Feuillette1, Anaïs Audebrand2,3, Magalie Lecourtois1, Stéphane Rousseau1, Anne‑Claire Richard1, Kévin Cassinari1, Vincent Deramecourt7,8, Charles Duyckaerts9,10, Anne Boland11, Jean‑François Deleuze11, Vincent Meyer11, Jordi Clarimon Echavarria12,13, Ellen Gelpi14,15, Haruhiko Akiyama16, Masato Hasegawa16, Ito Kawakami16, Tsz H. Wong17, Jeroen G. J. Van Rooij17, John C. Van Swieten17,18, Dominique Campion1, Paul A. Dutchak2,3, David Wallon19, Flavie Lavoie‑Cardinal2,3,5, Annie Laquerrière4, Anne Rovelet‑Lecrux1* and Chantelle F. Sephton2,3* Abstract Frontotemporal dementia (FTD) is a heterogeneous clinical disorder characterized by progressive abnormalities in behavior, executive functions, personality, language and/or motricity. A neuropathological subtype of FTD, frontotem‑ poral lobar degeneration (FTLD)-FET, is characterized by protein aggregates consisting of the RNA-binding protein fused in sarcoma (FUS). The cause of FTLD-FET is not well understood and there is a lack of genetic evidence to aid in the investigation of mechanisms of the disease. The goal of this study was to identify genetic variants contributing to FTLD-FET and to investigate their effects on FUS pathology. We performed whole-exome sequencing on a 50-year- old FTLD patient with ubiquitin and FUS-positive neuronal inclusions and unaffected parents, and identified a de novo postzygotic nonsense variant in the NCDN gene encoding Neurochondrin (NCDN), NM_014284.3:c.1206G > A, p.(Trp402*). The variant was associated with a ~ 31% reduction in full-length protein levels in the patient’s brain, sug‑ gesting that this mutation leads to NCDN haploinsufficiency. We examined the effects of NCDN haploinsufficiency on FUS and found that depleting primary cortical neurons of NCDN causes a reduction in the total number of FUS-posi‑ tive cytoplasmic granules. Moreover, we found that these granules were significantly larger and more highly enriched with FUS. We then examined the effects of a loss of FUS function on NCDN in neurons and found that depleting cells of FUS leads to a decrease in NCDN protein and mRNA levels. Our study identifies the NCDN protein as a likely contributor of FTLD-FET pathophysiology. Moreover, we provide evidence for a negative feedback loop of toxicity *Correspondence: anne.roveletlecrux@univ-rouen.fr; chantelle. sephton.1@ulaval.ca † Gaël Nicolas, Myriam Sévigny have contributed equally to this work 1 Inserm U1245 and CHU Rouen, Department of Genetics and CNR‑MAJ, Normandie University, UNIROUEN, F‑76000 Rouen, France 2 Department of Psychiatry and Neuroscience, Laval University, Quebec City, QC, Canada, 2325, rue de l’Université, G1V 0A6 Full list of author information is available at the end of the article © The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. The Creative Commons Public Domain Dedication waiver (http://​creat​iveco​ mmons.​org/​publi​cdoma​in/​zero/1.​0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Nicolas et al. Acta Neuropathologica Communications (2022) 10:20 Page 2 of 18 between NCDN and FUS, where loss of NCDN alters FUS cytoplasmic dynamics, which in turn has an impact on NCDN expression. Keywords: FUS (fused in sarcoma), Frontotemporal dementia (FTD), NCDN (neurochondrin), Norbin, Genetic variant, De novo mutation, FTLD-FET, Neurodegeneration, mGluR1/5, Cytoplasmic granules Introduction repair [42]. In the cytoplasm, FUS regulates mRNA Frontotemporal dementia (FTD) is a common form of transport and stability, miRNA processing and transla- dementia characterized by a progressive neuronal loss, tion regulation through its interactions with RNA [20, primarily across the frontal and temporal lobes, leading 62, 69, 79]. FUS binds several thousand RNAs at coding, to changes in executive functions, personality, abnor- non-coding and 5’- and 3’-UTR regions [27, 32, 46, 55, mal behaviors and language impairments [43, 54]. FTD 63], mediated through its RNA recognition motif (RRM), is a heterogeneous clinical, genetic and pathological dis- zinc finger (ZnF) domain and three arginine-glycine-gly- order and is broadly referred to as frontotemporal lobar cine (RGG) boxes [49]. At its N-terminal, FUS contains a degeneration (FTLD). Subtypes of FTLD are further low complexity domain (LCD) that contributes to its liq- categorized based on the composition of major proteins uid–liquid phase separation (LLPS) properties as well as found in insoluble cellular inclusions. The three main its interactions with RNA and other proteins [17] subtypes of FTLD are defined by the protein composition In this study, we identified a postzygotic de novo muta- of these inclusions: FTLD-Tau, FTLD-TDP (TAR-DNA tion in the NCDN gene (also known as Norbin), which binding protein-43), and FTLD-FET [51]. FTLD-TDP was predicted to result in haploinsufficiency in a patient and FTLD-FET are both characterized by Tau-negative, with sporadic FTLD-FET. Our study investigates its con- ubiquitin-positive deposits and differ by the actual pro- tribution to FTLD-FET and the effects on FUS pathol- teins involved in these deposits. FET refers to the fam- ogy. We show that depleting primary cortical neurons ily of proteins composed of FUS (Fused in Sarcoma), of NCDN leads to significant changes in FUS solubility EWS (Ewing Sarcoma Breakpoint region 1), and TAF15 and association with cytoplasmic granules. Moreover, the (TATA-binding protein-associated factor 15) proteins, depletion of FUS from cells leads to a decrease in NCDN which are the main components of the aggregates [74]. levels. Together, our data suggest that there is a nega- FTLD-FET accounts for 5–10% of all FTLD cases and it tive feedback loop of toxicity between NCDN and FUS, is further subdivided into three subtypes: atypical-FTLD where loss of NCDN alters FUS cytoplasmic dynamics with ubiquitin-positive inclusions (aFTLD-U), basophilic and loss-of-function or aggregation of FUS could pro- inclusion body disease (BIBD), and neuronal intermedi- mote neuronal dysfunction through the misregulation of ate filament inclusion body disease (NIFID) (for review, NCDN expression. see reference [58]). Loss of function or missense variants in the gene Materials and methods encoding FUS, one of the main components of the aggre- Source of materials gates of FTLD-FET cases, typically cause amyotrophic Reagents were obtained from the following sources: Inhibitor Cocktail (11836170001), MISSION® shRNA: lateral sclerosis (ALS) [44, 82, 91], a motor neuron dis- PhosSTOP (4906845001), cOmplete EDTA-free Protease ease that shares overlapping genetic and pathological fea- tures with FTD [3]. Such ALS mutations are inherited or TRCN0000225722 (FUS-KD-1), TRCN0000225724 (FUS- occur de novo [60] and post-mortem analysis of patient KD-2), TRCN0000010598 (hFUS-KD-1), TRCN0000119421 tissues show nuclear depletion and cytoplasmic aggrega- (NCDN-KD-1), TRCN0000119417 (NCDN-KD-2) and tion of FUS [35, 78], similar to that of FTLD-FET [31, 51, SHC002 (pLKO.1-puro Non-Mammalian shRNA Con- 66]. However, very few FUS mutations have been identi- trol Plasmid) are from Sigma-Aldrich; Cycloheximide fied in FTLD-FET patients or even in cases of ALS-FTLD (CHX) (AC357420010) from Thermo Fisher Scientific; [31, 81]. PDL- and Laminin-PDL coated coverslips from Neuvitro FUS is a ubiquitously expressed DNA/RNA binding Corporation (GG-12-PDL and GG-12-LAMININ). The fol- protein that is predominantly localized to the nucleus of lowing antibodies were used for immunofluorescence exper- cells [11, 74]. It is capable of nucleocytoplasmic shuttling iments: antibodies to FUS/TLS (HPA008784) and MAP2 Fluor® 488 (A-11034) and goat anti-Mouse Alexa Fluor® via its C-terminal proline-tyrosine nuclear localization (MAB3418) from Sigma-Aldrich; goat anti-rabbit Alexa sequence (PY-NLS) and nuclear export sequence (NES) [16, 41, 93]. In the nucleus, FUS regulates transcription, 546 (A-11030) from Thermo Fisher Scientific. The follow- [4, 13, 75, 89] splicing [14, 32, 45] and DNA damage ing antibodies were used for western blotting experiments: Nicolas et al. Acta Neuropathologica Communications (2022) 10:20 Page 3 of 18 FUS/TLS (A300-302A) from Bethyl Laboratories;and FUS/ Protein extraction and western blotting from brain TLS (sc-47711) from Santa-Cruz; NCDN (gift from Dr Fla- Sequential extraction of proteins from frontal cortex was performed as described in [57]. Briefly, frozen brain Secondary antibodies: IRDye® 680RD Goat anti-Mouse IgG jolet), β-actin (2066) and GAPDH (G9545) from Sigma. samples were extracted at 100 mg brain samples in 1 mL (926-68070), IRDye® 800CW Goat anti-Rabbit IgG (926- volume buffer, each extraction step being followed by a 32211) from LI-COR Biosciences. Immunohistochemical centrifugation step et 120 000 g, 30 min, 4 °C. Extraction studies were carried out using antibodies directed against buffer all contained a protease inhibitor cocktail (P8340, α-synuclein (diluted 1/75, Eurobio, les Ulis, France), the Sigma, Saint-Louis, MO, USA) and included (i) high-salt PHF tau (AT8, 1/20, Innogenetics, Gent, Belgium), ubiqui- (HS) buffer containing 50 mM Tris–HCl, 750 mM NaCl, tin (1/100, Agilent, Les Ulys, France), α-internexin (1/75; Life 10 mM NaF, 5 mM EDTA, pH 7.4, (ii) HS buffer contain- technologies- Invitrogen, Courtaboeuf, Villebon-sur-Yvette, ing 1% Triton-X (HS-TX), (iii) RIPA Lysis and extrac- France), TDP43 (1/1000; Proteintech Europ, Manches- tion buffer (from ThermoFisher Scientific, Waltham, ter, UK) and FUS (1/100; Euromedex, Souffelweyersheim, MA USA), (iv) 2% sodium dodecyl sulphate buffer (2% France). SDS). At each extraction step pellets were washed with the corresponding buffer to prevent carry over. Finally, the 2% SDS insoluble pellet was extracted in 70% for- Neuropathological examination mic acid (Ac. Form) and evaporated in a SpeedVac sys- and immunohistochemistry tem. The dried pellet was resuspended in 3X Laemmli An autopsy restricted to the brain and cervical spinal buffer containing 100 mM DTT. For immunoblot analy- cord was performed. After extraction of the brain which sis, equivalent amounts of each fraction were resolved by weighed 1080 g, and after excluding gross macroscopi- 10% sodium dodecyl sulphate–polyacrylamide gel elec- cal asymmetries, 1 cm-thick coronal slices obtained from trophoresis and transferred to nitrocellulose membranes the left hemisphere as well as a cerebellar sample were (Bio-Rad Laboratories, Hercules, CA, USA). Membranes stored at −70 °C until use. The right hemisphere as well were blocked with phosphate-buffered saline contain- as the whole brainstem and cerebellum were fixed in a ing 0.05% Tween and 5% non-fat dried milk and probed 10% formaldehyde solution buffer. Tissue samples were with anti-FUS antibody (1/5000; Bethyl Laboratories, by Stain-Free detection of total proteins using a Geldoc™ taken from multiple areas including anterior upper and Inc. Montgomery, TX, USA). Gel loading was normalized middle frontal gyrus, superior temporal gyrus, tempo- obtained in each lane was quantified with the ImageLab™ ral pole, inferior parietal gyrus, anterior cingular gyrus, EZ Imager (BioRad laboratories). The Stain-Free signal insular and motor cortex, calcarine fissure, hippocampus, nucleus basalis of Meynert, amygdala, basal ganglia, cer- software (Bio-Rad Laboratories). Primary antibody was ebral peduncles, pons, medulla oblongata and cerebellum detected with horseradish peroxidase-conjugated anti- (vermis, right hemisphere and dentate nucleus). Seven- rabbit (Jackson ImmunoResearch, West Grove, PA, USA). micrometer sections were cut from paraffin-embedded Signals were detected with chemiluminescence reagents blocks and stained with haematoxylin–eosin. Immuno- (ECL Clarity, Bio-Rad Laboratories) and acquired with histochemical studies were carried out using antibod- a GBOX (Syngene, Cambridge, UK), monitored by the ies directed against α-synuclein (diluted 1/75, Eurobio, Gene Snap software (Syngene). The signal intensity was les Ulis, France), the PHF tau (AT8, 1/20, Innogenet- quantified using the Genetools software (Syngene). ics, Gent, Belgium), ubiquitin (1/100, Agilent, Les Ulys, France), α-internexin (1/75; Life technologies- Invitro- gen, Courtaboeuf, Villebon-sur-Yvette, France), TDP43 Exome sequencing and genetic analyses (1/1000; Proteintech Europ, Manchester, UK), FUS The legal guardian of the patient and both parents pro- (1/100; Euromedex, Souffelweyersheim, France), and vided written consent for genetic analyses in a research TAF15 (1/200; Ozyme, St Cyr l’école, France). Immu- setting (RBM-0259; this study was approved by the Ile nohistochemical procedures included a microwave de France II ethics committee). DNA was isolated from pre-treatment protocol to aid antigen retrieval (pre- whole blood of the patient and both unaffected parents treatment CC1 kit, Ventana Medical Systems Inc, Tucson using standard procedures. Exomes were sequenced AZ). Incubations were performed for 20, 32 or 60 min at using the Illumina technology following capture using room temperature using the Ventana Benchmark XT sys- an Agilent Human all exons capture kit, V4UTR, by the tem. After incubation, slides were processed by means of Integragen society (Evry, France) with an average depth the Ultraview Universal DAB detection kit (Ventana). of coverage of 100x. Parenthood was checked prior to Nicolas et al. Acta Neuropathologica Communications (2022) 10:20 Page 4 of 18 WES analysis using informative microsatellite analy- same procedures as the patient carrying the NCDN non- sis in all three members of the trio and was verified in sense variant. The other patients were also sequenced exome sequencing data. Reads were mapped to the 1000 using the Illumina technology following capture using Genomes GRCh37 build using BWA 0.7.5a.10. Picard an Agilent Human all exons capture kit, V5UTR, at the Tools 1.101 was used to flag duplicate reads. We applied CNRGH (CEA, Evry, France) with an average depth of GATK for indel realignment, base quality score recalibra- coverage of 130x. Bioinformatics pipelines applied to all tion and single nucleotide polymorphisms, and indels WES were the same as described above with the excep- discovery using the Haplotype Caller across all samples tion of sequencing of parents, whose DNA was not simultaneously according to GATK 3.3 Best Practices available. recommendations. Variants were annotated using a In addition, we performed Sanger sequencing of the homemade pipeline using SnpEff, SnpSift and numer- NCDN gene in three patients from Spain, two previously ous data sources including clinVar, OMIM, DenovoDB reported Japanese patients [40] and in four patients from and GTEX tissular expression and gnomAD variants France, then we analyzed the NCDN gene sequencing frequency. In order to identify de novo candidate vari- data out of WES data previously generated in 5 Dutch ants, we selected high confidence variants in the proband patients. The last 5 patients underwent exome sequenc- (genotype quality ≥ 90 and read depth ≥ 10), then sub- ing using a SeqCap V2 capture kit and Illumina sequenc- tracted all variants detected in at least one parent. ing with an average depth of coverage of 60x. Rare (allele frequency in gnomAD < 1%) variants in the following genes were interpreted and no putatively path- Primary neuron culture ogenic variant was found: MAPT, GRN, VCP, FUS, TAR- Dissociated rat or mouse cortical neurons were pre- DBP, CHMP2B, SQSTM1, OPTN, CHCHD10, HNRNPA, pared from neonatal pup brains as described previously HNRNPA2B1, UBQLN2, TBK1. [56]. Cells were plated at a density of 62 cells/mm2 on For targeted Sanger sequencing and Snapshot analy- poly-D-lysine (PDL)-coated culture plates or at a den- ses, the proband’s DNA was isolated from blood using sity of 50 cells/mm2 on glass coverslips laminin-coated supplemented with serum-free B-27™ (50:1; Gibco, the Flexigen kit (Qiagen), and from frozen brain sam- on PDL layer (Neuvitro Corporation). Neurobasal media ples using the DNA Blood and Tissue kit (Qiagen). Brain regions included frontal cortex, hippocampus, occipi- 17504001), penicillin/streptomycin (50U/mL; 50 μg/mL; tal cortex, parietal cortex and temporal cortex. Snap- Gibco, 15140148) and 0.5 mM L-GlutaMAX (Invitrogen, shot analyses were applied to DNA isolated from blood 35050061) was used as growth medium. When plating and from all the above-mentioned brain regions. The the cells, fetal bovine serum (5%; Hyclone SH30071.03) presence of the G > A mutation was analysed using the was added. At DIV5, half of the culture medium was SNaPshot (PE Applied Biosystems) technique, accord- removed and replaced by serum-free growth medium ing to the manufacturer’s instructions. Briefly, blood and containing Ara-C (5 μM; Sigma, C1768) to limit prolifer- brain-extracted DNA were PCR amplified using prim- ation of non-neuronal cells. Neurons were then fed twice ers flanking the mutation (Forward primer: 5’-TGG​TCC​ a week by replacing half of the culture medium with fresh TGC​TCC​ATC​TCA​AG-3’; Reverse primer: 5’-TAG​AGG​ serum- and Ara-C-free growth medium. GTC​TTG​GCA​TAG​CG-3’). The PCR product was then submitted to primer extension with fluorescent ddNTPs Cell culture and the following primer: 5’-GTG​CGG​ATC​CTG​GGT​ Neuro-2a and HEK293T cell were cultured in complete GCC​TG-3’. The extended primer was finally submitted medium: 10% fetal bovine serum (Gibco, 12483020) and to electrophoresis on an automated sequencer (ABI 3500; DMEM (Gibco, 11965-092) and grown under standard Applied Biosystems), and fluorescence was analyzed culture conditions (37 °C, 5% CO2, 95% air). using the GeneMapper software (Applied Biosystems). Lentivirus production and infection Replication analysis For lentivirus production, HEK293T cells were cultured We gathered exome sequencing data or DNA samples in complete medium and grown to 70% confluence, fol- from multiple international cases with FTLD-FET (Addi- lowed by co-transfection with lentivirus packaging vec- tional file 1: table 1) [40]. All patients or legal guardians tors (VSVG and Δ8.9) and pLKO.1-puro vectors (CTL, provided written consent for genetic analyses. FUS-KD1, FUS-KD2, hFUS-KD1, NCDN-KD-1 and Overall, we performed WES (i) from DNA isolated NCDN-KD-2) using FUGENE6 (Promega, E2691) fol- from blood of two patients and from brain of 4 patients lowing the manufacturer’s instructions. 24 h post-trans- from France and (ii) from DNA isolated from brain of 5 fection, the medium was replaced by neuron growth patients from Spain. One WES was sequenced using the medium and 48 h post-transfection, the condition Nicolas et al. Acta Neuropathologica Communications (2022) 10:20 Page 5 of 18 medium was filtered through a 0.45 μm filter, snap frozen rus titer was obtained using the NucleoSpin® RNA Virus in liquid nitrogen and stored at -80 °C until use. Lentivi- prep kit (Takara Bio, 740956.10) and Lenti-X™ qRT-PCR Titration Kit (Takara Bio, 631235). Upon titer determi- nation, approximately 2–4 × ­108 viral copies/mL were added to DIV8-9 primary cortical neurons. Immunofluorescence staining of neurons Immunofluorescence of primary cortical neurons were performed as previously described [67, 69]. Cells were grown on PDL-coated coverslips and fixed with 4% para- formaldehyde (PFA) solution (4% sucrose, 100 mM phos- phate buffer, 2 mM NaEGTA pH7.4) for 10 min at room temperature (RT). Samples were washed three times for 5 min with phosphate buffer saline (PBS) and then incubated during 30 min in blocking/permeablization solution (0.1% Triton X-100 and 2% normal goat serum (NGS)) at RT. Samples were incubated in primary anti- 10 min with 1X PBS and incubated in Alexa Fluor® 488 body at 4°C, overnight. Samples were washed three times secondary antibodies diluted in blocking/permeabiliza- Fig. 1 Representative macroscopic and microscopic findings from tion solution for 1 h, RT. Coverslips were washed three the FTLD patient’s brain. a External view of the right hemisphere ProLong™ Gold Antifade mounting media containing times for 10 min with 1X PBS and then mounted with showing frontal and temporal atrophy with sparing of the cingular cortex along with the parietal and occipital lobes. b–c Coronal DAPI (Thermo Fisher, P36935). sections passing through the atrophic caudate nucleus, putamen and pallidum (arrow, b) and atrophic hippocampus (arrow, c). d–g Histological examination [H&E, OM × 20] showing absent lamination Confocal microscopy image acquisition of the anterior frontal cortex with vacuolization of layers II and III Neurons were imaged using a ZEISS LSM700 inverted (arrow, d), severe gliosis in the caudate nucleus (e) and neuronal confocal microscope using constant acquisition set- depletion replaced by gliosis of the CA1 field of the Ammon’s tings within image groups from each type of immuno- horn (arrows, f and g; H&E, OM × 20; OM × 100 respectively). H&E: haematoxylin and eosin stains; OM: original magnification fluorescence measurement. Images were acquired using a 63 × oil immersion objective (1.4 NA, Zeiss), in z-stack mode with an interval of 0.45 μm and analysed by ZEN imaging software (Zeiss) to generate maximal intensity followed by the application of a triangle threshold and projections (MIP) or 3D images. Confocal images show- removal of objects smaller than 30 pixels. For the analy- ing the cytosolic localization of FUS were captured using sis of NCDN intensity, the nucleus segmentation map settings where the nuclear FUS signal was saturated was subtracted from the foreground mask to obtain the (Fig. 5). Whereas the confocal images showing the deple- segmentation mask of cytosolic NCDN. For each experi- tion of FUS by shRNA-FUS lentiviruses were captured mental week, the fluorescence intensity in each image using settings that did not contain saturated pixels for was normalized by the median cytosolic NCDN inten- FUS (Fig. 7). sity of all neurons treated with the non-targeted scram- ble. Outliers greater than 2 standard deviations from the Quantitative image analysis mean of each condition were removed. To estimate the Quantification of cytoplasmic NCDN signal: Using a tri- median distribution with a 95% confidence interval (CI) angle threshold on the MIP of the DAPI signal, we cre- we resampled the data points with replacement 10 000 ated a segmentation mask of the nuclei. Objects with times. areas smaller than 20 pixels were removed. Binary hole Quantification of FUS granules fluorescence intensity: filling and dilation (kernel size: 3) was performed on the For the segmentation of FUS granules in the confocal remaining objects to generate the nuclei segmentation z-stacks, we used the 3D spot filter of the Allen Cell map. The foreground of the NCDN channel signal was Structure Segmenter (Chen et al., 2020 https://​doi.​org/​ extracted in each image slice by 3D Gaussian smooth- 10.​1101/​491035). We used auto contrast pre-processing ing of all the frames in the z-stack with a sigma of 0.5 and 3D gaussian smoothing (σ = 1). The segmentation Nicolas et al. Acta Neuropathologica Communications (2022) 10:20 Page 6 of 18 replacement 10,000 times and their bootstrapped dif- ferences were plotted. Quantification of FUS granules volume, number and area: Confocal images of primary neurons were used for 3D analysis using Imaris software. A region of interest (ROI) was defined for all cell bodies of each data set. For each ROI, FUS signal was detected using 488 nm chan- nel using the Imaris spot detection module. An arbitrary mean intensity threshold was used to detect the number of granule spots, defining a minimum detectable size of 0.01um and subtracting the detection coming from the cell nucleus. A minimum number of 16 cells per group were analysed. The volume, number and area of FUS- positive granules were calculated and represented as mean per cell analysed. An Ordinary One-way ANOVA with multiple comparisons using Turkey test was used to compare the different groups for the statistical analysis. Western blotting and subcellular fractionation Cortical neurons (DIV16) were washed once with ice cold 1X PBS pH7.4 (Gibco, 10010–023) and then lysed Fig. 2 Immunostaining of brain tissue from the FTLD-FET patient. in radioimmunoprecipitation assay buffer (RIPA) buffer a–d Immunohistochemical lesions in the hippocampus and frontal (20 mM Tris–HCL pH 8.0, 1 mM NaEDTA pH8.0, cortex. Ubiquitin immunolabeling displays rounded intra-cytoplasmic 0.5 mM NaEGTA pH 8.0, 1% Triton X-100, 150 mM inclusions within the dentate gyrus (black arrows, a) but with no deformation of the nucleus [OM × 400]. Similar intra-cytoplasmic NaCL, 1X protease inhibitors EDTA-free, 1X PhosSTOP inclusions were observed in the anterior frontal cortex using (Roche)) to obtain total cell lysates (TCL). Neuro-2a cells FUS immunohistochemistry (black arrows, b), associated to were lysed in polyribosome lysis buffer (PLB) (20 mM FUS-positive either rounded (black arrow, c) or elongated granular Tris–HCL pH7.4, 5 mM MgCl2, 100 mM KCl, 1% NP-40, intracytoplasmic inclusions (red arrow, c) [OM × 400] or filamentous 1 mM DTT, 20U/µl SUPERase Inhibitor, 1X protease curvilinear intranuclear inclusions in the dentate gyrus (black arrow, d) [OM × 400]. e TAF15 immunostaining displaying normal, finely inhibitors EDTA-free, 1X PhosSTOP (Roche)), then sub- granular nuclear staining (asterisk) contrasting with loss of TAF15 cellular fractionation of lysates was performed by centrif- nuclear immunostaining resulting in cytoplasmic aggregates (black ugation (10,000 × g, 10 min, 4 °C) to obtain supernatant arrows) with f, intracortical neuritic accumulations (black arrow). OM: (S1) and pellet (P1) fractions. The P1 fraction was then original magnification resuspended in RIPA buffer. All protein extracts were prepared in 1X Laemmli buffer and boiled (5 min, 95 °C) for western blot analysis. Samples were resolved on SDS– parameters were adjusted for each experimental week. polyacrylamide gels and transferred to nitrocellulose (range: 0.75–1,0.015–0.045). Thresholds were applied membranes followed by standard western blotting proce- to remove detections of non-specific staining or meas- dure as described in [69]. Membranes were imaged using urement artifacts from the segmented clusters (area the LI-COR Odyssey imaging system. Analysis of signal under 5 pixels or over 300 pixels, Intensity above intensity was done using Image Studio Lite Software Ver- 30,000 photon counts). The maximum intensity of each sion 5.2. Quantification of FUS expression in the subcel- cluster was computed to generate the intensity distri- lular fractions was normalized to GAPDH for TCL and bution histogram of the FUS granules. The histogram S1 and ponceau red for P1. was fitted using a skew normal function and the peak of the fitted distribution was calculated. For each experi- Total RNA from cells were extracted using TRIzol™ Quantitative RT‑PCR mental week, the peak position of the fitted distribution of all the images of the CTL condition were used for Reagent (Invitrogen) following the manufacturer pro- normalization. Outliers greater than 2 standard devia- tocol. RNA extracts were treated with DNase I (Roche, tions from the mean of each condition were removed. 4716728001), and cDNA was synthesized using High- To estimate distribution with a 95% confidence inter- Capacity cDNA Reverse Transcription Kit (Applied Bio- val (CI) we bootstrapped the data points [5]. For the systems, 4368813). Quantitative RT-PCR reactions of 10 bootstrap analysis the data points were resampled with μL contained 25 ng cDNA, 150 nM of each primer and 5 Nicolas et al. Acta Neuropathologica Communications (2022) 10:20 Page 7 of 18 µL PowerUp™ SYBR™ Green Master Mix (Applied Bio- sporadic behavioral variant of FTD (bvFTD). The patient system, A25741). All reactions were performed in tripli- had no personal history of psychiatric or neurologi- cate on a QuantStudio 5 Real-Time PCR System (Applied cal diseases including a normal development except for Biosystem) and relative mRNA levels were calculated by a history of dyslexia. In her early forties, she started to the comparative threshold cycle method using U36B as show behavioral troubles with a progressive psychomo- the internal control. tor slowing, impairment of interpersonal relationships, The following primer pairs were used: mistakes at work, and loss of self-care. By the age of 44, U36B (Forward 5’- CGT​CCT​CGT​TGG​AGT​GAC​A-3’; she experienced a sharp decline in her performance at Reverse 5’- CGG​TGC​GTC​AGG​GATTG -3’), work and severe interpersonal relationship issues led to TDP-43 (Forward 5’-AAC​TGA​GCA​GGA​TCT​GAA​ termination of her employment. Shortly thereafter, she AGA​CTA​TTT-3’; Reverse 5’-CCC​TTT​CGA​GTG​ACC​ was hospitalized at the age of 45 years and displayed AGT​TTTAA-3’), apragmatism, loss of social adaptation, lack of motivation FUS (Forward 5’-CAG​CTA​TCG​ACT​GGT​TTG​ATG-3’; and aggressive behavior. Treatment with the antidepres- Reverse 5’-CGA​TTG​AAG​TCA​GCT​CGG​CG-3’). sant citalopram or antipsychotic drugs aripiprazole and NCDN (Forward 5’-GCG​CCA​TCG​TGA​AGT​GTG​A-3’; olanzapine or sismotherapy, did not lead to any improve- Reverse 5’-CCC​GCC​TTG​CCC​AAC​TGT​-3’). ment in behavior. A CT scan performed at age 47 showed fronto-temporal atrophy (data not shown). Neurological Statistical analyses and psychological assessment showed motor and verbal At least n = 3 biological experiments were performed initiation impairment gathering apathy, reduced spon- for every statistical analysis using Microsoft Excel 2013; taneous speech, and loss of emotional reactivity. How- this means independent cell cultures were performed ever, she occasionally presented with episodes of extreme for each biological experiment. A Student’s t-test at 95% impulsivity, aberrant motor behavior and unmotivated confidence was used for the comparison of two groups. laughter. Signs of delusion or mood disorders were never Statistical analysis was performed for fractionation reported. Gradually, the cognitive and behavioral syn- experiments comparing the relative protein intensities drome increased leading to severe perseverations, hyper- and for QPCR. Each statistical analysis and the number orality and a complete loss of language. The patient died of biological experiments are indicated in the figure leg- at the age of 50 and an autopsy restricted to the brain was ends. All statistical analysis giving p < 0.05 are significant. performed. She had no family history of any neurodegen- erative or psychiatric disease, except late onset dementia Results in the maternal grandfather who died at age 96. Both par- Identification of a de novo variant in the NCDN gene ents were unrelated and were healthy at age 80 and 75, in a FTLD‑FET patient respectively for the father and the mother, at the time of We aimed to identify a genetic determinant of FTLD- death of their daughter. FET in a female patient (AH-11-02) with early-onset, Fig. 3 Increased insolubility of FUS in the FTLD patient’s brain compared to controls a Western blot analysis of proteins extracted from frozen brain tissue, from controls (CTL-1, -2) or affected patient (AH-11–02). FUS staining was performed on samples resulting from sequential brain extraction, using buffers with increasing solubilisation properties. Total protein was used as loading control with Stain-Free technology (SF). Three independent extractions per individual were performed. Representative blots and Stain-Free staining are shown. b Graph representing the ratio of FUS bands intensity in the soluble fraction (HS—HS-Tx—RIPA) relative to the insoluble fractions (2%SDS—Ac.Form) after SF normalization [means ± standard error of the mean (SEM)] Nicolas et al. Acta Neuropathologica Communications (2022) 10:20 Page 8 of 18 Post-mortem examination of the brain revealed severe atrophy in the frontal and temporal lobes (Fig. 1a). The substantia nigra and the locus coeruleus were moderately depigmented. On coronal sections of the right hemi- sphere, the cerebral white matter was irregularly discol- oured. The caudate nucleus, putamen and pallidum were atrophic (Fig. 1b), as well as the hippocampus (Fig. 1c). Ventricular dilatation was also noted, more pronounced in the anterior horns. The cerebellum and brainstem dis- played no macroscopical lesions (data not shown). Histological examination revealed neuronal loss in the pigmented nuclei of the brainstem, particularly the sub- stantia nigra pars compacta, associated with interstitial pigment deposition and reactive gliosis, but Lewy bod- ies were never observed, even by means of α-synuclein immunohistochemistry (data not shown). In the cer- ebellum, neither spongiosis nor internal granular cell layer atrophy were noted. Purkinje cells were preserved. No spinal cord lesions were noted. Frontal and tempo- ral cortices except for the superior and middle temporal gyri were severely affected (Fig. 1d). The lesions associ- ated severe neuronal loss, microvacuolization of layers II and III, as well as vanishing cortical lamination. The rare residual pyramidal neurons of layer III contained lipofus- cin accumulation. Severe neuronal loss was also observed in the basal ganglia, with almost no neurons and severe gliosis of the caudate nucleus (Fig. 1e). In the hippocam- pus, the dentate gyrus, the CA4, CA3 and CA2 fields of the Ammon’s horn were preserved, whereas the CA1 field, presubiculum and subiculum were almost devoid of neurons and were replaced by gliosis, consistent with hippocampal sclerosis (Fig. 1f, g). The parietal and cingu- lar cortices had a normal lamination with mild neuronal loss and the occipital cortex was spared. Pathological protein aggregates were observed in the Fig.4 Detection of the NCDN de novo variant (c.1206G > A; affected brain regions. Ubiquitin-positive, round-shaped p.(Trp402*)) in the FTLD-FET patient. a Confirmation of the variant inclusions were observed in the neuronal perikarya of the in NCDN by Sanger sequencing (left panels) and SNaPshot (right panels) in blood and in different brain regions. In SNaPshot panels, dentate gyrus, anterior frontal cortex and temporal pole blue represents the WT allele (G) and green the mutated allele (Fig. 2a and b). These inclusions were negative for Tau (A). b Western blot analysis of NCDN and β-Actin from control and TDP-43 (data not shown). However, intra-cytoplas- (CTL-1, -2) or FTLD-FET affected patient (AH-11–02) frontal cortical mic and intranuclear FUS-positive neuronal inclusions tissues. Arrowhead indicates full-length NCDN and the star shows were observed in the dentate gyrus and in the frontal cor- the absence of truncated NCDN at the expected size of 43KDa. c Quantification of NCDN expression shows a ~ 31% decrease in NCDN tex (Fig. 2c and d). Intra-cytoplasmic round-shaped neu- protein level from n = 4 experimental replicates per group. Error bars ronal inclusions were concentrated in the vicinity of the represent the mean ± SEM nuclei. They were relatively homogeneous whereas intra- nuclear inclusions had a granular or vesicular pattern, some of them with filamentous or curvilinear appear- ance (Fig. 2c and d). Moreover, some glial inclusions To gain insights into changes of FUS solubility, bio- were observed in the white matter. These inclusions were chemical analysis was performed on frozen brain extracts immunolabelled by the anti-TAF15 antibody (Fig. 2e and of patient AH-11-02 and in two control individuals f ). These findings led to a final diagnosis of FTLD-FET, devoid of neurodegenerative disease. Proteins were atypical FTLD-U subtype according to Neumann and extracted sequentially, using buffers with increasing solu- Mackenzie [58]. bilisation properties, as described in [57]. As expected, Nicolas et al. Acta Neuropathologica Communications (2022) 10:20 Page 9 of 18 FUS could be detected as a 70-kDa band, and was mainly NMD if altered splicing would introduce a premature recovered in the most soluble fractions, as well as in the stop codon. Overall, this suggests that there is no human 2% SDS fraction. In the patient extracts, we noted a shift being reported in a control database so far with a trun- of FUS staining towards the more insoluble fractions, cating variant predicted to trigger NMD as is the case for which resulted in a clear decrease of the soluble/insoluble the c.1206 > A, p.(Trp402*) variant. ratio in the patient compared to controls (Fig. 3a and b). We hypothesized that this de novo variant would To better understand the cause leading to FTLD- result in haploinsufficiency. The NCDN gene encodes FET, we performed Sanger sequencing of the FUS gene the 75 kDa Neurochondrin protein, also known as Nor- in DNA isolated from blood and identified no candi- bin. NCDN is highly expressed in the brain and even date variant. In order to target de novo genetic vari- appears to be brain-specific according to the Genotype- ants responsible for causing FTLD-FET, we performed Tissue Expression (GTEX) database [10]. We assessed the whole exome sequencing (WES) on DNA isolated from expression of NCDN in the patient’s brain samples and blood from the patient and her unaffected parents. We found a ~ 31% reduction in full-length protein levels com- first explored all neurodegenerative dementia genes pared to control brains (Fig. 4b and c), consistent with and found no putatively pathogenic variant in patient our hypothesis. AH-11-02 (see methods). We then focused on can- In a replication attempt, we searched for NCDN vari- didate de novo variants. This led to the identification ants in 25 additional unrelated unsolved FTLD-FET of a unique de novo variant, in the NCDN gene. This patients (see methods, Additional file 1: table 1) by whole novel variant (NM_014284.3:c.1206G > A, p.(Trp402*); exome or Sanger sequencing and identified no non-syn- Chr1(GRCh37):g.36028055G > A) was indeed absent onymous NCDN variant, suggesting that such variants in the parental samples and present in 24.5% (12/49) of remain extremely rare. the reads in the proband, suggesting that the mutation Overall, our analysis on this FTLD-FET patient led to could be post-zygotic in ~ 39% of the patient’s cells. We the identification of a new rare NCDN variant associated confirmed the presence of the variant by Sanger sequenc- with decreased NCDN protein levels in the brain, FUS ing and using the SNaPshot technique in DNA extracted inclusions and altered FUS solubility. from blood and from different regions of the brain cortex of the patient (Fig. 4a), confirming that the mutation is NCDN haploinsufficiency affects FUS expression an early post-zygotic event with an allelic ratio ranging and localization in neurons from 24.6% to 31.1% in different brain cortical regions NCDN is highly expressed and distributed throughout and 24.2% in blood. the cytoplasm, axons, dendrites, spines and perisynapse This variant introduces a premature stop codon in exon and has roles in dendrite morphogenesis, neural out- 4 out of 7 exons regarding this transcript and is hence growth and synaptic plasticity [34, 61, 71, 72, 83, 84]. It predicted to trigger nonsense-mediated decay (NMD). has been shown to modulate metabotropic glutamate It is absent from the genome aggregation database (gno- receptor 5 (mGluR5), through positively regulating mAD, v2.1), gathering exomes and whole genomes of mGluR5 cell surface expression and downstream signal- 141,456 individuals of diverse ethnicities [38]. More ing [84]. Importantly, FUS has also shown to be regulated importantly, the NCDN gene appears to be very intol- through mGLUR1/5 activity and its downstream signal- erant to loss-of-function variants, with an observed/ ing pathways [19, 68, 69]. Upon, mGluR1/5 stimulation, expected ratio of 0.04 (upper boundary of the 95% con- FUS expression increases in dendrites and synapses of fidence interval: 0.17) and a maximum loss of function neurons [2, 19, 68]. Acting downstream of mGluR1/5 is intolerance score of 1 in gnomAD [38]. This indicates that the mechanistic target of rapamycin (mTOR) signaling the general population is highly depleted in truncating pathway, which is the major regulator of cytoplasmic variants (nonsense, frameshift), as also further confirmed and local translation at synapses [28]. Under conditions by the absence of any deletion in copy number variant where the mTOR pathway is inhibited, FUS is shown to databases including the Database of Genomic Variants localize to the cytoplasm where it promotes translation (DGV, [50]) and gnomAD-SV [9]. When looking further repression [69]. into gnomAD, we could identify two frameshift variants To determine the impact of NCDN haploinsufficiency seen 4 times each and a splicing variant observed once. on FUS pathology, we depleted primary rat cortical Both frameshift variants mapped to the penultimate exon neurons of NCDN using lentiviruses containing shR- (exon 6/7) within the last 50 base pairs and were hence NAs targeted against NCDN (NCDN-KD) and exam- not predicted to trigger NMD. The splice site variant ined the effects on FUS localization and expression. was observed once and mapped to the 5’ canonical splic- We found that depleting neurons of NCDN resulted in ing site of intron 6; it was also not predicted to trigger changes in FUS-positive cytoplasmic granules (Fig. 5a Nicolas et al. Acta Neuropathologica Communications (2022) 10:20 Page 10 of 18 Fig. 5 NCDN depletion in neurons affects FUS cytoplasmic granule dynamics. Lentivirus containing shRNAs targeting NCDN (NCDN-KD1 or -KD2) or non-targeted scramble (CTL-KD) were used to deplete primary rat cortical neurons (RCN) of NCDN. a Confocal images of RCN (DIV16) stained with antibodies against FUS (HPA008784, green), MAP2 (red) and DAPI (blue). Scale bar = 10 μm. b IMARIS generated 3D surface images of FUS-positive cytoplasmic granules from RCN. Quantification of the number of granules per neuron (c), mean area (d) and volume of granules (e). f Bootstrapped difference of maximum intensity of segmented FUS cytoplasmic granules from NCDN-KD compared to CTL-KD RCN showing the kernel density plot of the bootstrapped differences (shaded grey area), the minimum and maximum resampled differences (black horizontal lines), and the 95% confidence interval (red horizontal lines). NCDN-KD1,95% CI: 0.0062- 0.1937 p = 0.0098; NCDN-KD2,95% CI: 0.0044–0.1372 p < 0.0001. g Western blot of NCDN, FUS and GAPDH proteins from RCN. h Quantification of FUS protein levels from RCN relative to GAPDH. Statistical analysis was performed using a one-way ANOVA with multiple comparisons using Turkey test (c–e) or a Student’s t test (h) (a, p < 0.05; b, p < 0.01; c, p < 0.005; d, p < 0.001 vs CTL; ns, not significant, p > 0.05 vs CTL). Error bars represent the mean ± SEM. Each experiment was performed from n = 3–4 biological replicates per group Nicolas et al. Acta Neuropathologica Communications (2022) 10:20 Page 11 of 18 Fig. 6 Loss of NCDN affects FUS localization and solubility. Lentivirus containing shRNAs targeting NCDN (NCDN-KD1 or -KD2) or non-targeted scramble (CTL-KD) were used to deplete Neuro-2a cells of NCDN. a CTL-KD and NCDN-KD cells were lysed in PLB followed by fractionation to generate soluble (S1) and insoluble fractions (P1). TCL: total cell lysates. PLB: polyribosome lysis buffer. RIPA: radioimmunoprecipitation assay buffer. b Western blot of NCDN, FUS and GAPDH proteins and corresponding ponceau red staining of membranes. 5% of each fraction was loaded on the gel. The arrow indicates the band corresponding to NCDN. c Quantification of FUS protein levels in fractions. Statistical analysis was performed using a Student’s t test (a, p < 0.05; b, p < 0.01; ns, not significant, p > 0.05 vs CTL). Error bars represent the mean ± SEM. Each experiment was performed from n = 3–4 biological replicates per group and Additional file 1: Fig. S1 and S2). Analysis of these rat cortical neurons (Fig. 7a), but significantly decreased granules in NCDN-deficient neurons showed a signifi- the median fluorescence intensity of cytosolic NCDN cant reduction in the number and an increase in size of in FUS-depleted neurons (Fig. 7a and b). These findings the FUS-positive granules when compared with control corresponded with global decreases in NCDN protein neurons (Fig. 5b–e). Moreover, the signal intensity of expression in these neurons (Fig. 7c and d). FUS within these structures was significantly increased FUS is shown to stabilize the expression of mRNAs [64, (Fig. 5f ), without any significant changes to the protein 79, 90] and dysregulation of FUS can lead to destabiliza- expression of FUS (Fig. 5g and h). These findings suggest tion and NMD of its mRNA targets [25, 36, 92]. There- that depleting neurons of NCDN affects FUS localization fore, we examined the mRNA expression of NCDN in and association with cytoplasmic granules. FUS-depleted neurons and found there to be a reduc- To determine if the localization and solubility of FUS is tion in total NCDN mRNA compared to control neurons affected by a reduction in NCDN, we isolated the soluble (Fig. 7e), a finding that was also consistent in Neuro-2a (S1) and insoluble (P1) fractions from NCDN-depleted cells (Additional file 1: Fig. S3). Consistent with previ- Neuro-2a cells by subcellular fractionation (Fig. 6a). We ous reports, we did not observe any changes in TDP- examined the distribution of FUS in these fractions and 43 mRNA levels in FUS-depleted neurons [39, 46, 65]. found that FUS was enriched in the P1 fraction, of these Together, these findings suggest that FUS is involved cells (Fig. 6b and c), similar to what was observed in the in the regulation of NCDN protein expression and/or patient (Fig. 3). Together, these data suggest that loss of mRNA stability. NCDN can affect FUS localization and granule dynamics in cells. Discussion We report here a de novo NCDN mutation in a spo- FUS loss‑of‑function affects NCDN expression in neurons radic FTLD-FET patient. This nonsense mutation was A pathological hallmark of FTLD-FET as well as ALS- associated with reduced protein levels in the patient’s FUS is the nuclear loss and cytoplasmic aggregation of brain, hence haploinsufficiency as a main mecha- FUS, which is thought to contribute to the disease due to nism of disease. We thus investigated the relation- a toxic loss of nuclear function and a toxic gain-of-cyto- ship between NCDN and FUS and determined that a plasmic function [33, 48, 65, 68, 70]. As an approach to decrease in NCDN expression causes changes in FUS mimic the loss of FUS function shown to contribute to cytoplasmic granule dynamics. Moreover, our findings disease, we depleted cells of FUS and examined the effect reveal that a decrease in FUS expression promotes a on NCDN expression. We found that knocking-down reduction in NCDN levels. Collectively, our data pro- FUS with lentiviruses containing FUS shRNAs (FUS- vide evidence for a negative feed-back loop of toxicity KD) did not affect the distribution of NCDN in primary between NCDN and FUS, where loss of NCDN alters Nicolas et al. Acta Neuropathologica Communications (2022) 10:20 Page 12 of 18 Fig. 7 FUS depletion in neurons affects NCDN protein and mRNA levels. Lentivirus containing shRNAs towards FUS (FUS-KD1 or KD2) or non-targeted scramble (CTL) were used to infect primary rat neurons (RCN). a Confocal images of RCN (DIV16) stained with antibodies against FUS (sc-47711, red), NCDN (green) and DAPI (blue). Scale bar = 10 μm. b Bootstrapped difference of NCDN cytoplasmic intensity of medians from FUS-KD compared to CTL-KD RCN showing the kernel density plot of the bootstrapped differences (shaded grey area), the minimum and maximum resampled differences (black horizontal lines), and the 95% confidence interval (red horizontal lines) in neurons. FUS-KD1,95%CI: 0.0050- 0.1525 p = 0.0003; FUS-KD2,95% CI: 0.0057–0.1629 p = 0.001. c Western blot of NCDN, FUS and GAPDH proteins from RCN. d Quantification of NCDN protein levels from RCN relative to GAPDH. e Quantitative RT-PCR for NCDN, TDP-43 and FUS relative to U36B from primary mouse cortical neurons (DIV16). Statistical analysis was performed using a Student’s t test (a, p < 0.05; d, p < 0.001 vs CTL; ns, not significant, p > 0.05 vs CTL). Error bars represent the mean ± SEM. Each experiment was performed from n = 3–4 biological replicates per group FUS cytoplasmic dynamics, and that the misregulation to further our understanding of the disease and for the of FUS localization further promotes misregulation of development of therapeutic targets. NCDN expression (Additional file 1: Fig. S4). We con- The majority of familial FTD (~ 60%) cases are caused clude that FUS is downstream of NCDN. Exploring by autosomal dominant mutations in GRN, MAPT and this pathway in the context of FTLD-FET is relevant C9orf72 [21, 22]. The genetic cause of FTLD-FET is less Nicolas et al. Acta Neuropathologica Communications (2022) 10:20 Page 13 of 18 clear and FUS mutations are exceptionally found in FTD variant being most likely hypomorphic, and de novo vari- patients [31, 66, 81]. To our knowledge, the NCDN muta- ants affected NCDN function more severely. The authors tion c.1206G > A, p.(Trp402*) identified in this study suggested that two of the variants, located in the mGluR5 has not been previously reported in humans. Moreover, binding domain, specifically altered this pathway. It we could not identify any other case carrying a non- remains unclear whether haploinsufficiency is the most synonymous NCDN variant despite gathering 25 FTLD/ accurate model to mimic these missense variants’ effects, ALS-FUS cases through international collaboration. although it is clear that they are pathogenic through loss FTLD-FET is among the rarest subtypes of FTLD and can of NCDN function. In addition, although there are no itself be subdivided into different subgroups. Although truncating NCDN variant predicted to trigger NMD in our case was diagnosed with aFTLD-U, we conservatively variant databases or deletions in controls, a few patients explored other subtypes of FTLD-FET, namely NIFID with a developmental disorder were reported with dele- and BIBD cases, along with aFTLD-U cases. However, tions encompassing NCDN and multiple other genes. it remains unclear whether NIFID, BIBD and aFTLD- However, it is unclear (i) whether NCDN plays a role U are closely related disorders that could share strong in the developmental phenotype of the latter patients genetic determinants or if the molecular bases are dis- and (ii) whether the same patients as well as recently tinct. Hence, the replication step of our analysis can be described patients with missense variants will eventually considered as limited, if NCDN truncating variants may develop FTLD-FET. Here, the nonsense mutation was be exclusively associated to aFTLD-U. detected as a mosaic. We hypothesize that such an NMD- FTD shares clinical, genetic and pathogenic features triggering variant, if carried in the germ line, might be with ALS [37, 52]. ALS is caused by degeneration of lethal during in-utero development, which could explain upper and lower motor neurons leading to progres- absence from control databases and absence of associa- sive paralysis and death [3]. Unlike FTLD-FET, auto- tion with developmental disorders. somal dominant mutations in FUS account for 5–10% In a non-Mendelian manner, NCDN has been impli- ALS [15, 44, 82]. Most ALS-FUS mutations result in an cated in several neurological disorders including epilepsy, increase in cytoplasmic FUS and the formation of FUS- depression, schizophrenia and spinal muscular atrophy positive cytoplasmic inclusions [15, 44, 82], similar to (SMA) [1, 12, 53, 59, 76, 83–85, 87]. Previous studies those found in FTLD-FET [31, 51, 66]. Our identifica- in mice have shown that forebrain-specific knockout of tion of this de novo mutation in NCDN suggests that a NCDN causes impaired mGluR5 dependent long-term pathway related to NCDN could contribute to FUS mis- potentiation (LTP) and long-term depression (LTD), as localization, which may partly explain how patients can well as schizophrenia-like phenotypic behaviors [84]. display FUS-associated degenerative defects without Neuron-specific NCDN knockout mice have reduced having mutations in the FUS gene. A number of de novo hippocampal neurogenesis and depressive-like behav- mutations in sporadic ALS patients have been identified iours [83]. Overexpression of NCDN in neuronal cells in ALS trios studies, although none affected the NCDN induces neurite outgrowth [72]. Collectively, these stud- gene, to our knowledge [8, 73, 80]. In the ALS variant ies show that NCDN has important roles in dendrite server, which gathers genetic data of 1277 ALS cases, morphogenesis, neural outgrowth and synaptic plastic- three missense NCDN ultra rare variants are reported ity [34, 61, 71, 72, 83, 84]. Our data show that depleting (NM_014284.3:c.1417G > A, p.(Val473Ile) in a sporadic NCDN from neurons affects FUS cellular distribution ALS patient, c.509G > A, p.(Arg170Gln) in a familial ALS and association with cytoplasmic granules. While the patient and c.1507C > G, p.(Pro503Ala) in 3 familial ALS changes in localization of FUS in our neuron cultures did patients), but their inheritance is not mentioned and they not cause cell death, mislocalization of FUS to the cyto- are mainly predicted as benign by bioinformatics tools. plasmic compartment has been shown to cause loss of No truncating variant is reported on this server (ALS dendritic branching, loss of mature spines and contribute Variant Server, Worcester, MA (URL: http://​als.​umass​ to neurodegeneration [32, 48, 65, 68, 70]. Taken together, med.​edu/) [nov, 2021 accessed]. these findings support the hypothesis that NCDN and Until very recently, NCDN was not associated with any FUS are part of the same regulator pathway important for Mendelian disorder. Missense variants were however neuronal function. reported in patients with a neurodevelopmental phe- NCDN is also linked with SMA, a degenerative mus- notype with epilepsy. One family exhibited a missense cular disease caused by a reduction in the survival motor homozygous variant, with unaffected heterozygous par- neuron (SMN) full-length protein [47, 76]. The findings ents, while 3 unrelated patients showed missense de novo from one study show that the function and expression variants with a more severe phenotype [18]. Functional of NCDN and SMN are co-dependent [76]. When SMN analyses suggested a loss-of-function effect, the bi-allelic levels are reduced in cells, it causes a decrease in the Nicolas et al. Acta Neuropathologica Communications (2022) 10:20 Page 14 of 18 amount of cytoplasmic NCDN-positive granules. Con- expression [19, 68, 77], NCDN expression also affects versely, depleting cells of NCDN leads to an increase in FUS localization and activity, which might be due to the SMN nuclear foci in SH-SY5Y cells. Moreover, this study lack of signaling of mGluR5 at the cell surface (Additional identified that NCDN and SMN interact within mobile file 1: Fig. S4) [84]. cytoplasmic granules [76]. Interestingly, FUS and SMN are shown to share a common pathway in the context of Conclusions neurobiology and ALS [6, 7, 23, 88]. FUS is found to asso- In summary, we have identified a rare de novo NCDN ciate with the SMN complex through a direct interaction variant in a FTLD-FET patient that results in haploin- with SMN [88]. Functionally, overexpression of SMN can sufficiency of NCDN. We conclude that NCDN has an rescue the neurite growth defects and reduced dendritic important function in regulating FUS granule dynam- branching induced by ALS-FUS mutants [7]. In our study ics, and that these changes cause a misregulation of we show that depleting cells of FUS affects NCDN pro- NCDN expression. Based on the common biological tein and mRNA levels (Fig. 7). Our findings are consist- functions of both NCDN and FUS, as well as their link ent with published supplemental data from the brains of with FTLD-FUS and ALS-FUS, there is an intrigu- FUS KO mice that shows a decrease in NCDN mRNA ing possibility that these proteins are part of a common [65]. Moreover, NCDN mRNA is among the list of genes regulatory pathway for the maintenance of synaptic found to interact with FUS [46], an interaction that homeostasis. Taken together, our findings suggest that remains to be validated. Further analysis is required to disruption of this pathway would lead to neuronal defects determine if FUS and SMN are directly regulating NCDN and neurodegeneration. translation and whether NCDN:FUS:SMN are part of a converging pathway involved in maintaining synaptic Abbreviations homeostasis. FTD: Frontotemporal dementia; FTLD: Frontotemporal lobar degenera‑ A previous study showed that NCDN interacts with tion; TDP: TAR-DNA binding protein-43; FUS: Fused in sarcoma; EWS: Ewing a subset of group 1 mGluRs, including mGluR5, where Sarcoma Breakpoint region 1; TAF15: TATA-binding protein-associated factor 15); aFTLD-U: Atypical-FTLD with ubiquitin-positive inclusions; BIBD: Basophilic NCDN has been shown to promote mGluR5 cell surface inclusion body disease; NIFID: Neuronal intermediate filament inclusion body expression and activation of mGluR5 through a direct disease; ALS: Amyotrophic lateral sclerosis; PY-NLS: Proline-tyrosine nuclear interaction with this receptor [84]. mGluR5 activation localization sequence; NES: Nuclear export sequence; RRM: RNA recognition motif; ZnF: Zinf finger; RRG​: Arginine-glycine-glycine; LCD: Low complexity at spines has been shown to regulates local protein syn- domain; LLPS: Liquid–liquid phase separation (LLPS); mGluR1/5: Metabo‑ thesis at the synapse, which promotes synaptic strength tropic glutamate receptor group 1 subtypes 1 and 5; mTOR: Mechanistic and promote LTD [30, 86]. Ligand activation of mGluR5 target of rapamycin; NCDN: Neurochondrin; WES: Whole exome sequencing; NMD: Nonsense-mediated decay; DGV: Database of genomic variants; GTEX: initiates several downstream signaling pathways includ- Genotype-tissue expression; SMA: Spinal muscular atrophy; LTP: Long-term ing the PI3K/AKT/mTOR pathway [28, 29]. In neurons, potentiation; LTD: Long-term depression; SMN: Survival motor neuron. the PI3K/AKT/mTOR pathway is the central pathway for regulating local translation and synaptic strength [24, Supplementary Information 26]. The activity of FUS is shown to be modulated by The online version contains supplementary material available at https://​doi.​ mGluR1/5 [19, 68, 77]. When neurons are treated with org/​10.​1186/​s40478-​022-​01314-x. an mGluR1/5 agonist, FUS localization and expression increases locally at spines [19, 68], which does not occur Additional File 1: Supplementary tables and figures: Table S1. Exome sequencing data or DNA samples from multiple international cases with in mGluR5−/− neurons [19]. In response to glutamate FTLD-FET. Fig. S1. Validation of secondary antibody specificity for immu‑ excitotoxicity, where mGluRs are overactivated, FUS nocytochemistry studies. Fig. S2. Immunocytochemistry validation of localizes to the cytoplasm of neurons, which correlates NCDN knock-down in neurons. Fig. S3. Knock-down of FUS in N2a affects NCDN protein and mRNA levels. Fig. S4. Model for NCDN haploinsuffi‑ with a global decrease in protein synthesis [77]. Under ciency and FTD-FET. Supplementary References. Citations for Table S1. conditions where the mTOR pathway is pharmacologi- cally inhibited, FUS is shown to localize to the cytoplasm Acknowledgements where it associates with stalled polyribosomes and pro- Collaboration CEA-DRF-Jacob-CNRGH-CHU de Rouen. This work did benefit motes translation inhibition [69]. In our study we find from support of the France Génomique National infrastructure, funded as part that under conditions where NCDN is depleted in neu- of the « Investissements d’Avenir » program managed by the Agence Nation‑ ale pour la Recherche (contract ANR-10-INBS-09). We thank the neurological rons, the localization and cytoplasmic dynamics of FUS tissue bank of the Biobanc-Hospital Clinic-IDIBAPS, Barcelona, Spain for sample is affected. Moreover, our findings in cell culture are con- and data procurement. sistent with the presence of cytoplasmic FUS inclusions Authors’ contributions in a FTLD-FET patient with NCDN haploinsufficiency. ARL, GN and CFS conceived the study. All authors designed the experiments Our findings suggest that similar to mGluR1/5 activa- and interpreted the data. MS, ARL, KC, SF, ML, FL, FM, AL, RO, and AA per‑ tion, which affects FUS cytoplasmic distribution and formed the experiments. MC, AD and FLC performed the quantitative image Nicolas et al. Acta Neuropathologica Communications (2022) 10:20 Page 15 of 18 analysis CFS and MS, GN and ARL wrote the manuscript with input from all References authors. 1. Aronica E, Gorter JA (2007) Gene expression profile in temporal lobe epi‑ lepsy. Neurosci Rev J Bring Neurobiol Neurol Psychiat 13:100–108. https://​ Funding doi.​org/​10.​1177/​10738​58406​295832 This work was supported by Fonds de recherche du Québec Santé (FRQS), 2. Belly A, Moreau-Gachelin F, Sadoul R, Goldberg Y (2005) Delocalization Alzheimer Society of Canada Young Investigator Research Grant (15–29), of the multifunctional RNA splicing factor TLS/FUS in hippocampal Alzheimer’s Association New Investigator Research Grant (NIRG-14-321584), neurones: exclusion from the nucleus and accumulation in dendritic Natural Sciences and Engineering Research Council of Canada (NSERC, RGPIN- granules and spine heads. Neurosci Lett 379:152–157. https://​doi.​org/​10.​ 2020-06376 and DGECR-2020-00060), Brain Canada Future Leaders Grant to 1016/j.​neulet.​2004.​12.​071 C.F.S.; CERVO foundation and Natural Sciences and Engineering Research 3. Broce IJ, Castruita PA, Yokoyama JS (2021) Moving toward patient-tailored Council of Canada discovery grant (RGPIN-2019-06704) to F.L.C, who is a treatment in ALS and FTD: the potential of genomic assessment as a tool Canada Research Chair Tier II; Fonds de recherche du Québec Santé (FRQS) to for biological discovery and trial recruitment. Front Neurosci. https://​doi.​ P.A.D; and CNRMAJ to G.N. and D.W. org/​10.​3389/​fnins.​2021.​639078 4. Bronisz A, Carey HA, Godlewski J, Sif S, Ostrowski MC, Sharma SM (2014) Availability of data and materials The multifunctional protein fused in sarcoma (FUS) is a coactivator of The microscopy image datasets supporting the conclusions of this article microphthalmia-associated transcription factor (MITF)*. J Biol Chem are available at https://​s3.​valer​ia.​scien​ce/​ncdn-​fus/​index.​html. All codes 289:326–334. https://​doi.​org/​10.​1074/​jbc.​M113.​493874 required to perform quantitative image analysis are available at https://​github.​ 5. Buonaccorsi JP, Romeo G, Thoresen M (2018) Model-based bootstrap‑ com/​FLClab/​NCDN-​FUS. All other data generated or analysed during this ping when correcting for measurement error with application to logistic study are included in this published article. regression. Biometrics 74:135–144. https://​doi.​org/​10.​1111/​biom.​12730 6. Cacciottolo R, Ciantar J, Lanfranco M, Borg RM, Vassallo N, Bordonné R, Cauchi RJ (2019) SMN complex member Gemin3 self-interacts and has a Declarations functional relationship with ALS-linked proteins TDP-43, FUS and Sod1. Sci Rep 9:18666. https://​doi.​org/​10.​1038/​s41598-​019-​53508-4 Ethics approval and consent to participate 7. Casci I, Krishnamurthy K, Kour S, Tripathy V, Ramesh N, Anderson EN, All animal experiments were approved by the ethical committee of CPAUL, Marrone L, Grant RA, Oliver S, Gochenaur L et al (2019) Muscleblind acts Laval University, Canada (protocol 18-069-2) and comply with the current laws as a modifier of FUS toxicity by modulating stress granule dynamics of Canada. This study was approved by the Ile de France II ethics committee and SMN localization. Nat Commun 10:5583. https://​doi.​org/​10.​1038/​ (RBM-0259). s41467-​019-​13383-z 8. Chesi A, Staahl BT, Jovičić A, Couthouis J, Fasolino M, Raphael AR, Competing interests Yamazaki T, Elias L, Polak M, Kelly C et al (2013) Exome sequencing to The authors declare that they have no conflicts of interest with the contents identify de novo mutations in sporadic ALS trios. Nat Neurosci 16:851– of this article. 855. https://​doi.​org/​10.​1038/​nn.​3412 9. Collins RL, Brand H, Karczewski KJ, Zhao X, Alföldi J, Francioli LC, Khera AV, Author details Lowther C, Gauthier LD, Wang H et al (2020) A structural variation refer‑ 1 Inserm U1245 and CHU Rouen, Department of Genetics and CNR‑MAJ, ence for medical and population genetics. Nature 581:444–451. https://​ Normandie University, UNIROUEN, F‑76000 Rouen, France. 2 Department doi.​org/​10.​1038/​s41586-​020-​2287-8 of Psychiatry and Neuroscience, Laval University, Quebec City, QC, Canada, 10. Consortium GT (2013) The Genotype-Tissue Expression (GTEx) project. 2325, rue de l’Université, G1V 0A6. 3 CERVO Brain Research Centre, Laval Nat Genet 45:580–585. https://​doi.​org/​10.​1038/​ng.​2653 University, Quebec City, QC, Canada, 2601, chemin de la Canardière, G1J 11. Crozat A, Åman P, Mandahl N, Ron D (1993) Fusion of CHOP to a novel 2G3. 4 Inserm U1245 and CHU Rouen, Department of Pathology, Normandie RNA-binding protein in human myxoid liposarcoma. Nature 363:640–644. University, UNIROUEN, F‑76000 Rouen, France. 5 Institut Intelligence et Don‑ https://​doi.​org/​10.​1038/​36364​0a0 nées, Laval University, Quebec City, QC, Canada. 6 Department of Biochemistry, 12. Dateki M, Horii T, Kasuya Y, Mochizuki R, Nagao Y, Ishida J, Sugiyama F, Microbiology and Bioinformatics, Laval University, Quebec City, QC, Canada. Tanimoto K, Yagami K, Imai H et al (2005) Neurochondrin negatively 7 Lille Neuroscience and Cognition (Inserm UMRS1172) Alzheimer and Tauopa‑ regulates CaMKII phosphorylation, and nervous system-specific gene dis‑ thies, Laboratory of Excellence Distalz (Development of Innovative Strategies ruption results in epileptic seizure. J Biol Chem 280:20503–20508. https://​ for a Transdisciplinary Approach to ALZheimer’s Disease), University of Lille, doi.​org/​10.​1074/​jbc.​M4140​33200 CHU Lille, Lille, France. 8 Department of Neuropathology, University of Lille, 13. Dhar SK, Zhang J, Gal J, Xu Y, Miao L, Lynn BC, Zhu H, Kasarskis EJ, St. Clair CHU Lille, Lille, France. 9 INSERM, CNRS U1127, Institut du Cerveau, Sorbonne DK, (2013) FUsed in sarcoma is a novel regulator of manganese superox‑ Université, ICM, Paris, France. 10 Laboratoire de Neuropathologie R. Escourolle, ide dismutase gene transcription. Antioxid Redox Signal 20:1550–1566. AP-HP, Hôpital de la Pitié-Salpêtrière, Paris, France. 11 CEA, Centre National https://​doi.​org/​10.​1089/​ars.​2012.​4984 de Recherche en Génomique Humaine, Université Paris-Saclay, 91057 Evry, 14. Dichmann DS, Harland RM (2012) fus/TLS orchestrates splicing of France. 12 Memory Unit, Department of Neurology, Hospital de la Santa Creu developmental regulators during gastrulation. Genes Dev 26:1351–1363. i Sant Pau, Biomedical Research Institute Sant Pau, Universitat Autònoma de https://​doi.​org/​10.​1101/​gad.​187278.​112 Barcelona, Barcelona, Spain. 13 Centro de Investigación Biomédica en Red 15. Dormann D, Haass C (2013) Fused in sarcoma (FUS): an oncogene goes Enfermedades Neurodegenerativas (CIBERNED), Madrid, Spain. 14 Neurological awry in neurodegeneration. Molecular cellular Neurosci. https://​doi.​org/​ Tissue Bank of the Biobank-Hospital Clinic-IDIBAPS, Barcelona, Spain. 15 Division 10.​1016/j.​mcn.​2013.​03.​006 of Neuropathology, Department of Neurology, Medical University of Vienna, 16. Dormann D, Rodde R, Edbauer D, Bentmann E, Fischer I, Hruscha A, Than Vienna, Austria. 16 Dementia Research Project, Department of Brain and Neu‑ ME, Mackenzie IR, Capell A, Schmid B et al (2010) ALS-associated fused in rosciences, Tokyo Metropolitan Institute of Medical Science, Setagaya City, sarcoma (FUS) mutations disrupt Transportin-mediated nuclear import. Japan. 17 Department of Neurology and Alzheimer Center, Erasmus Medical EMBO J 29:2841–2857. https://​doi.​org/​10.​1038/​emboj.​2010.​143 Center, Rotterdam, The Netherlands. 18 Department of Internal Medicine, 17. Farina S, Esposito F, Battistoni M, Biamonti G, Francia S (2021) Post- Erasmus Medical Center, Rotterdam, The Netherlands. 19 UNIROUEN, Inserm Translational modifications modulate proteinopathies of TDP-43, FUS and U1245 and CHU Rouen, Department of Neurology and CNR‑MAJ, Normandie hnRNP-A/B in amyotrophic lateral sclerosis. Front Mol Biosci 8:585 University, F‑76000 Rouen, France. 18. Fatima A, Hoeber J, Schuster J, Koshimizu E, Maya-Gonzalez C, Keren B, Mignot C, Akram T, Ali Z, Miyatake S et al (2021) Monoallelic and bi-allelic Received: 27 September 2021 Accepted: 18 January 2022 variants in NCDN cause neurodevelopmental delay, intellectual disability, and epilepsy. Am J Hum Genet 108:739–748. https://​doi.​org/​10.​1016/j.​ ajhg.​2021.​02.​015 Nicolas et al. Acta Neuropathologica Communications (2022) 10:20 Page 16 of 18 19. Fujii R, Okabe S, Urushido T, Inoue K, Yoshimura A, Tachibana T, Nishikawa amyotrophic lateral sclerosis and diseases of the frontotemporal demen‑ T, Hicks GG, Takumi T (2005) The RNA binding protein TLS is translocated tia spectrum. JAMA Neurol 75:860–875. https://​doi.​org/​10.​1001/​jaman​ to dendritic spines by mGluR5 activation and regulates spine morphol‑ eurol.​2018.​0372 ogy. Current biology : CB 15:587–593. https://​doi.​org/​10.​1016/j.​cub.​2005.​ 38. Karczewski KJ, Francioli LC, Tiao G, Cummings BB, Alföldi J, Wang Q, Col‑ 01.​058 lins RL, Laricchia KM, Ganna A, Birnbaum DP et al (2020) The mutational 20. Fujii R, Takumi T (2005) TLS facilitates transport of mRNA encoding an constraint spectrum quantified from variation in 141,456 humans. Nature actin-stabilizing protein to dendritic spines. J Cell Sci 118:5755–5765. 581:434–443. https://​doi.​org/​10.​1038/​s41586-​020-​2308-7 https://​doi.​org/​10.​1242/​jcs.​02692 39. Kawaguchi T, Rollins MG, Moinpour M, Morera AA, Ebmeier CC, Old 21. Goldman JS, Van Deerlin VM (2018) Alzheimer’s disease and frontotem‑ WM, Schwartz JC (2020) Changes to the TDP-43 and FUS Interactomes poral dementia: the current state of genetics and genetic testing since Induced by DNA Damage. J Proteome Res 19:360–370. https://​doi.​org/​10.​ the advent of next-generation sequencing. Mol Diagn Ther 22:505–513. 1021/​acs.​jprot​eome.​9b005​75 https://​doi.​org/​10.​1007/​s40291-​018-​0347-7 40. Kawakami I, Kobayashi Z, Arai T, Yokota O, Nonaka T, Aoki N, Niizato K, 22. Greaves CV, Rohrer JD (2019) An update on genetic frontotempo‑ Oshima K, Higashi S, Katsuse O et al (2016) Chorea as a clinical feature of ral dementia. J Neurol 266:2075–2086. https://​doi.​org/​10.​1007/​ the basophilic inclusion body disease subtype of fused-in-sarcoma-asso‑ s00415-​019-​09363-4 ciated frontotemporal lobar degeneration. Acta Neuropathol Commun 23. Groen EJ, Fumoto K, Blokhuis AM, Engelen-Lee J, Zhou Y, van den Heuvel 4:36–36. https://​doi.​org/​10.​1186/​s40478-​016-​0304-9 DM, Koppers M, van Diggelen F, van Heest J, Demmers JA et al (2013) 41. Kim HJ, Taylor JP (2017) Lost in transportation: nucleocytoplasmic ALS-associated mutations in FUS disrupt the axonal distribution and transport defects in ALS and other neurodegenerative diseases. Neuron function of SMN. Hum Mol Genet 22:3690–3704. https://​doi.​org/​10.​1093/​ 96:285–297. https://​doi.​org/​10.​1016/j.​neuron.​2017.​07.​029 hmg/​ddt222 42. Kok JR, Palminha NM, Dos Santos SC, El-Khamisy SF, Ferraiuolo L (2021) 24. Henry FE, McCartney AJ, Neely R, Perez AS, Carruthers CJ, Stuenkel EL, DNA damage as a mechanism of neurodegeneration in ALS and a con‑ Inoki K, Sutton MA (2012) Retrograde changes in presynaptic function tributor to astrocyte toxicity. Cell Mol Life Sci 78:5707–5729. https://​doi.​ driven by dendritic mTORC1. J Neurosci 32:17128–17142. https://​doi.​org/​ org/​10.​1007/​s00018-​021-​03872-0 10.​1523/​JNEUR​OSCI.​2149-​12.​2012 43. Kurz A, Kurz C, Ellis K, Lautenschlager NT (2014) What is frontotemporal 25. Ho WY, Agrawal I, Tyan S-H, Sanford E, Chang W-T, Lim K, Ong J, Tan BSY, dementia? Maturitas 79:216–219. https://​doi.​org/​10.​1016/j.​matur​itas.​ Moe AAK, Yu R et al (2021) Dysfunction in nonsense-mediated decay, 2014.​07.​001 protein homeostasis, mitochondrial function, and brain connectivity in 44. Kwiatkowski TJ Jr, Bosco DA, Leclerc AL, Tamrazian E, Vanderburg CR, Russ ALS-FUS mice with cognitive deficits. Acta Neuropathol Commun 9:9–9. C, Davis A, Gilchrist J, Kasarskis EJ, Munsat T et al (2009) Mutations in the https://​doi.​org/​10.​1186/​s40478-​020-​01111-4 FUS/TLS gene on chromosome 16 cause familial amyotrophic lateral 26. Hoeffer CA, Klann E (2010) mTOR signaling: at the crossroads of plasticity, sclerosis. Science 323:1205–1208. https://​doi.​org/​10.​1126/​scien​ce.​11660​ memory and disease. Trends Neurosci 33:67–75. https://​doi.​org/​10.​1016/j.​ 66 tins.​2009.​11.​003 45. Lagier-Tourenne C, Polymenidou M, Cleveland DW (2010) TDP-43 and 27. Hoell JI, Larsson E, Runge S, Nusbaum JD, Duggimpudi S, Farazi TA, Hafner FUS/TLS: emerging roles in RNA processing and neurodegeneration. M, Borkhardt A, Sander C, Tuschl T (2011) RNA targets of wild-type and Hum Mol Genet 19:R46-64. https://​doi.​org/​10.​1093/​hmg/​ddq137 mutant FET family proteins. Nat Struct Mol Biol 18:1428–1431. https://​doi.​ 46. Lagier-Tourenne C, Polymenidou M, Hutt KR, Vu AQ, Baughn M, Huelga org/​10.​1038/​nsmb.​2163 SC, Clutario KM, Ling SC, Liang TY, Mazur C et al (2012) Divergent roles 28. Hou L, Klann E (2004) Activation of the phosphoinositide 3-kinase- of ALS-linked proteins FUS/TLS and TDP-43 intersect in processing long Akt-mammalian target of rapamycin signaling pathway is required for pre-mRNAs. Nat Neurosci 15:1488–1497. https://​doi.​org/​10.​1038/​nn.​3230 metabotropic glutamate receptor-dependent long-term depression. J 47. Lefebvre S, Bürglen L, Reboullet S, Clermont O, Burlet P, Viollet L, Benichou Neurosci 24:6352–6361. https://​doi.​org/​10.​1523/​JNEUR​OSCI.​0995-​04.​ B, Cruaud C, Millasseau P, Zeviani M et al (1995) Identification and 2004 characterization of a spinal muscular atrophy-determining gene. Cell 29. Huber KM, Klann E, Costa-Mattioli M, Zukin RS (2015) Dysregulation of 80:155–165. https://​doi.​org/​10.​1016/​0092-​8674(95)​90460-3 mammalian target of rapamycin signaling in mouse models of Autism. J 48. López-Erauskin J, Tadokoro T, Baughn MW, Myers B, McAlonis-Downes M, Neurosci 35:13836–13842. https://​doi.​org/​10.​1523/​JNEUR​OSCI.​2656-​15.​ Chillon-Marinas C, Asiaban JN, Artates J, Bui AT, Vetto AP et al (2018) ALS/ 2015 FTD-Linked mutation in FUS suppresses intra-axonal protein synthesis 30. Huber KM, Roder JC, Bear MF (2001) Chemical Induction of mGluR5- and and drives disease without nuclear loss-of-function of FUS. Neuron Protein Synthesis-Dependent Long-Term Depression in Hippocampal 100:816-830.e817. https://​doi.​org/​10.​1016/j.​neuron.​2018.​09.​044 Area CA1. J Neurophysiol 86:321–325. https://​doi.​org/​10.​1152/​jn.​2001.​ 49. Loughlin FE, Lukavsky PJ, Kazeeva T, Reber S, Hock E-M, Colombo M, Von 86.1.​321 Schroetter C, Pauli P, Cléry A, Mühlemann O et al (2019) The solution 31. Huey ED, Ferrari R, Moreno JH, Jensen C, Morris CM, Potocnik F, Kalaria RN, structure of FUS bound to RNA reveals a bipartite mode of RNA recogni‑ Tierney M, Wassermann EM, Hardy J et al (2012) FUS and TDP43 genetic tion with both sequence and shape specificity. Mol Cell 73:490-504.e496. variability in FTD and CBS. Neurobiol Aging 33:1016.e1019-1017. https://​ https://​doi.​org/​10.​1016/j.​molcel.​2018.​11.​012 doi.​org/​10.​1016/j.​neuro​biola​ging.​2011.​08.​004 50. MacDonald JR, Ziman R, Yuen RKC, Feuk L, Scherer SW (2014) The Data‑ 32. Ishigaki S, Masuda A, Fujioka Y, Iguchi Y, Katsuno M, Shibata A, Urano F, base of Genomic Variants: a curated collection of structural variation in Sobue G, Ohno K (2012) Position-dependent FUS-RNA interactions regu‑ the human genome. Nucleic Acids Res 42:D986–D992. https://​doi.​org/​10.​ late alternative splicing events and transcriptions. Sci Rep 2:529. https://​ 1093/​nar/​gkt958 doi.​org/​10.​1038/​srep0​0529 51. Mackenzie IR, Neumann M, Bigio EH, Cairns NJ, Alafuzoff I, Kril J, Kovacs 33. Ishigaki S, Sobue G (2018) Importance of Functional Loss of FUS in FTLD/ GG, Ghetti B, Halliday G, Holm IE et al (2010) Nomenclature and nosol‑ ALS. Front Mol Biosci 5:44. https://​doi.​org/​10.​3389/​fmolb.​2018.​00044 ogy for neuropathologic subtypes of frontotemporal lobar degenera‑ 34. Istvánffy R, Vogt Weisenhorn DM, Floss T, Wurst W (2004) Expression of tion: an update. Acta Neuropathol 119:1–4. https://​doi.​org/​10.​1007/​ neurochondrin in the developing and adult mouse brain. Dev Genes Evol s00401-​009-​0612-2 214:206–209. https://​doi.​org/​10.​1007/​s00427-​004-​0396-2 52. Masrori P, Van Damme P (2020) Amyotrophic lateral sclerosis: a clinical 35. Ito D, Hatano M, Suzuki N (2017) RNA binding proteins and the patho‑ review. Eur J Neurol 27:1918–1929. https://​doi.​org/​10.​1111/​ene.​14393 logical cascade in ALS/FTD neurodegeneration. Sci Translatio Med 9: 53. Matosin N, Fernandez-Enright F, Fung SJ, Lum JS, Engel M, Andrews JL, eaah5436. https://​doi.​org/​10.​1126/​scitr​anslm​ed.​aah54​36 Huang XF, Weickert CS, Newell KA (2015) Alterations of mGluR5 and its 36. Kamelgarn M, Chen J, Kuang L, Jin H, Kasarskis EJ, Zhu H (2018) ALS muta‑ endogenous regulators Norbin, Tamalin and Preso1 in schizophrenia: tions of FUS suppress protein translation and disrupt the regulation of towards a model of mGluR5 dysregulation. Acta Neuropathol 130:119– nonsense-mediated decay. Proc Natl Acad Sci USA 115:E11904-e11913. 129. https://​doi.​org/​10.​1007/​s00401-​015-​1411-6 https://​doi.​org/​10.​1073/​pnas.​18104​13115 54. Mulkey M (2019) Understanding frontotemporal disease progression and 37. Karch CM, Wen N, Fan CC, Yokoyama JS, Kouri N, Ross OA, Höglinger G, management strategies. Nurs Clin North Am 54:437–448. https://​doi.​org/​ Müller U, Ferrari R, Hardy J et al (2018) Selective genetic overlap between 10.​1016/j.​cnur.​2019.​04.​011 Nicolas et al. Acta Neuropathologica Communications (2022) 10:20 Page 17 of 18 55. Nakaya T, Alexiou P, Maragkakis M, Chang A, Mourelatos Z (2013) FUS 72. Shinozaki K, Maruyama K, Kume H, Kuzume H, Obata K (1997) A novel regulates genes coding for RNA-binding proteins in neurons by binding brain gene, norbin, induced by treatment of tetraethylammonium in rat to their highly conserved introns. RNA 19:498–509. https://​doi.​org/​10.​ hippocampal slice and accompanied with neurite-outgrowth in neuro 1261/​rna.​037804.​112 2a cells. Biochem Biophys Res Commun 240:766–771. https://​doi.​org/​10.​ 56. Nault F, De Koninck P (2010) Dissociated Hippocampal Cultures. In: 1006/​bbrc.​1997.​7660 Doering LC (ed) Protocols for Neural Cell Culture, Fourth Edition. Humana 73. Steinberg KM, Yu B, Koboldt DC, Mardis ER, Pamphlett R (2015) Exome Press, City, pp 137–159 sequencing of case-unaffected-parents trios reveals recessive and de 57. Neumann M, Bentmann E, Dormann D, Jawaid A, DeJesus-Hernandez M, novo genetic variants in sporadic ALS. Sci Rep 5:9124. https://​doi.​org/​10.​ Ansorge O, Roeber S, Kretzschmar HA, Munoz DG, Kusaka H et al (2011) 1038/​srep0​9124 FET proteins TAF15 and EWS are selective markers that distinguish FTLD 74. Svetoni F, Frisone P, Paronetto MP (2016) Role of FET proteins in neuro‑ with FUS pathology from amyotrophic lateral sclerosis with FUS muta‑ degenerative disorders. RNA Biol 13:1089–1102. https://​doi.​org/​10.​1080/​ tions. Brain J Neuro 134:2595–2609. https://​doi.​org/​10.​1093/​brain/​awr201 15476​286.​2016.​12112​25 58. Neumann M, Mackenzie IRA (2019) Review: neuropathology of non- 75. Tan AY, Riley TR, Coady T, Bussemaker HJ, Manley JL (2012) TLS/FUS tau frontotemporal lobar degeneration. Neuropathol Appl Neurobiol (translocated in liposarcoma/fused in sarcoma) regulates target gene 45:19–40. https://​doi.​org/​10.​1111/​nan.​12526 transcription via single-stranded DNA response elements. Proc Natl Acad 59. Newell K, Mackay C, Lum J, Millard S, Huang X-F, Fernandez F (2016) Sci 109:6030. https://​doi.​org/​10.​1073/​pnas.​12030​28109 PS146. Norbin: an emerging player in the pathophysiology and treatment 76. Thompson LW, Morrison KD, Shirran SL, Groen EJN, Gillingwater TH, of depression? Int J Neuropsychopharmacol 19:50–50. https://​doi.​org/​10.​ Botting CH, Sleeman JE (2018) Neurochondrin interacts with the SMN 1093/​ijnp/​pyw043.​146 protein suggesting a novel mechanism for spinal muscular atrophy 60. Nicolas G, Veltman JA (2019) The role of de novo mutations in adult-onset pathology. J Cell Sci. https://​doi.​org/​10.​1242/​jcs.​211482 neurodegenerative disorders. Acta Neuropathol 137:183–207. https://​doi.​ 77. Tischbein M, Baron DM, Lin YC, Gall KV, Landers JE, Fallini C, Bosco DA org/​10.​1007/​s00401-​018-​1939-3 (2019) The RNA-binding protein FUS/TLS undergoes calcium-mediated 61. Ohoka Y, Hirotani M, Sugimoto H, Fujioka S, Furuyama T, Inagaki S (2001) nuclear egress during excitotoxic stress and is required for GRIA2 mRNA Semaphorin 4C, a transmembrane semaphorin, [corrected] associates processing. J Biol Chem 294:10194–10210. https://​doi.​org/​10.​1074/​jbc.​ with a neurite-outgrowth-related protein, SFAP75. Biochem Biophys Res RA118.​005933 Commun 280:237–243. https://​doi.​org/​10.​1006/​bbrc.​2000.​4080 78. Tyzack GE, Luisier R, Taha DM, Neeves J, Modic M, Mitchell JS, Meyer I, 62. Pham J, Keon M, Brennan S, Saksena N (2020) Connecting RNA-Modifying Greensmith L, Newcombe J, Ule J et al (2019) Widespread FUS mislo‑ Similarities of TDP-43, FUS, and SOD1 with MicroRNA Dysregulation calization is a molecular hallmark of amyotrophic lateral sclerosis. Brain : Amidst A Renewed Network Perspective of Amyotrophic Lateral Sclerosis a journal of neurology 142:2572–2580. https://​doi.​org/​10.​1093/​brain/​ Proteinopathy. Int J Mol Sci. https://​doi.​org/​10.​3390/​ijms2​11034​64 awz217 63. Rogelj B, Easton LE, Bogu GK, Stanton LW, Rot G, Curk T, Zupan B, Sugi‑ 79. Udagawa T, Fujioka Y, Tanaka M, Honda D, Yokoi S, Riku Y, Ibi D, Nagai T, moto Y, Modic M, Haberman N et al (2012) Widespread binding of FUS Yamada K, Watanabe H et al (2015) FUS regulates AMPA receptor function along nascent RNA regulates alternative splicing in the brain. Sci Rep and FTLD/ALS-associated behaviour via GluA1 mRNA stabilization. Nat 2:603. https://​doi.​org/​10.​1038/​srep0​0603 Commun 6:7098. https://​doi.​org/​10.​1038/​ncomm​s8098 64. Sahadevan S, Hembach KM, Tantardini E, Pérez-Berlanga M, Hruska- 80. van Doormaal PTC, Ticozzi N, Weishaupt JH, Kenna K, Diekstra FP, Verde F, Plochan M, Megat S, Weber J, Schwarz P, Dupuis L, Robinson MD et al Andersen PM, Dekker AM, Tiloca C, Marroquin N et al (2017) The role of (2021) Synaptic FUS accumulation triggers early misregulation of synap‑ de novo mutations in the development of amyotrophic lateral sclerosis. tic RNAs in a mouse model of ALS. Nat Commun 12:3027–3027. https://​ Hum Mutat 38:1534–1541. https://​doi.​org/​10.​1002/​humu.​23295 doi.​org/​10.​1038/​s41467-​021-​23188-8 81. Van Langenhove T, van der Zee J, Sleegers K, Engelborghs S, Vanden‑ 65. Scekic-Zahirovic J, Sendscheid O, El Oussini H, Jambeau M, Sun Y, berghe R, Gijselinck I, Van den Broeck M, Mattheijssens M, Peeters K, De Mersmann S, Wagner M, Dieterlé S, Sinniger J, Dirrig-Grosch S et al (2016) Deyn PP et al (2010) Genetic contribution of FUS to frontotemporal lobar Toxic gain of function from mutant FUS protein is crucial to trigger cell degeneration. Neurology 74:366–371. https://​doi.​org/​10.​1212/​WNL.​ autonomous motor neuron loss. EMBO J 35:1077–1097. https://​doi.​org/​ 0b013​e3181​ccc732 10.​15252/​embj.​20159​2559 82. Vance C, Rogelj B, Hortobagyi T, De Vos KJ, Nishimura AL, Sreedharan J, 66. Seelaar H, Klijnsma KY, de Koning I, van der Lugt A, Chiu WZ, Azmani A, Hu X, Smith B, Ruddy D, Wright P et al (2009) Mutations in FUS, an RNA Rozemuller AJ, van Swieten JC (2010) Frequency of ubiquitin and FUS- processing protein, cause familial amyotrophic lateral sclerosis type 6. positive, TDP-43-negative frontotemporal lobar degeneration. J Neurol Science 323:1208–1211. https://​doi.​org/​10.​1126/​scien​ce.​11659​42 257:747–753. https://​doi.​org/​10.​1007/​s00415-​009-​5404-z 83. Wang H, Warner-Schmidt J, Varela S, Enikolopov G, Greengard P, Flajolet 67. Sephton CF, Good SK, Atkin S, Dewey CM, Mayer P III, Herz J, Yu G (2010) M (2015) Norbin ablation results in defective adult hippocampal neu‑ TDP-43 is a developmentally regulated protein essential for early embry‑ rogenesis and depressive-like behavior in mice. Proc Natl Acad Sci USA onic development. J Biol Chem 285:38740–38740. https://​doi.​org/​10.​ 112:9745–9750. https://​doi.​org/​10.​1073/​pnas.​15102​91112 1074/​jbc.​A109.​061846 84. Wang H, Westin L, Nong Y, Birnbaum S, Bendor J, Brismar H, Nestler E, 68. Sephton CF, Tang AA, Kulkarni A, West J, Brooks M, Stubblefield JJ, Liu Aperia A, Flajolet M, Greengard P (2009) Norbin is an endogenous regula‑ Y, Zhang MQ, Green CB, Huber KM et al (2014) Activity-dependent FUS tor of metabotropic glutamate receptor 5 signaling. Science 326:1554– dysregulation disrupts synaptic homeostasis. Proc Natl Acad Sci USA 1557. https://​doi.​org/​10.​1126/​scien​ce.​11784​96 111:E4769-4778. https://​doi.​org/​10.​1073/​pnas.​14061​62111 85. Wang HY, MacDonald ML, Borgmann-Winter KE, Banerjee A, Sleiman P, 69. Sevigny M, IB Julien, Venkatasubramani JP, Hui JB, Dutchak PA, Sephton Tom A, Khan A, Lee KC, Roussos P, Siegel SJ et al (2020) mGluR5 hypo‑ CF (2020) FUS contributes to mTOR-dependent inhibition of translation. function is integral to glutamatergic dysregulation in schizophrenia. Mol The Journal of biological chemistry. https://​doi.​org/​10.​1074/​jbc.​RA120.​ Psychiatry 25:750–760. https://​doi.​org/​10.​1038/​s41380-​018-​0234-y 013801 86. Weiler IJ, Greenough WT (1993) Metabotropic glutamate receptors trig‑ 70. Sharma A, Lyashchenko AK, Lu L, Nasrabady SE, Elmaleh M, Mendelsohn ger postsynaptic protein synthesis. Proc Natl Acad Sci 90:7168. https://​ M, Nemes A, Tapia JC, Mentis GZ, Shneider NA (2016) ALS-associated doi.​org/​10.​1073/​pnas.​90.​15.​7168 mutant FUS induces selective motor neuron degeneration through toxic 87. Xu Y, Li Z, Yao L, Zhang X, Gan D, Jiang M, Wang N, Chen G, Wang X (2017) gain of function. Nat Commun 7:10465. https://​doi.​org/​10.​1038/​ncomm​ Altered Norbin Expression in Patients with Epilepsy and a Rat Model. Sci s10465 Rep 7:13970. https://​doi.​org/​10.​1038/​s41598-​017-​13248-9 71. Shinozaki K, Kume H, Kuzume H, Obata K, Maruyama K (1999) Norbin, 88. Yamazaki T, Chen S, Yu Y, Yan B, Haertlein TC, Carrasco MA, Tapia JC, Zhai B, a neurite-outgrowth-related protein, is a cytosolic protein localized in Das R, Lalancette-Hebert M et al (2012) FUS-SMN protein interactions link the somatodendritic region of neurons and distributed prominently in the motor neuron diseases ALS and SMA. Cell Rep 2:799–806. https://​doi.​ dendritic outgrowth in Purkinje cells. Brain Res Mol Brain Res 71:364–368. org/​10.​1016/j.​celrep.​2012.​08.​025 https://​doi.​org/​10.​1016/​s0169-​328x(99)​00181-3 Nicolas et al. Acta Neuropathologica Communications (2022) 10:20 Page 18 of 18 89. Yang L, Gal J, Chen J, Zhu H (2014) Self-assembled FUS binds active chro‑ matin and regulates gene transcription. Proc Natl Acad Sci 111:17809. https://​doi.​org/​10.​1073/​pnas.​14140​04111 90. Yokoi S, Udagawa T, Fujioka Y, Honda D, Okado H, Watanabe H, Katsuno M, Ishigaki S, Sobue G (2017) 3’UTR length-dependent control of SynGAP isoform α2 mRNA by FUS and ELAV-like proteins promotes dendritic spine maturation and cognitive function. Cell Rep 20:3071–3084. https://​ doi.​org/​10.​1016/j.​celrep.​2017.​08.​100 91. Yousefian-Jazi A, Seol Y, Kim J, Ryu HL, Lee J, Ryu H (2020) Pathogenic genome signatures that damage motor neurons in amyotrophic lateral sclerosis. Cells. https://​doi.​org/​10.​3390/​cells​91226​87 92. Zhou Y, Liu S, Liu G, Oztürk A, Hicks GG (2013) ALS-associated FUS muta‑ tions result in compromised FUS alternative splicing and autoregulation. PLoS Genet 9:e1003895–e1003895. https://​doi.​org/​10.​1371/​journ​al.​pgen.​ 10038​95 93. Zinszner H, Sok J, Immanuel D, Yin Y, Ron D (1997) TLS (FUS) binds RNA in vivo and engages in nucleo-cytoplasmic shuttling. J Cell Sci 110(Pt 15):1741–1750 Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub‑ lished maps and institutional affiliations. Ready to submit your research ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions