Theme 04 - IN VIVO EXPERIMETAL MODELS

Background: The most common pathological hallmark of ALS is the cytoplasmic mislocalisation and aggregation of phosphorylated TDP-43, with the concurrent loss of functional nuclear TDP-43 in the motor neurons of the spinal cord and motor cortex. The cerebellum shares vast connectivity with these areas and functions in fine motor tuning yet seems spared from TDP-43 pathology, although inclusions of the ubiquitin-binding scaffold protein p62 have been identified in the cerebellar cortex of C9orf72 patients1. Previous work in our lab has demonstrated that TDP-43 protein levels are much higher in the cerebellum vs. the cortex in both control and transgenic mouse models. Furthermore, we observed a relative increase in cytoplasmic TDP-43 load in the cerebellum compared to the cortex in all mice, particularly in the large Purkinje cells. Objectives: To compare pathological hallmarks of phosphorylated TDP-43 and p62 in the cortex and the cerebellum of hTDPQ331K mice, and to compare the TDP-43 interactome using immunoprecipitation and mass spectrometry analysis of TDP-43 from both tissues in control animals. Methods: Brains were harvested from 3 and 24-month-old NTg, and hTDPQ331K animals and processed for cortex and cerebellar specific cellular fractionation for immunoprecipitation, or cryofrozen and sectioned for immunohistochemistry. Phosphorylated TDP-43 (pTDP-43) (Biolgened) and p62 (abcam) antibodies were used to analyse pathology for immunohistochemistry. TDP-43 antibodies (Proteintech/abcam) were used to pull-down TDP-43 from nuclear and cytoplasmic fractions from the cortex and cerebellum of 3 months NTg mice. Results: Immunohistochemical staining shows a generally punctate staining pattern of p62 that colocalises with pTDP in the motor cortex and cerebellum of both NTg and hTDPQ331K animals at 3 months. At 24 months these small puncta assemble into large aggregates of colocalised pTDP-43/ p62 in the motor cortex of the hTDPQ331K mice but not the cerebellum. Mass spectrometry analysis of TDP-43 pulldowns shows some differences in TDP-43 binding partners between the cortex and cerebellum. Interactions between the paraspeckle proteins NONO and SFPQ were observed in both brain regions however interactions with other paraspeckle proteins FUS and PSPC-1 only occurred within the cortex. The fibrogranular network protein Matrin-3 was also seen to interact more consistently with TDP-43 in the cortex. Several synaptic were also only detected as TDP-43 interactors in the cortex. Discussion: The cerebellum can evade the characteristic large, pathological, cytoplasmic aggregates of TDP-43 and other ALS-related proteins seen within the motor cortex, despite having a higher cytoplasmic load of the protein. It is possible that the differential TDP-43 interactome within the motor cortex contributes to this pathological aggregation, or that the cerebellum possesses a more efficient way of handling the higher concentration of cytoplasmic TDP-43.

Background: The most common pathological hallmark of ALS is the cytoplasmic mislocalisation and aggregation of phosphorylated TDP-43, with the concurrent loss of functional nuclear TDP-43 in the motor neurons of the spinal cord and motor cortex. The cerebellum shares vast connectivity with these areas and functions in fine motor tuning yet seems spared from TDP-43 pathology, although inclusions of the ubiquitin-binding scaffold protein p62 have been identified in the cerebellar cortex of C9orf72 patients1. Previous work in our lab has demonstrated that TDP-43 protein levels are much higher in the cerebellum vs. the cortex in both control and transgenic mouse models. Furthermore, we observed a relative increase in cytoplasmic TDP-43 load in the cerebellum compared to the cortex in all mice, particularly in the large Purkinje cells. Objectives: To compare pathological hallmarks of phosphorylated TDP-43 and p62 in the cortex and the cerebellum of hTDPQ331K mice, and to compare the TDP-43 interactome using immunoprecipitation and mass spectrometry analysis of TDP-43 from both tissues in control animals. Methods: Brains were harvested from 3 and 24-month-old NTg, and hTDPQ331K animals and processed for cortex and cerebellar specific cellular fractionation for immunoprecipitation, or cryofrozen and sectioned for immunohistochemistry. Phosphorylated TDP-43 (pTDP-43) (Biolgened) and p62 (abcam) antibodies were used to analyse pathology for immunohistochemistry. TDP-43 antibodies (Proteintech/abcam) were used to pull-down TDP-43 from nuclear and cytoplasmic fractions from the cortex and cerebellum of 3 months NTg mice. Results: Immunohistochemical staining shows a generally punctate staining pattern of p62 that colocalises with pTDP in the motor cortex and cerebellum of both NTg and hTDPQ331K animals at 3 months. At 24 months these small puncta assemble into large aggregates of colocalised pTDP-43/ p62 in the motor cortex of the hTDPQ331K mice but not the cerebellum. Mass spectrometry analysis of TDP-43 pulldowns shows some differences in TDP-43 binding partners between the cortex and cerebellum. Interactions between the paraspeckle proteins NONO and SFPQ were observed in both brain regions however interactions with other paraspeckle proteins FUS and PSPC-1 only occurred within the cortex. The fibrogranular network protein Matrin-3 was also seen to interact more consistently with TDP-43 in the cortex. Several synaptic were also only detected as TDP-43 interactors in the cortex. Discussion: The cerebellum can evade the characteristic large, pathological, cytoplasmic aggregates of TDP-43 and other ALS-related proteins seen within the motor cortex, despite having a higher cytoplasmic load of the protein. It is possible that the differential TDP-43 interactome within the motor cortex contributes to this pathological aggregation, or that the cerebellum possesses a more efficient way of handling the higher concentration of cytoplasmic TDP-43.
tilly.hawkins@kcl.ac.uk supplemented with either GTE or cocoa flavanols. Histological, immunohistochemical, and western blot procedures, along with motor behavioral and electrophysiological tests were carried out.
Results: Compared to control mice, both GTE-and cocoa-supplementation significantly improved the survival rate of mice (p < 0.05, GTE and p < 0.01, cocoa), and elicited a significant reduction in the proportion of fibers with lipofuscin aggregates (p < 0.001) and central nuclei (p < 0.001), and a conspicuous increase in the density of satellite cells (p < 0.001) and the expression of PGC-1a (p < 0.05) in skeletal muscles. Additionally, both supplements significantly increased the number of innervated NMJs (p < 0.001 in tibialis anterior (TA) and soleus (Sol) muscles) and augmented their degree of maturity (p < 0.001, GTE and p < 0.01, cocoa) compared to controls. GTE, but not cocoa, prominently increased the density of VAChT (p < 0.001) and VGluT2 (p < 0.001) synaptic afferents on MNs, which were lost in control aged spinal cords; conversely, cocoa, but not GTE, significantly augmented the proportion of VGluT1 afferents on MNs (p < 0.001). Moreover, GTE but not cocoa reduced aging-associated microgliosis (p < 0.05) and increased the proportion of neuroprotective microglial phenotypes (p < 0.001). No overt improvements were observed in agingrelated motor activity decline and electrophysiological alterations, possibly reflecting the need for longer-term interventions. Discussion: Our data indicate that certain plant flavonoids may be beneficial in the nutritional management of agerelated deterioration of the neuromuscular system towards addressing sarcopenia progression. These results open the door to a nutraceutical therapy for aging-related neuromuscular alterations accompanying sarcopenic muscle and should be also considered for the management of patients affected by muscular and MN diseases. Amyotrophic lateral sclerosis (ALS) is a sporadic or genetic disease associated with peripheral loss of synaptic connectivity at neuromuscular junctions and central loss of motor neurons. The SOD1G93A mouse model has been widely used to study a familial form of ALS. C1q, the initiating molecule of the classical complement cascade, marks synapses in the central nervous system for glial elimination during normal development while triggering aberrant synapse loss in neurodegenerative disorders. In ALS, C1q additionally tags the neuromuscular junction within the peripheral nervous system before its removal by macrophages. Despite data showing that embryonic knock out of C1q was not protective in a mutant SOD1 model, we hypothesized that excessive synaptic pruning initiated by C1q contributes to motor deficits in ALS and that pharmacologically inhibiting this process would be beneficial. We treated adult SOD1G93A mice with an anti-C1q blocking antibody from 7 to 16 weeks of age. Treatment resulted in the reduction of C1q in plasma, spinal cord, and muscle tissue, along with inhibition of downstream classical complement activation after 9 weeks of treatment. Treated mice showed significant improvement in the amplitude and reduced latency of compound muscle action potential, demonstrating increased synaptic connectivity at the neuromuscular junction upon C1q inhibition. Furthermore, anti-C1q treatment reduced Nf-L levels in CSF and plasma of SOD1G93A mice, marking the reduction in neuronal damage in this model.  (2). Reduced level of C9orf72 protein was associated with an elevated level of AMPAR glutamate receptor 1 (GluR1) in IPSC with C9orf72 mutation (3), C9orf72 knockout (C9-KO) iMN (4) and mice (5), and post-mortem motor cortex of C9-ALS patients (4). However, C9orf72 haploinsufficiency, by itself, is not sufficient to cause ALS phenotypes in mice (6). Since excitotoxicity is a major pathogenic mechanism in ALS (7), we hypothesize that elevation in GluR1 makes C9-KO mice susceptible to excitotoxicity.
Objective: The objective of the project is to test the susceptibility of C9orf72 KO mice to excitotoxicity. Methods: To induce excitotoxicity, a cohort of mice was injected with 20 mg/kg kainic acid (KA) i.p. and scored on the Racine scale of 6 stages. Another cohort of mice was EEGrecorded before and after KA treatment.
Results: C9-KO mice exhibited more seizure manifestations when the average Racine stage, number, and duration of seizure at stages 5 and 4 were quantified. Recordings from the hippocampal CA3 region show baseline hyperexcitability and enhanced susceptibility to KA. Discussion: Overall, C9-KO mice are vulnerable to kainic acid injection on behavioral and electroencephalographic assessments. A previous study shows that C9-KO rats treated with a small dose of kainic acid claimed not to cause seizure, developed motor deficits, and showed loss of motor neurons (MNs) (8). In the current study, we used a higher dose to induce excitotoxicity in the hippocampus, because over 50% of ALS patients also develop cognitive impairment caused by degeneration of the frontal and temporal lobes (9). In line with this, the elevation in GluR1 in C9-KO mice was localized to the CA3 of the dorsal hippocampus in mice. Our data provide functional consequences of elevated GluR1, supporting the hypothesis that loss of C9orf72 increases susceptibility to excitotoxicity.
belay.gebregergis@mail.utoronto.ca IVV-08 Tongue denervation atrophy and dysphagia penetrance, but not overall survival, are affected by limb phenotype in a mouse model of ALS and sired equal proportions of mixed vs. hindlimb phenotype offspring. Both LCN-SOD1 phenotypes had significant dysphagia compared to WT mice (p < 0.05): slower lick and swallow rates, longer inter-lick and inter-swallow intervals, and longer pharyngeal transit times. Dysphagia penetrance differed between phenotypes (mixed ¼100%; hindlimb ¼64%), yet survival times were similar. LCN-SOD1 mice, regardless of phenotype, had significantly smaller genioglossus myofibers and more centralized myonuclei compared to WT mice (p < 0.05). These biomarkers of denervation atrophy were significantly correlated with VFSS metrics (lick and swallow rates, p < 0.05), but not survival time.
Discussion: Limb phenotype heritance was only chance level for LCN-SOD1 mice, whereas dysphagia penetrance was complete for mixed (100%) vs. incomplete for hindlimb (64%) phenotype mice. Remarkably, both phenotypes had tongue denervation atrophy, even hindlimb phenotype mice without dysphagia. This finding recapitulates human ALS, providing the robust rationale for using this preclinical model to explore targeted treatments for tongue denervation atrophy and ensuing dysphagia. Unlike human ALS, LCN-SOD1 mice with dysphagia and tongue atrophy did not have shorter survival times, likely due to our humane end-point criterion.
Nonetheless, this mouse model recapitulates key aspects of human ALS and therefore provides a translational platform to accelerate the discovery of targeted treatments for tongue denervation atrophy and dysphagia.
rrthf2@umsystem.edu IVV-09 Splicing factor proline and glutamine rich factor-an opportunity to investigate mechanisms of motor neuron disease from a novel perspective Background: Splicing factor proline and glutamine rich (SFPQ) pathology has recently emerged as a common feature of familial and genetic MND (1-4). Reported pathology includes nuclear depletion of SFPQ in multiple animal models of MND (1-3), increased SFPQ intron retention (1,2), reduced expression at the RNA level (1), and significantly, aggregation of SFPQ within spinal cord motor neurons (1). SFPQ is a DNA-RNA binding protein with roles in multiple cellular pathways implicated in the progression of MND including RNA transcription, processing and transport, DNA repair, and the cellular stress response. We hypothesise that SFPQ dysregulation presents an opportunity to examine dysfunction in these pathways from a novel perspective.
Methods: To confirm that SFPQ pathology contributes to the progression of MND (as opposed to developing as a consequence), the effects of reported pathologies are being assessed in cell and animal (zebrafish) models. Following phenotypic assessment, dysfunction associated with each pathology will be assessed at the proteome-wide level through mass spectrometry. We are also performing a targeted assessment of the effects of SFPQ dysregulation on motor neuron RNA transport. Induced failure of SFPQ-dependant RNA transport has been associated with fragmentation of sensory neuron axons (5) and we hypothesise that similar processes contribute to neuronal death in MND. To investigate this hypothesis, we are using a suite of transgenic zebrafish to perform the first characterisation of SFPQ-dependant RNA transport in motor axons under healthy and disease-relevant conditions. The optical transparency of the zebrafish makes them ideal models to study these dynamic processes in vivo in real-time.
Results: Preliminary evidence suggests that SFPQ pathology actively contributes to the progression of MND.
Overexpression of an intron retaining SFPQ constructs in cell lines produced an SFPQ protein isoform with a marked cytoplasmic shift (p < 0.0001), a high propensity to form cytoplasmic puncta, and a significant increase in cell death. Analysis of additional pathologies and characterisation of SFPQdependant RNA granules is in progress.

Conclusions:
We have recently shown that dysregulation of SFPQ is a feature of MND patient central nervous system tissue. Evidence to date indicates that this pathology contributes to disease progression and therefore, SFPQ offers a means of investigating MND-affected pathways from a novel perspective.
alison.hogan@mq.edu.au Biochemical analysis of these protein inclusions or aggregates has revealed the presence of full-length proteins and smaller protein fragments, such as ataxin-3 in MJD patients and TDP-43 in MND patients. Full-length proteins can be cleaved into smaller protein fragments by protease enzymes, and potentially contribute to disease pathogenesis. Calpain proteases are activated by changes in calcium concentrations and cleave many essential proteins, including TDP-43, ataxin-3, a-spectrin, caspase 3, and autophagy-related proteins (beclin-1 and p62). Cleavage of these proteins is known to play a role in protein activation but is also hypothesised to sometimes alter the protein function and drive neurotoxicity, leading to protein aggregate formation and neurodegeneration. We hypothesise that overactivity of calpain proteases is a disease mechanism that occurs early within the pathogenesis of neurodegenerative diseases, such as MJD and MND, leading to later neuropathology. In this study, we aimed to determine if overactivity of calpain proteases was a disease mechanism present in; CMVMJD135 transgenic mice, modelling MJD, and inducible NEFH-tTA/tetO-hTDP-43DNLS (NLS-TDP-43) transgenic mice, modelling MND. We sought to identify if a particular calpain cleavage substrate, a-spectrin, could be detected in the plasma of mice, providing an in vivo indicator of calpain activity during disease progression. Further, we aimed to examine whether the transgenic mouse models exhibited altered levels of the cleavage product. We found that the MJD mice had similar concentrations of a-spectrin breakdown products (SBDP) to wild-type littermates at 5 weeks of age. In contrast, MJD animals were found to display significantly increased levels of SBDPs when compared to wildtype littermates at 12 weeks of age, correlating with the onset of motor and neurological symptoms. Further, SBDPs were also found to be significantly increased in homogenised cerebella obtained from 15-week MJD mice compared with age-matched littermate controls. Findings from this study, relating to the level of SBDPs present in the plasma and brain lysates of transgenic mouse models of neurodegenerative diseases hypothesised to have calpain overactivity as a disease mechanism, will be informative for the investigation of mechanisms of these diseases, discovery of disease biomarkers and testing of potential therapeutics. Background: Aggregation and cytoplasmic mislocalization of TDP-43 are pathological hallmarks of amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD) spectrum. However, the molecular mechanism by which TDP-43 aggregates form and cause neurodegeneration remains poorly understood. Cyclophilin A, also known as peptidyl-prolyl cistrans isomerase A (PPIA), is a foldase and molecular chaperone. We previously found that PPIA interacts with TDP-43 and governs some of its functions, its deficiency accelerated disease progression in the SOD1 G93A mouse model of ALS (1), and a low level of PPIA correlated with a worse disease phenotype in ALS patients, suggesting a prognostic potential (2). Objectives: To investigate the role of PPIA in the cellular processes leading to TDP-43 pathology and its impact on ALS and ALS-FTD pathogenesis.
Methods: PPIA knock-out mice were characterized throughout their entire lifespan by MRI analysis, histology, electrophysiology, cognitive and motor tests, evaluation of neuroinflammation, and TDP-43 pathology. PPIA and TDP-43 protein and mRNA levels were measured in several ALS and ALS-FTD patients (n ¼ 151) and healthy controls (n ¼ 128). ALS patients were screened for coding, non-synonymous and loss-of-function SNPs in the PPIA gene. Wild-type and mutant PPIA structures were studied by molecular dynamics simulations in silico and biochemical assays in vitro.
Results: PPIA knock-out mice develop a neurodegenerative disease with key behavioural features of FTD, marked TDP-43 pathology, characterized by cytoplasmic mislocalization, abundant accumulation of C-terminal fragments, and hyperphosphorylation, and late-onset motor dysfunction. In the mouse brain, deficient PPIA induces mislocalization and aggregation of the GTP-binding nuclear protein Ran, a PPIA interactor and a master regulator of nucleocytoplasmic transport, also for TDP-43. Moreover, in absence of PPIA, TDP-43 autoregulation is perturbed and TDP-43 and proteins involved in synaptic function, synaptophysin, and PSD95, are downregulated, leading to impairment of synaptic plasticity. Finally, we found that PPIA was downregulated in several ALS and ALS-FTD patients and identified a PPIA loss-of-function mutation, PPIA K76E, in a sporadic ALS patient. The mutant PPIA has low stability, altered structure, and an aberrant interaction with TDP-43 and Ran.  (1). Paraptotic-like MNs display an accumulation of Flotilin and CD81 particles which are considered as markers for extracellular vesicles. These degenerating MNs display pMLKL positivity, the effector protein of the necroptotic pathway. Paraptotic MNs showed a significant loss of afferent synapsis and recruit microglial cells. We have established the type 2 phenotype when mfSOD1 accumulation and vacuolization was restricted to MN processes, and type 1 when mfSOD1 accumulation was not evident neither in MN somas or processes. The most severe type 3 phenotype appears to be more frequent between P60 and P90 than in the terminal stages. We have also found a progressive microgliosis displaying neurotoxic markers of activation. The relative amount of mfSOD1 phenotypes can be experimentally altered by modulating the microglial neuroinflammatory response.
Discussion: This study reveals a dissociation between mfSOD1 accumulation in degenerating MNs and the clinical progression of the disease. We show that at pre-symptomatic stages (p60) a substantial amount of MNs with phenotype type 3 (paraptotic-like) exist which can be determinant for the ulterior spread of the disease by an exosome-based mechanism. There is also a link between mfSOD1 expression in MNs and neuroinflammation. Recently, we have noticed that Y172-immunostaining is also present in the cytoplasm of SCs.

Objectives: To obtain new insights into the role of Y172related antigen in MNs and SCs.
Methods: An immunocytochemical characterization of the Y172-related-protein was performed on the spinal cord and sciatic nerve sections from CD1 mice in basal conditions and following sciatic nerve transection, and from mouse models of ALS (SOD1G93A) and SMA (Smn2B/À). All procedures were approved by the Committee for Animal Care and Use of the University of Lleida.
Results: In injured MNs after peripheral nerve transection, and in MNs from ALS and SMA mice at advanced stages of the disease, the Y172-immunostaining associated to C-boutons showed a significant depletion ($65% p < 0.001 vs. contralateral (nerve axotomy); $50% p < 0.001 vs. WT (SOD1G93A); $75% p < 0.01 vs. WT (Smn2B/À)). In the adult sciatic nerve, Y172-positive vesicle-like structures were found in the SC cytoplasm (ie perinuclear region and Cajal bands) closely related to proteins of the secretory pathway. Developing sciatic nerve exhibited a gradual increase of the cytoplasmic Y172-immunoreactivity with age. The Y172immunolabeling in SCs was also affected by the phenotypic changes that these cells exhibit upon nerve injury: the Y172vesicles were decreased in number but exhibited larger size and were particularly located at the periphery of the cytoplasm of repair SCs. Discussion: We show a novel unidentified molecular component of the C-bouton organization, which expression is lost in damaged MNs even before the occurrence of cholinergic deafferentation. Moreover, the presence of Y172-related protein in SCs suggests that this protein may play a role in axon-glial communication and MN maintenance. Our results lay the foundation for further studies aimed at identifying the Y172-related protein and determining its role in the context of the development, maintenance, plasticity, and pathology of the neuromuscular system. Objectives: We decided to study the potential involvement of these lncRNAs in ALS pathology.
Results: The lncRNAs expression profile was deregulated in all analyzed areas. Linc-p21 presented an alteration in its expression levels in all analyzed areas at 18 Weeks. We also found a p53 upregulation in symptomatic SOD1-G93A mice, indicating deregulation in the p53/linc-p21/p21 pathway. Tug1, which was found upregulated in the lumbar spinal cord at both 8 and 18 weeks of age. In relation to this, we found an increased expression of its downstream targets Tril/Tlr4, indicating an ongoing inflammatory phenomenon. For linc-Brn1a and linc-Brn1b, we found predominant deregulation in the pre-symptomatic mice, suggesting they could be relevant in the early phases of CNS development of G93A disease model. A reduced, although present, deregulation was found for Hottip, Eldrr, Linc-Enc1, and Fendrr. In SH-SY5Y-SOD1G93A, the 6 human homologues resulted deregulated vs. the wild-type cell line.  . Immunohistochemistry has been performed for specific markers of oligodendrocytes (NG2, GPR17, MBP, and GSTpi), neurons (SMI-32 and Hb9), microglia (Iba-1, CD16/32, and YM1), and astrocytes (GFAP). Results: The low MTK dose neither increased survival probability nor ameliorated disease progression in MTK-treated and separately analyzed SOD1G93A males and females vs. distinct gender vehicle-treated mice. Accordingly, immunohistochemical analyses in the lumbar spinal cord did not highlight MTK-induced modification of OL differentiation, myelin integrity, or neuroinflammatory parameters. On the contrary, the high MTK dose positively affected survival probability and strongly delayed weight loss in SOD1G93A female mice with respect to vehicle-treated mice. A similar amelioration was registered in motor behavioral tests. Moreover, immunohistochemical analysis showed that the high MTK dose positively affected specific cellular markers. Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disorder in which premature loss of upper and lower motor neurons leads to fatal paralysis. A landmark contribution to understanding cellular mechanisms of ALS came from the discovery of cytoplasmic accumulation and nuclear loss of the RNA binding protein TDP-43 from affected neurons in most instances of ALS. TDP-43 disruption represents a common pathological hallmark in both sporadic and familial forms of ALS, as well as in a growing number of other neurodegenerative disorders including frontotemporal dementia and Alzheimer's disease. Remarkably, the loss of neuron-specific tubulin-associated protein stathmin-2 has recently been identified as a major alteration linked to TDP-43 loss of nuclear function. The relevant role of stathmin-2 in the development of the nervous system and the axonal regeneration after injury is widely recognized, however, whether it places any role in an adult nervous system is currently unknown. Here, we unravel the crucial role of stathmin-2 in motor neuron axon maintenance using different approaches to deplete murine stathmin-2 in an otherwise normal adult mouse nervous system. Sustained reduction of stathmin-2 in spinal cords of adult wild-type mice by intraventricular delivery of antisense oligonucleotides, or subpial delivery of AAV9 encoding an shRNA against murine stathmin-2-focally in the lumbar spinal cord-resulted in progressive hindlimb motor deficits, reduction on nerve conduction velocity, and muscle denervation by disruption of neuromuscular junctions. Our results show that loss of stathmin-2 alone is sufficient to trigger premature axonal degeneration of motor neurons without provoking neuronal death. We also define the importance of stathmin-2 in the sensory system. The reduction of stathmin-2 expression in neurons on the dorsal root ganglion resulted in mechanical stimuli-mediated touch and pain response impairment. Furthermore, loss of stathmin-2 diminished the markers on sensory neuron terminals innervating the dorsal spinal cord. Collectively, our work uncovers the importance of stathmin-2 in the adult motor and sensory nervous systems to preserve functional innervated axon terminals. Most importantly reinforces the restoration of stathmin-2 as a critical therapeutic approach for TDP-43-dependent neurodegenerative diseases. Background: Matrin 3 is a nuclear matrix protein that has many roles in RNA processing including splicing and transport of mRNA. Missense mutations in the Matrin 3 gene (MATR3) have been linked to familial forms of amyotrophic lateral sclerosis (ALS) and distal myopathy with vocal cord and pharyngeal weakness (VCPDM) (1). However, the exact role of MATR3 mutations in ALS and myopathy pathogenesis is not understood. Protein interactome studies have demonstrated that Matrin 3 interacts with numerous proteins involved in RNA processing. Our lab has previously demonstrated that ALS-linked mutations alter MATR3 protein interactions and also decrease the nuclear export of mRNA. It is important to establish new models to understand the role of Matrin 3 and MATR3 mutations in neuromuscular disease pathogenesis. We hypothesize that MATR3 mutations cause RNA processing deficits which contribute to motor neuron (ALS) and muscle (myopathy) centric disease. Objectives: To elucidate the role of S85C and P154S MATR3 mutations in ALS and myopathy pathogenesis. Methods: We generated novel knock-in mouse models that express either S85C (ALS/VCPDM associated) or P154S (ALS associated) MATR3 mutations. Importantly, these new knockin mouse models (generated in collaboration with The Jackson Laboratory using CRISPR-Cas technology) avoid overexpression and maintain physiologic levels of mutant Matrin 3. To determine longitudinal phenotypic effects induced by MATR3 mutations, motor, and cognitive functions were tested using the rotarod, open field, and novel object recognition assays. CNS and muscle tissue were collected at early and late time points to determine the effects of MATR3 mutations on neuromuscular pathology and changes in gene expression. Results: Behavioral analysis demonstrated that homozygous S85C knock-in caused significant motor impairment detectable as early as 3 months of age. Conversely, the P154S mutation did not produce significant motor impairment. However, both mutations produced significant neuroinflammation at 8 months of age. Further analysis of muscle tissue revealed significantly reduced myofiber size in both mutant mice compared to control mice. Discussion: This study presents two separate models that maintain physiologically relevant levels of Matrin 3 while producing a strong motor phenotype in the S85C mutant mice. Interestingly, both models demonstrate hallmarks of neurodegenerative phenotypes in mutant mice, such as neuroinflammation, and our analysis demonstrates that muscle tissues are also affected. We propose that identifying the different molecular changes in both models could lead to a deeper understanding of the role of Matrin 3, and its mutations, in pathogenesis. Further, these new two models represent potentially valuable tools to neuromuscular disease field. Background: Impaired tongue function due to progressive degeneration of hypoglossal nucleus (HGN) neurons in amyotrophic lateral sclerosis (ALS) commonly gives rise to dysphagia and increased risk of malnutrition, aspiration pneumonia, and feeding tube dependence. Importantly, there are currently no effective treatments to prevent or slow the progression of HGN degeneration. Here, we used a translational mouse model of ALS to develop a targeted excitatory optogenetic stimulation approach to preserve HGN neurons and prevent corresponding tongue atrophy and dysphagia. Methods and results: To facilitate targeted opsin expression in the HGN, we stereotaxically injected the adeno-associated viral vector AAV5-hSyn-ChR2-eYFP (ChR2) into the HGN (4 WT, 4 SOD1-G93A mice) and implanted an optrode into the HGN 3 weeks post-injection. Opsin expression was confirmed in all mice by evoked HGN activity and tongue movement in response to photostimulation of the HGN (450-nm laser pulses at 40 Hz and 50% duty cycle). Tongue movements were also triggered by surface illumination of the tongue (6/8 mice), meaning the opsin spread to the distal hypoglossal nerve. Post-mortem histological examination confirmed strong opsin expression in the HGN and hypoglossal nerve, thus demonstrating that they can be reliably targeted for functional optogenetic stimulation. Videofluoroscopic swallow studies (VFSS) revealed our surgical procedures had no significant effect on licking and swallowing function or survival (p < 0.05) for WT or SOD1 mice compared to archived data from the same colony. To evaluate the effects of HGN photostimulation on behavior and neural activity, we examined tongue function in real-time by measuring lick rate and monitoring local field potentials (LFPs) in the HGN during voluntary drinking. We found that lick rate can be modulated in a graded fashion by tuning the intensity of HGN photostimulation (n ¼ 4/4). Our analysis of neural data showed LFPs in the HGN had strong lickingentrained signals, which can be used as a real-time indicator of lick rate. We hypothesize that optogenetic stimulation alters the lick rate by modulating hypoglossal drive to the tongue and causing muscle stiffness, where higher laser intensities generate more resistive force against tongue movement to cause slower licking. Conclusion: The HGN can be safely and reliably targeted for optogenetic stimulation in a mouse model of ALS. Optogenetic responses in the HGN were strong and could evoke tongue movements. Therefore, optogenetic stimulation of the HGN holds promise as a targeted tongue resistance exercise program where the resistive force can be tuned by adjusting the laser intensity during drinking. Given that optogenetic stimulation has been shown to promote axonal growth and proliferation in previous studies, we hypothesize that controlled optogenetic stimulation of the HGN may significantly delay neuronal death and preserve licking/swallowing function, which if translatable, could prolong the survival of human ALS patients and enhance the quality of life. Background and objective: One challenge in ALS research and drug discovery is the lack of fast and reliable in vivo models. Because aging is the main risk factor for ALS and many other neurodegenerative diseases, we evaluated the suitability of aged mice as a disease model for ALS, taking advantage of the commercial availability of C57BL/6J mice at different ages up to 90 weeks from the Jackson Laboratory (JAX). In readouts tested for this model, we focused on biofluid biomarkers with the potential to be translated to human patients. Methods: All mice in this study were directly purchased from JAX. There are 3 age groups: 8 weeks (young), 52 weeks (middle age), and 90 weeks (old). Each group consisted of 8 male and 8 female C57BL/6J mice with a group size N of 16. The following potential biomarkers in specific sample types have been measured: neurofilament light (NFL) in cerebrospinal fluid (CSF); Interleukin 12 p40 (IL-12p40) in plasma; Di-22:6-bis(monoacylglycerol)phosphate (BMP) in urine; Sphingomyelins and ceramides in plasma; Acid sphingomyelinase (ASM) activity in striatum; Transcriptomic analysis in rostral cortex and lumbar spinal cord. Results: For statistical analysis 1-way or 2-way ANOVA was used with the Dunnett's test and the 8 week age group as the control. We have observed a significant increase of NFL and IL-12p40 in the 90 weeks group, an increase of BMP in the 52 weeks group, and an increase of ASM activity in both 52 and 90 weeks groups. 7 sphingomyelin and 4 ceramide species showed a significant decrease in 52 and 90 weeks groups. Interestingly transcriptomic analysis and KEGG pathway enrichment identified pathways of neurodegeneration and several neurodegenerative diseases including ALS as top dysregulated pathways over age in both cortex and spinal cord. Discussion: We evaluated transcriptomic changes and several potential biomarkers in aged mice. Overall KEGG enrichment indicates similar pathway dysregulation in aged mice as in neurodegeneration. We demonstrated the increase of potential ALS disease biomarker NFL (1) and IL-12p40 (2) in aged mouse biofluids. The decrease of some sphingomyelins and ceramides is consistent with early-stage sphingolipid changes in a genetic ALS mouse model with SOD1 mutation (3). The changes in BMP (4) and ASM activity (5) indicate defects in the lysosomal pathway. Our work supports the use of aged mice as a quick model to test biomarkers for disease progression and lysosomal pathway engagement in ALS drug discovery. Background: TDP-43 can undergo liquid-liquid phase-separation (LLPS) in vitro, a process that is considered to play a role in the transformation of cytoplasmic condensates into insoluble aggregates. Post-translational modifications (PTMs) further influence the proteins' native structure, localisation, and overall function. In vitro studies show that lysine acetylation impairs TDP-43 RNA-binding ability and thus promotes the accumulation of insoluble TDP-43 species upon cellular stress (1). Whether acetylation changes TDP-43 phase-separation and aggregation propensities in vivo is yet to be explored. Objectives: To develop an in vivo model that allows us to monitor the dynamic changes of TDP-43 over time. Specifically, we aimed to (i) characterise quantitatively and qualitatively hTDP-43 compartmentalisation and phase-separation propensities in vivo, and (ii) determine the effects of RNA-binding inhibition and cellular stress on TDP-43 localisation, phase-separation, and cytoplasmic aggregation in motor neurons (MNs) in a living animal. Methods: We overexpressed human wildtype TDP-43 and RNA-binding deficient mutants specifically in motor neurons in the zebrafish spinal cord. We applied high-resolution dynamic imaging at a sub-cellular level as well as fluorescence recovery after photobleaching (FRAP) experiments to determine the molecular characteristics of condensates, TDP-43 compartmentalisation, and aggregation. Results: We established a comprehensive zebrafish model of TDP-43 pathology that allows the characterisation of nuclear and cytoplasmic TDP-43 in real-time. Confocal microscopy revealed that TDP-43 undergoes phase separation into biomolecular condensates in vivo. RNA-binding deficiency affected the localisation of TDP-43 within spinal MNs. We further observed aberrant phase separation in TDP-43 variants that are implicated in human pathology. Acetylated and therefore RNA-binding deficient TDP-43 revealed changes in LLPS characteristics, such as droplet size, dynamics, and fluorescence recovery. Discussion: Our in vivo approach demonstrates that hTDP-43 undergoes phase-separation in MNs in a zebrafish model and provides critical insights into an important molecular pathway potentially underlying aggregation formation in ALS. Understanding the underlying biophysical principles and specific properties of biological condensates has important implications for a range of biological processes. For TDP-43 specifically, it may hold the key to understanding how mutations or mislocalisation of TDP-43 can influence aggregation and trigger neurodegeneration. In summary, we established a crucial in vivo platform for testing therapeutic strategies to modulate TDP-43 pathology in the future. natalie.scherer@hdr.mq.edu.au Background: DNA damage is increasingly implicated in neurodegeneration in ALS, and we and others have recently shown that TDP-43 normally performs a role in DNA repair (1). However, this function is impaired by ALS-associated mutant forms of TDP-43. Previous studies in our group have also established that protein disulphide isomerase (PDI) is protective against dysfunction to proteostasis induced by multiple mutant proteins in vitro in ALS (2,3). However, it remains unclear if PDI is also protective against other cellular mechanisms associated with ALS, including DNA damage. Objectives: In this study, we aimed to examine whether PDI is protective against DNA damage both in vivo and in vitro, thus further defining the protective properties of PDI in ALS.
Methods: Two cell lines, motor neuronal-like NSC-34 and Neuro-2a, were transfected with PDI tagged with V5 (PDI WT-V5) or PDI with ALS-associated TDP-43 mutants (TDP-43M337V). At 48-h post-transfection the cells were treated with 13.5 uM etoposide, a topoisomerase inhibitor, for 30 min to induce DNA damage. Then, immunocytochemistry was performed using an anti-V5 antibody to detect PDI, and antibodies against DNA damage markers phosphorylated histone H2AX (cH2AX) and p53 binding protein (53BP1), which form characteristic foci following DNA damage. For in vivo studies, zebrafish embryos were microinjected with either RNA encoding PDI WT or a PDI mutant lacking redox activity. At 24 h post-fertilization, embryos were incubated with 3 mM H 2 O 2 for 24 h, then lysed to perform western blotting analysis. Results: PDI was protective against DNA damage in Neuro-2a and NSC-34 cells; significantly fewer 53 BP1 and cH2AX foci were formed in cells expressing PDI compared to controls following etoposide treatment. Moreover, PDI was also protective in cells expressing mutant TDP-43M337V. In addition, in zebrafish, the expression of PDI mRNA led to less DNA damage (assessed by cH2AX expression) induced by H 2 O 2 . PDI-WT also rescued both motor impairment and axonopathy in zebrafish expressing mutant superoxide dismutase (SOD1) A4V, unlike the PDI mutant lacking redox activity. Hence, these data reveal that PDI is protective against ALSlike phenotypes and DNA damage in vivo. Discussion: These results demonstrate the protective role of PDI against DNA damage both in vivo and in vitro. This study, therefore, has implications for future therapeutic studies based on preventing the induction of DNA damage in ALS.
sina.shadfar@mq.edu.au IVV-24 The unexplored effects of SUMOylation on TDP43 phase separation, and the subsequent impact on its localization and aggregation Background: TDP-43 localizes predominantly in the nucleus, arranging itself in highly dynamic condensates through a process called liquid-liquid phase separation (LLPS). Recently, the concept has emerged that LLPS dysregulation could be involved in the formation of pathological aggregates. However, it has been a challenge to study this concept in vivo. While it is well-evidenced that post-translational modifications (PTMs) can affect TDP-43 localization, very little is known whether PTMs can influence phase separation. We recently reported the in vitro importance of the putative SUMOylation site (Lysine136) for TDP-43 localization (1). Objective: Here, we aimed to assess the influence of SUMOylation on TDP-43 compartmentalisation, aggregation, and phase separation in vivo. Methods: Using our unique zebrafish platform and confocal microscopy, we assessed the localisation and aggregation properties of eGFP-tagged human TDP-43 (wild type, G294V and their non-SUMOylated variants with the K136R mutation) in spinal motoneurons (MNs) of 3-5 days old fish. Using FRAP techniques and 3D rendering, we tracked the recovery of fluorescence within condensates, fission/fusion events, dynamics of these condensates, and how mutations, PTMs, and stress impacted these processes. Results: Our preliminary in vivo results confirm and extend our previous in vitro studies. Lysine 136 had a strong effect on TDP-43 cellular compartmentalization in our zebrafish spinal MNs. The SUMO impaired variant (K136R) presented with a higher percentage of nuclear localisation compared to their non-mutated counterparts (WT-and G294V-TDP-43). MNs expressing SUMO impaired TDP-43 also displayed less cytoplasmic aggregates. Moreover, the SUMOylation pathway impacted TDP-43 aggregation and dynamics of phase separated TDP-43. The molecular condensates displayed different characteristics and the fluorescence recovery after photobleaching was faster, indicating for the first time a role of the putative SUMOylation site K136 in the phase separation of TDP-43 in vivo. Discussion: Understanding how LLPS regulates TDP-43 accumulation and how processes, such as PTMs and stress can affect TDP-43 compartmentalisation has important implications for our understanding of TDP-43 pathology. Here we demonstrate for the first time in vivo how TDP-43 undergoes phase separation and how SUMOylation can affect mislocalization and aggregate formation. Understanding these fundamental molecular underpinnings that drive TDP-43 assembly and disassembly may hold the key to developing novel therapies for patients in the future.

cindy.maurel@mq.edu.au
Background: Aberrantly expressed fused in sarcoma (FUS) is a hallmark of FUS-related amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD) (1). Wildtype FUS localises to synapses (2) and interacts with mitochondrial proteins (3,4) while mutations have been shown to cause pathological changes affecting mitochondria, synapses, and the neuromuscular junction (NMJ) (5). This indicates a crucial physiological role for FUS in regulating synaptic and mitochondrial function that is currently poorly understood. Objectives: Identify synaptic and mitochondrial downstream effects of mutant FUS both in vitro and in vivo; analyse a probable novel interaction between FUS and Syntaphilin (SNPH); investigate global protein translation in primary cortical neurons expressing mutant FUS. Methods: Neurite complexity and synaptic density were studied in vitro in primary rat neurons expressing eGFP-FUSWT, eGFP-FUSR514G, or eGFP-FUSDNLS to determine synaptic and mitochondrial downstream effects. To investigate this in vivo, a motor neuron-specific zebrafish model was developed and co-injected with MNX1:Gal4 and UAS: eGFP-FUSWT, UAS: eGFP-FUSR514G or UAS: eGFP-FUSDNLS and axon growth in real-time, branching and NMJ density were measured. Using live imaging, the mitochondrial movement was probed in in vitro neuronal models expressing mutant FUS. Complimentary proximity ligation assays assessed the endogenous interaction of FUS and SNPH while overexpression of mutant FUS and eGFP-SNPH evaluated if mutant FUS led to alterations. Lastly, Puromycin assays investigated how mutant FUS caused differences in global protein translation. Results: We found that overexpressing mutant FUS alters synaptic numbers and neuronal complexity in both primary rat neurons and zebrafish models. The degree to which FUS is mislocalised leads to differences in the synaptic changes which are mirrored by changes in mitochondrial density and transport. Furthermore, we show that FUS interacts and localises with Syntaphilin (SNPH) and that mutations in FUS affect this relationship, which may lead to the synaptic and mitochondrial phenotypes observed. Finally, we show that mutant FUS led to changes in global protein translation. Conclusions: We provide evidence that mislocalised cytoplasmic FUS causes mitochondrial and synaptic changes and that FUS plays a vital role in maintaining neuronal health in vitro and in vivo. Moreover, we show that FUS and SNPH localisation could explain the synaptic and mitochondrial defects observed leading to global protein translation defects. Importantly, our results support the 'gain-of-function' hypothesis for disease pathogenesis in FUS-related ALS. However, controversy has arisen over the nature of the phenotype with laboratories reporting contradictory findings (2). Objectives: To resolve this controversy, we have begun the most detailed phenotypic characterisation of the model so far. Methods: A cohort of C9orf72 BAC animals from Jackson laboratories have undergone a range of behavioural and electrophysiological tests assessing motor function, and natural behaviours, such as nesting, burrowing, and social recognition. Results: Over the course of 1 year, C9orf72 BAC mice have displayed no discernible abnormalities compared to nontransgenic controls. Discussion: Due to the lack of any discernible phenotype in C9orf72 BAC mice from Jackson laboratories, we have obtained transgenic C9orf72 BAC mice directly from an academic laboratory reporting a prominent neurodegenerative phenotype for comparison. This cohort will help us understand if, for example, epigenetics or methodological differences are responsible for the controversy surrounding the model. Background: CFAP410 has been identified as a risk gene for ALS (1). CFAP410 is a basal body protein shown to be important for primary cilia formation and also shown to bind to cytoskeletal elements. The N-terminal conserved region of CFAP410 has a predicted mitochondria localization signal, two tandem leucine-rich repeats followed by a leucine-rich cap. There is also a short, conserved stretch at the C-terminus of the protein. CFAP410 has been shown to interact with NEK which is also found to be mutated in ALS. Objectives: ALS-associated variants in CFAP410 occur throughout the protein and it is unclear which of these variants are causal for the disease and how they contribute if all to the aetiology of ALS. Our objective is to understand the cellular and molecular mechanisms by which variants in CFAP410 cause ALS. Methods: Using CRISPR/Cas9 we introduced an epitope tag and the ALS-associated variants into the endogenous Cfap410 gene in mouse embryonic stem cells. The expression of the tagged mutant gene was studied by immunocytochemistry and western blotting. The characterised tagged knockin clones were differentiated to motor neurons and the consequences of the ALS-associated variant were studied using molecular and cellular approaches. Results: We have successfully tagged the endogenous Cfap410 gene in mouse embryonic stem cells with a c-terminal epitope tag using the gene editing approach and show the expression of the tagged Cfap410 protein which localises to the basal body of the primary cilia. Using these tagged ES cells, we successfully knocked in two ALS-associated variants into the tagged gene and show that there is no loss of Primary cilia in the motor neurons derived from the Cfap410 mutant ES cells. However, we find that in these neurons the DNA damage response is affected causing their death. We also find the TDP43 is redistributed to the cytoplasm and axons and interactions of Cfap410 with some of its known interacting partners are affected. Discussion: Proteins associated with the basal body of primary cilia are increasingly being implicated in the DNA damage response. We see in the case of Cfap410 that this is also the case. Point Mutations in the gene do not affect the formation of primary cilia but cause the death of motor neurons by affecting the DNA damage response. bssvss@bath.ac.uk