Theme 02 - GENETICS AND GENOMICS

title: Mitochondrial copy number in Motor neuron disease Background: Mitochondrial DNA copy number is a well validated biomarker of mitochondrial function, known to correlate with overall mortality as well as prevalent and incident neurodegenerative disease (1). Previously, multiple lines of evidence have implicated mitochondrial function has in the pathogenesis of both familial and sporadic motor neuron disease and altered mitochondrial function has been observed in both relevant post-mortem tissue and in vitro models of MND (2). Measurement of mitochondrial DNA copy number can be reliably obtained from whole genome sequencing data, allowing mitochondrial function in MND to be investigated at a far greater scale than previously attempted. Objectives: To investigate changes in mitochondrial copy number within MND, between genetic subtypes, and its correlation with phenotype. Methods: Mitochondrial copy number and genotype was extracted from whole genome sequencing data from the 52 Genetics and Genomics 32nd International Symposium on ALS/MND

Introduction: Amyotrophic lateral sclerosis (ALS) is a progressive and fatal neurodegenerative disease characterized by the lost of upper and lower motor neurons. Approximately 10% of ALS cases are familial and remaining cases are apparently sporadic. The knowledge of genetic basis of Amyotrophic Lateral Sclerosis (ALS) has made a great progress over the last 10 years, including the detection of dominant and recesive variants at several new genetic loci. In fact, today it is thought that up to 10% of sporadic ALS cases may be due to pathogenic mutations in different genes. Material and methods: ALS patients (familial cases n ¼ 11; sporadic cases n ¼ 271) and controls (n ¼ 127) were recruited through the Madrid group of Project MinE Consortium, after signing the informed consent. All patients were diagnosed according to the revised El Escorial criteria. After the isolation of the blood DNA, samples were sequenced using the Illumina HiSeq platform with a 30X coverage. Binary sequence alignment/map formats (BAM) were generated for each individual and genomes were aligned to hg19. The annotation of exonic variants in a panel of 22 ALS genes (C9orf72, SOD1, TARDBP, FUS, TBK1, TUBA4A, MATR3, CCNF, NEK1, ANXA11, TIA1, ALS2, NEFH, SIGMAR1, SETX, VAPB, VCP, DCTN1, ANG, SQSTM1, FIG4, OPTN) was done by using Bystro genomic tool (https://bystro.io/), considering a MAF <0.01 as cut-off. Intronic and synonymous variants were excluded for further analysis. The pathogenicity of a variant was categorized according to in silico scores associated to protein damage and conservation by the use of Varsome search engine. The Genome Aggregation Database (gnomAD) was used to provide an additional large control population reference. Expansion Hunter-v2.5.1 25 was used to assay the hexanucleotide repeat expansion in the C9orf72 gene. Results: 45,5% of familial ALS (fALS) cases presented likely pathogenic variants, detected in SOD1 or C9orf72 genes. 107 patients out of 271 with sporadic ALS had variants (n ¼ 134) in their sequences. Eleven carried the pathogenic hexanucleotide repeat expansion in C9orf72 (4%). After filter exonic variants with Varsome, twelve were selected. However, after comparing with variants found in local controls, only 10 in DCTN1 (3), FIG4 (1), NEK1(2), OPTN (1), SETX (1) and SQSTM1(2) mantained their clasification as potentialy causatives of the disease. Conclusions: In our cohort 45.5% of fALS and near of 8% of sALS carried potentialy pathogenic variants, in acordance with other studies. However, mutations in SOD1 gene were less frequent than expected in sALS patients. The introduction of a native control population in the study is important in order to avoid possible variant classification errors.

Acknowledgments
This work has been supported by FUNDELA and Instituto de Salud Carlos III. gerardoalonso.m@gmail.com GEN-02 Three cases of FUSassociated amyotrophic lateral sclerosis with different phenotypes Background: Alterations in the functionality of the Fused in Sarcoma FUS gene have been put in relation with an array of neurodegenerative disorders, among which familial amyotrophic lateral sclerosis (ALS) with aggressive progression. The FUS protein belongs to a class of highly conserved and ubiquitously expressed RNA-binding proteins of the FET complex that shuttle between the cell nucleus and cytoplasm. The FET complex regulates mRNA transcription, alternative splicing, RNA stability, microRNA biogenesis, apoptosis, and cell division. The majority of ALS-associated FUS mutations interfere with the Transportin 1 binding site, leading to cytoplasmic accumulation of FUS and eventually to a block of nuclear importing processes. Previous literature shows that FUS can also cause other muscular diseases and sarcoma. Objectives: Here, we describe three different cases of FUS mutations with contrasting phentoypes. First case: a 43 years old man with a strong familial history for ALS (including mother and grandmother) that came to our attention in 2019 for a history of progressive hyoposthenia and muscle hypotrophy of his right leg with signs of lower motor neuron impairment. After one year of monomelic involvement, the patient showed a diffuse muscular involvement, requiring PEG and tracheostomy in 2021. He has no cognitive impairment. The genetic test showed a heterozygous mutation in the FUS gene in variants R > 521C of exon 15. Second case: a man that came to our attention in 2007, at the age of 36, for the occurrence of upper limbs hyposthenia. Interestingly, he suffered from drug-resistant psychotic depression since 1990. His sister died at the age of 40 for cerebral malignancy and his grandfather in the maternal line died for ALS at the age of 66, after 3 years of typical disease. The patient soon developed tetraparesis and needed the positioning of PEG and tracheostomy, but he is still alive. He has no cognitive impairment. The NGS analysis showed a heterozygous mutation in c.667_672 of CGGCGG in exon 6 of FUS. Introduction: Amyotrophic lateral sclerosis (ALS) is an incurable disease with a significant genetic underpinning. Despite heritability estimates of 52%, GWAS studies have discovered only seven genome-wide significant hits, which are relevant to <10% of ALS patients. Recently, a novel machine learning model called RefMap was developed to increase the power of gene discovery by integrating motor neuron functional genomics with ALS genetics. 690 candidate ALS genes were identified using RefMap. Genetic convergence between common and rare variation highlighted BNC2 and KANK1 as new ALS genes. Aims and objectives: To develop cell models to characterise the pathogenicity of ALS-associated KANK1 and BNC2 mutations. To identify novel therapeutic targets for the treatment of ALS. Results: In neuronally differentiated SH-SY5Y and iPSC cells we used CRISPR/SpCas9 editing to recapitulate ALS-associated regulatory and coding mutations within KANK1. Coding and non-coding CRISPR/SpCas9 perturbations reduced the expression of KANK1 mRNA and produced neurotoxicity with disruption of the distal axon. Moreover, KANK1 coding mutations in iPSC-derived motor neurons led to hypoexcitability and TDP-43 cytoplasmic mislocalization. Coding mutations within BNC2 in neuronally differentiated SH-SY5Y cells reduced the expression of BNC2 mRNA and produced neurotoxicity. Conclusions: KANK1 and BNC2 are new ALS risk genes, which are enriched with common and rare ALS-associated genetic variation across multiple domains and datasets. KANK1 is functionally related to a number of known ALS genes which are important for cytoskeletal function, including PFN1, KIF5A and TUBA4A. BNC2 is not well characterized in the context of human neurons. We have experimentally verified the link between variants linked by RefMap to ALS. Moreover, we have demonstrated that reduced expression of KANK1 and BNC2 in human CNS-relevant neuronal models is cytotoxic. Finally, our iPSC-derived motor neuron data suggest that TDP-43 mislocalization is downstream of dysfunction within the distal axon, which has implications for our understanding of the molecular pathogenesis of ALS more broadly. KANK1 and BNC2 upregulation are new therapeutic targets that may be broadly relevant for the treatment of sporadic ALS. Background: For over a decade it has been known that intronic variants in UNC13A, a critical synaptic gene, increase risk/severity of ALS and frontotemporal dementia (FTD). However, despite being one of the strongest genome-wide association study (GWAS) hits for these diseases, the molecular mechanism underpinning this association has remained unclear. Results: Loss of nuclear TDP-43 protein, which is a key regulator of splicing, is a molecular hallmark of 97% of ALS and approximately half of FTD cases. In this study, we identify a new cryptic exon (CE) in UNC13A which is robustly induced by loss of nuclear TDP-43. This CE, which contains a premature termination codon that promotes nonsense-mediated decay, is detected both in cellular disease models and in patients, and is present at exceptionally high levels in the nuclei of FTD patient neurons. Strikingly, two ALS/FTD-linked single nucleotide polymorphisms (SNPs) in UNC13A are in close proximity to the cryptic splice sites and reside within TDP-43 binding footprints. We demonstrate that these SNPs modulate the binding of TDP-43 to the CE-containing intron, promoting increased levels of this deleterious cryptic splicing event, thus explaining why these SNPs increase risk/severity of ALS/FTD. Discussion: Through a wide range of experimental and bioinformatic approaches, we have resolved the long-standing mystery as to why intronic variants in UNC13A increase risk/ severity of ALS/FTD. Excitingly, the UNC13A CE is an ideal candidate for splice-switching antisense oligonucleotides therapies. Because the CE is detected even in patients without the UNC13A risk SNPs, such approaches have therapeutic potential for the vast majority of ALS and FTLD-TDP patients.
oscar.wilkins.18@ucl.ac.uk contributing factors for the initiation and progression of disease pathology in amyotrophic lateral sclerosis (ALS). Objective: This study was designed to reveal how ALS cortex communicate their disease state via exosomes, and to reveal the content of the exosomal protein cargo. Methods: Well defined and characterized mouse models, which develop ALS due to misfolded SOD1 toxicity and TDP-43 pathology are utilized together with wildtype healthy littermates. To isolate cortex derived exosomes, dissociated cortical cells were prepared from both disease models and healthy controls, and exosomes secreted to the medium were collected after 10 days in vitro. Proteomic analyses of isolated exosomes suggested the presence of a "message" sent both by hSOD1G93A and prpTDP-43A315T motor cortex, which suggested the modulation of a key transcription factor. In an effort to investigate whether this message is "protective" or "detrimental", we performed gene modulation experiments in both healthy and diseased neurons and investigated the impact of this transcription factor on neuronal survival. Results: We find that the motor cortex of both hSOD1G93A and prpTDP-43A315T mouse models of ALS utilize exosomes to communicate their disease state as early as postnatal day 3 (P3). Our results suggest that proteins present in these exosomes appear to signal a message that codes for protection from upcoming neurodegeneration at this early age. Discussion: This study unravels how cortex is involved in ALS early in the course of disease, and how exosomes can modulate key cellular events that may be targeted to overcome neurodegeneration. mukesh.gautam@northwestern.edu GEN-07 Mutations in the SPTLC1 gene are a cause of juvenile amyotrophic lateral sclerosis Background/Introduction: SPTLC1 encodes a critical subunit of serine palmitoyltransferase, the enzyme catalyzing the first and rate-limiting step in de novo sphingolipid biosynthesis. Mutations in this gene are known to cause hereditary sensory autonomic neuropathy, type 1A (HSAN1), and shown to cause an increase in the production of a neurotoxic metabolite, deoxymethyl-sphinganine. This is due to an altered enzymatic utilization of L-alanine or glycine over the canonical L-serine. Methods: We had four unrelated individuals between the ages of 5 and 15 diagnosed with juvenile onset ALS. None of these cases had a family history of ALS or neuromuscular disorders, suggesting de novo variations as the underlying genetic mechanism. De novo mutations in novel genes might also be involved in adult onset ALS. To delineate the genetic causes of early onset ALS, we performed exome sequencing on each index case along with their unaffected parents. We also screened adult onset ALS cases for mutations in SPTLC1. Results: We identified four de novo mutations in SPTLC1, of which two mutations resulted in the same amino acid change, p.Ala20Ser, whereas the other two mutations were in p.Ser331Tyr and p.Leu39del. Building on previous work in cellular models of SPTLC1 in HSAN1, we show that similar mitochondrial defects reported for HSAN1 were also present in HEK293FT cells stably expressing mutant p.Ala20Ser. Interestingly, when cells were treated with excess serine, features of mitochondrial abnormalities were reverted. Whilst preliminary, we demonstrate that oral L-serine significantly improved one of our patient's failure to thrive, which is a promising early clinical proxy following the natural course of the disease. Plasma sphingolipid measurements in the patient were also taken over the course of serine supplementation and we did not find excess derivatives of ceramides indicating that there is no overt activation of the sphingolipid/ceramide pathway. Mutation screening in sporadic ALS cases reveal that SPTLC1 mutations in adult onset ALS is rare, accounting for 0.34% in 5607 adult onset ALS cases. Discussion/Conclusion: Despite this rarity, given the promising benefit of serine supplementation in our patient and the lack of effective treatments for ALS, suggests that it may be worthwhile to screen adult ALS patients for SPTLC1 mutations. In aggregate, our data broaden the phenotype associated with mutations in SPTLC1 to include juvenile ALS and implicates sphingolipid metabolism as a pathway in motor neuron disease and crucially we demonstrate an early proofof-concept to test the precision medicine approach in an otherwise fatal neurodegenerative disease. chiarp@mail.nih.gov GEN-08 Altered circular RNA expression and circular RNA-microRNA interactions in amyotrophic lateral sclerosis S. Brennan 1 , C. Aquilina-Reid 1,2 , A. Curry-Hyde 2 and M. Janitz 2,3 from a biomarker development perspective (2). The best described function of circRNAs is their ability to act as microRNA sponges (3). In this process, circRNAs bind to microRNAs which would normally target and degrade messenger RNAs. This then causes de-repression of these targeted mRNAs. Methods: In this study, we analysed RNA expression profiles of ALS patient samples supplied by the New York Genome Consortium. We selected a total of 50 ALS patient samples and 19 healthy control samples from this data set divided across the cervical, thoracic and lumbar spinal cord regions. Differentially expressed (DE) circRNAs were identified using CIRI2 and CircExplorer and circRNAs detected by both tools were used in subsequent steps. We then identified genomic loci that produced >10 unique DE circRNAs and determined the biological functions of these genomic loci. We also identified miRNA binding sites in our set of DE circRNAs and ranked them by frequency. In parallel, we identified DE mRNAs from the same data set using limma and EdgeR, and integrated mRNA and circRNAs data with the aim of identifying potential de-repression relationships mediated by miRNA. Results: We identified numerous hotspot genomic loci and 54.1% of these hotspot genes were identified in the thoracic spinal cord region. The genes from which these circRNAs were produced have roles in regulation of GTPase activity, cell junction assembly and DNA repair. Three genetic loci, in particular, seemed to produce a disproportionate number of circRNAs warranting further study as biomarkers of ALS. We also identified 92 differentially expressed circRNAs using stringent statistical parameters and found evidence for mRNA differential expression potentially due to the miRNA sponge effect. The mRNAs seemingly affected by this mechanism had roles in response to Nitrogen and Oxygen containing compounds, as well as response to peptide hormone signaling and wnt signaling. Discussion: These circRNAs offer novel insight into the mechanism by which oxidative stress and wnt signaling can become disrupted in ALS. This study demonstrates the potential for biomarker and therapeutic candidate discovery through exploration of the circular transcriptome.

sam@genieus.co
Background: To date, gene mutations remain the only proven cause of ALS, with more than 30 ALS genes identified. However, 90% of ALS cases (about 40% of familial and 90% of sporadic ALS) still have an unknown genetic cause. Most research in ALS has focused on small nucleotide level variants only.
Objectives: This project extends ALS gene discovery to also consider large structural variation by using bioinformatic pipelines to analyse whole genome sequencing (WGS) data for the identification of both small and large genetic alterations in a small ALS family that has limited power for traditional gene discovery methods. Methods: WGS was performed on two affected individuals from a family with a history of ALS (MQ52). To identify single nucleotide variants and small insertion/deletions potentially acting as the pathogenic gene mutation causing ALS within this family, shared variant analysis using custom bioinformatics was carried out to filter for novel nonsynonymous and indel variants absent or extremely rare from publicly available control databases. Variants were validated using the Integrative Genomics Viewer and direct sequencing. To identify structural variants (SVs), we have developed a comprehensive pipeline that calls SVs from WGS data using multiple tools such as Manta, Lumpy and MetaSV. The tool Duphold, is then used to filter SVs based on quality scores, followed by SV annotation using Reciprocal Overlap Annotator and AnnotSV to generate a list of shared or overlapping between the two affected individuals in the family. Results: Initial analysis of family MQ52 identified a total of 14 small candidate variants (13 single nucleotide variants and one indel). Using the SV discovery pipeline, 80 novel/rare SVs (from a total of over 8000 total family SVs) were identified as potential causes of disease in this family, out of which 44 SVs lie in coding regions of the genome. This SV pipeline is still undergoing refinement and following further filtering resultant SVs will be validated in the lab using established PCR techniques. Discussion: Our gene discovery pipelines can successfully identify both small and large variants with the potential to cause disease within small ALS families. Identifying novel ALS variants has great potential to identify novel ALS genes and pathways, that could help with the development of future treatments. Given  Background: Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disorder that predominately involves the motor neurons in the brain and spinal cord (1). The TARDBP gene, encoding TAR DNA-binding protein 43 (TDP-43) protein, has been identified as a major causative gene in ALS (2). Objective: To report one novel TARDBP mutation and three known TARDBP mutations in a Chinese cohort with 275 sporadic ALS patients and 20 unrelated familial ALS probands.
Methods: This study included 20 unrelated FALS probands and 275 SALS patients referred to the Department of Neurology, Fujian Medical University Union Hospital, and the Department of Neurology, Fujian Provincial Hospital between January 2017 and December 2020. Targeted next-generation sequencing (NGS) was performed to screen mutations within known ALS causative genes. Results: We identified a novel variant, p.K176I, in the TARDBP gene in a FALS family presenting with frontotemporal dementia and parkinsonism. The p.K176I variant was interpreted as likely pathogenic according to the ACMG guidelines (3). Moreover, we identified other three known mutations in one FALS proband and three SALS patients. The incidences of TARDBP mutations in our FALS cohort and SALS cohort were 10% and 1.1%, respectively. Discussion: Our results support that TARDBP mutations are one of the most common mutations in both FALS and SALS in China. Patients with TARDBP mutations may have a broad phenotype spectrum of ALS, FTD, and parkinsonism. The TARDBP gene should be included in genetic screening for ALS with FTD, and/or parkinsonism. Background: The genetic role in sporadic ALS, which accounts for 90% of all ALS cases, is yet to be discovered. In addition to single-nucleotide variations, a growing number of structural variations (SVs) have also been associated with the risk of both familial and sporadic ALS. These include the GGGGCC hexanucleotide repeat expansion in C9orf72 and CAG trinucleotide repeat expansion in ATXN2. Despite advances in next-generation sequencing technologies, the molecular mechanisms underlying ALS pathogenesis are still far from being understood. One plausible explanation is that the disease-causing genetic variants may be located in genomic regions that are difficult to discover using conventional sequencing technologies such as short-read sequencers. Objectives: We aimed to determine if long read sequencing (LRS) offers a means to discover potential disease-associated SVs in two ALS patients with no prior family risk of the disease. Methods: The genomic DNA from participants were sequenced on PacBio Sequel Platform with an average coverage of 10Â. Following pre-processing of sequencing results, the data were filtered to prioritize rare potentially deleterious SVs based on several criteria: (i) pass the default variant caller filters; (ii) not reciprocally overlapped with common SVs from population databases; and (iii) not predicted as benign by annotSV ranking. Candidate SVs were confirmed by Sanger sequencing. Expression of the genes with candidate SVs was analysed using RNA sequencing. Case control analysis of candidate genes were also performed using short read sequencing of 213 classical ALS and 27 matched control genomes from the New York Genome Centre. Results: We report candidate SVs in ten genes that may be involved in the aetiology of ALS in the two patients which may be applied to therapeutic strategies for these two patients, as well as other sporadic ALS cases. Discussion: Each of the candidate genes has the capacity to play a role in either the aetiology or disease progression of ALS, supporting current evidence that suggests polygenic/oligogenic model and multifactorial nature of sporadic ALS (1). We present here, the ability of long read sequencing to detect SVs that are relevant candidates for understanding ALS pathophysiology. While these variants may be identified through short-read technologies, the limitations in sequencing long reads (typically up to 600bp) and high false-positive rates commonly associated with the approach are not practical for genome wide high throughput SV detection. The strategy presented here shows that there is good reason to search the genome much more thoroughly than what is currently standard practice. Background: Amyotrophic lateral sclerosis (ALS) is a lateonset neurodegenerative disorder. Mitochondrial dysfunction is involved in the complex pathophysiology of ALS; however, the role of mitochondrial DNA (mtDNA) variants in ALS is poorly understood. We aimed to elucidate the role of mtDNA variants in the pathogenesis of ALS. Methods: The mitochondrial haplogroups of 585 ALS patients and 371 healthy controls were determined; 38 ALS patients and 42 controls underwent long-range polymerase chain reaction combined with next-generation sequencing technology to analyze whole mitochondrial genome variants. Results: A higher percentage of variants accumulated in ALS patients than in controls. Analysis of coding region variations that were further stratified by mtDNA genes revealed that nonsynonymous variants were more vulnerable in ALS patients than in controls, particularly in the ND4L, ND5, and ATP8 genes. Moreover, pathogenic nonsynonymous variants tended to over-represent in ALS patients. Unsurprisingly, nonsynonymous variants were not related to the phenotype. Haplogroup analysis did not found evidence of association between haplogroups with the risk of ALS, however, patients belonging to haplogroup Y and M7c were prone to develop later onset of ALS. Discussion: This is the first study to profile mtDNA variants in ALS patients from mainland China. Our results suggest that an increase in the number of nonsynonymous variants is linked to the pathogenesis of ALS. Moreover, haplogroup Y and M7c may modulate the clinical expression of ALS. Our findings provide independent, albeit limited, evidence for the role of mtDNA in the pathogenesis of ALS. Background: Neuroinflammation is a well characterised pathophysiological mechanism involved in the onset and progression of amyotrophic lateral sclerosis (ALS). Regulatory T cells (Tregs) are immune modulators that suppress inflammation and prevent the onset of autoimmune disorders. These cells dramatically and progressively decreased in ALS patients, with lower levels associated with shorted survival. Interleukin2 (IL2) is a crucial cytokine for Treg differentiation and functions and the administration of low dose IL2 (ld-IL2) has been proposed as an immune regulatory strategy for ALS to reduce neuroinflammation through the expansion of protective Tregs. Moreover, evidence of possible direct and prosurvival effects of IL2 on neurons and glial cells has been reported in the literature. Aims: To evaluate the effect of ld-IL-2 on the blood transcriptome of ALS patients included in the IMODALS clinical trial and to investigate gene expression changes associated with IL2 treatment in patient-derived astrocytes. Methods: 36 patients were recruited in the IMODALS trial and randomly assigned to three treatment arms: 1MIU, 2MIU-IL2 or placebo. They received one injection daily for five days every 28 days for a total of 3 treatment cycles. At selected time points blood was collected, white blood cells isolated and RNA was extracted. Microarray and NanoString data were generated to investigate transcriptional changes associated with the treatment and results were validated through qRT-PCR. Moreover, to investigate any potential effects of IL-2 on CNS cells, patient-derived astrocytes were differentiated from ALS or healthy volunteer fibroblasts and in vitro treated with IL2. Transcriptional differences were investigated using Oxford Nanopore cDNA sequencing. Findings: Gene expression and pathway analysis revealed longitudinal transcriptional changes throughout the IMODALS trial. In particular, evidence of a broad immune suppression was provided after the first treatment cycle (day8) while an activation of immune suppressive pathways reached a peak only at the end of the third administration (day64). In fact, a time-dependent and dose-dependent activation of Treg markers was reported which suggested a cumulative effect of ld-IL2. However, inter-individual differences were found amongst patients and participants were classified into fully, moderate and low responders. A predictive biomarker analysis was carried out and two genes were identified for which expression at recruitment was able to predict patient responsiveness to the ld-IL2 administration at day64. Finally, to investigate if IL-2 is able to mediate protective effects on CNS cells, patient-derived astrocytes from 3 C9ORF72-ALS patients, 3 sporadic ALS, and 3 healthy controls were generated and treated with IL2 following optimisation. IL2-associated transcriptional differences were evaluated through cDNA sequencing. Background: Hereditary transthyretin-related amyloidosis (hATTR) is an autosomal dominant disorder characterized by amyloid deposits in different tissues due to transthyretin (TTR) gene mutations. Although a length-dependent sensorymotor neuropathy with autonomic features is the disease hallmark, a predominantly motor involvement resembling a lower motor neuron disease has been described (1). Moreover, hATTR can mimic bulbar ALS (2). Clinical cases: We report three sporadic ALS cases who were asymptomatic Val50Met TTR gene carriers. All of them had hATTR familial history. The first and second patients were a 43-year-old man and a 37-year-old woman who presented with a spinal and bulbar onset ALS forms, respectively. Both fulfilled definitive diagnosis according to revised El Escorial criteria (3), supported by the neurophysiological Awaji guidelines (4). They died five and three years after symptoms onset, with respiratory failure, respectively. The third patient is a 52-year-old woman who presented with pure upper motor neuron signs in upper and lower limbs for two years. Electromyography did not disclose loss of motor units in bulbar, cervical, and lumbosacral-dependent muscles. A diagnosis of probable primary lateral sclerosis was proposed. In this patient, conventional neuropsychological evaluation indicated frontal dysfunction, and fluorodeoxyglucose positron emission tomography scan revealed frontotemporal hypometabolism, both favouring frontotemporal dementia. In none signs of neuropathy were detected on sequential nerve conductions studies. Brain and spinal cord MRI, and cerebrospinal fluid (CSF) analysis were also unremarkable in all. Transcranial magnetic stimulation was performed in the first and last patients, confirming involvement of the corticospinal tract. Genetic testing for C9orf72 gene expansion was requested for the last two patients, being negative for both. For the last patient, next generation sequencing ALS-related gene panel was also requested, being unrevealing. Discussion: We suggest that abnormal TTR may be involved in ALS pathogenesis, which has not been previously described, given the rarity of both TTR gene mutations and ALS in general population. Interestingly, high levels of oxidative modified transthyretin were found in CSF of ALS patients with no TTR gene mutation (5). Further studies are needed to elucidate if abnormal transthyretin protein has an additional role in motor neuron degeneration. Introduction: CAG repeat expansions of ATXN2 have been associated with both ALS and SCA2 with distinctive phenotypes presenting in association to different length of expansions. Currently, a CAG expansion with less than 30-33 repeats is linked to ALS, although with an incomplete penetrance, whereas expansions characterized by more than 30-33 repeats are associated with SCA2 (1). Case report: Here we describe the case of an ALS patient with an heterozygous missense mutation of ATXN2 gene, c.2860C>T, determining the aminoacidic substitution p.Pro954Ser, characterized by a fast progressive classic disease phenotype and dysexecutive cognitive impairment. In her family history, her mother died in the framework of behavioral dementia with concomitant extrapyramidal signs. No significant mutation was identified in any other ALS gene contained within our genetic panel. In silico analysis showed that this mutation could have a pathogenic effect on the protein.
Discussion: Recent studies point to the hypothesis that ALS is not driven by a single genetic factor, but rather by a complex interplay between several disease-associated gene products (2). ATXN2 has been shown to interact with known crucial gene products, including C9ORF72, TDP-43 and FUS (2), providing additional insight into ATXN2 synergistic contribute to disease. In this report, we enlighten the aggressive course of the disease in a patient without other known mutations in crucial ALS gene. The reason behind this peculiar manifestation is not simple to clarify. In fact, the occurrence of mutations in more than one gene supports the idea that ALS could be considered an oligogenic disease (2) with some of these genes displaying mendelian inheritance and a number of other genes acting as disease modifiers by influencing penetrance and pleiotropy. It's already known that intermediate CAG repeats of ATXN2 are strong modifiers in FTD-ALS and ALS (3), but only in the presence or absence of C9orf72 repeat expansions. Conclusion: We hope that our experience could widen knowledge about ATXN2 role in ALS disease, trusting that future studies will shed new light on the role of ATXN2 in processes that are typically perturbed in ALS/SCA2. Undoubtedly, further research is necessary to determine the prevalence of this mutation in ALS patients and its pathogenetic role. Background: Previous studies suggested that multiple rare variants in genes causing Amyotrophic Lateral Sclerosis (ALS) may synergistically interact and increase disease risk or reduce the age of onset. However, the identification of potential oligogenic ALS cases varies across studies and the oligogenic hypothesis has not been validated in a large ALS cohort. Aim: To investigate the potential role of oligogenic variation in ALS with Whole-Genome Sequencing (WGS). Methods: We analysed 1043 patients from the Piemonte and Valle d'Aosta Register for ALS (PARALS) and 750 healthy controls who underwent WGS. Genetic variants within a panel of 42 ALS-relevant genes were then assessed and patients were classified according to three different definitions of "oligogenic" cases: (a) individuals with two or more pathogenic variants according to the American College of Medical Genetics criteria; (b) individuals with at least two pathogenic or rare (Minor Allele Frequency <0.1%) variants that were absent from the internal control cohort and predicted deleterious in silico; (c) individuals carrying two or more pathogenic variants or rare variants of uncertain significance. The proportion of potential oligogenic cases was analyzed with the binomial test. Age at disease onset was then compared across oligogenic, monogenic and wild-type cases. Results: The rate of oligogenic ALS cases varies according to the criteria adopted from 1.2% of our cohort (a) to 4.8% (b) and up to 12.7% (c). However, in all the three proposed definitions of oligogenicity the proportion of patients with multiple genetic variants was not higher than expected on the basis of chance (p-value ¼0.84). We observed a higher percentage of oligogenic cases in ALS cases than in healthy controls, but this difference is driven by oligogenic combinations of a highly penetrant allele or known risk factor with another rare variant of uncertain significance: if the effect of the highly penetrant is considered, the frequency of oligogenic variants did not differ between ALS cases and controls (pvalue ¼0.26). Potential oligogenic cases are younger than wild-type ALS patients but they did not differ in both age at onset and disease severity from monogenic patients. Conclusions: We demonstrated that the frequency of potential oligogenic ALS cases depends on the criteria adopted for variant filtering. However, the presence of multiple rare variants did not increase disease risk after controlling for highly penetrant genetic risk factors. Additionally, in our cohort the presence of oligogenic variants did not influence the age of onset or disease progression. Our results highlight that a precise definition of oligogenicity in ALS is essential to interpret the results of genetic screening: understanding the role of multiple genetic variants will be increasingly necessary to optimize patients' treatment and counselling.

University Of Sheffield, Sheffield, United Kingdom
Abstract title: Mitochondrial copy number in Motor neuron disease Background: Mitochondrial DNA copy number is a well validated biomarker of mitochondrial function, known to correlate with overall mortality as well as prevalent and incident neurodegenerative disease (1). Previously, multiple lines of evidence have implicated mitochondrial function has in the pathogenesis of both familial and sporadic motor neuron disease and altered mitochondrial function has been observed in both relevant post-mortem tissue and in vitro models of MND (2). Measurement of mitochondrial DNA copy number can be reliably obtained from whole genome sequencing data, allowing mitochondrial function in MND to be investigated at a far greater scale than previously attempted. Objectives: To investigate changes in mitochondrial copy number within MND, between genetic subtypes, and its correlation with phenotype. Methods: Mitochondrial copy number and genotype was extracted from whole genome sequencing data from the project MINE cohort (9000 patients and controls) and the Answer ALS cohort (1000 patients and controls). Methylation data from whole blood from Paul Hop et al. (3) was used to investigate downstream consequences of changes in copy number. Findings: Mitochondrial copy number is greater in MND cases than controls and varies between genetic subtypes of MND demonstrating the relevance of mitochondrial function to MND. Mitochondrial genotypes known to cause changes in mitochondrial copy number are not associated with risk of MND, suggesting changes in mitochondrial function are a downstream consequence of MND pathology. We find changes in the methylation of nuclear DNA at sites previously shown to be caused by altered mitochondrial copy number, demonstrating that changes to mitochondrial copy number seen in ALS have functional effects on cellular processes. Background: Cytoplasmic dynein 1 is responsible for the majority of retrograde transport. Signalling endosomes, structures formed during receptor endocytosis, link dynein to processes of cell signalling, such as EGF/EGFR and BDNF/TrkB pathways (1). Their degradation is required to terminate the induced signal after ligand-receptor binding. Dynein trafficks signalling endosomes to lysosomes, playing a role in terminating receptor signalling. Signalling pathways activate a variety of different proteins and targets, and a major subclass of these proteins are kinasesincluding ERK1/2, and Akt. These kinases modulate a variety of cell functions such as survival, proliferation, or migration, through changes in gene expression (2). As a result, alterations in signalling pathways may have knock-on effects on this process, leading to potentially pathogenic changes in gene expression. Previous research indicates that impaired dynein function through the Legs-atodd-angles (Loa) mutation in mice (3) leads to delayed trafficking of signalling endosomes, with increases in ERK1/2 phosphorylation following EGF and BDNF stimulation in Mouse Embryonic Fibroblasts (MEFs) and motor neurons respectively (4), representing a potential alteration to the typical signalling pathway. Objectives: We hypothesise that, due to the axonal length of motor neurons, deficits in retrograde transport may substantially alter signalling pathways in these neurons and lead to potentially pathological changes in gene expression. This may lead to gradual cell death after development, or negatively impact development during gestation through changes in neuronal migration or axonal extension and contribute to the vulnerability of motor neurons in ALS. Thus, we aim to investigate the effect of dynein impairment on gene expression directly. Methods/Results: Through Nanopore sequencing, we are investigating changes in gene expression when dynein function is impaired in Loa MEFs and have found that over 300 genes are significantly (p < 0.05) dysregulated, combined with 2-3 fold increases in ERK1/2 phosphorylation, indicating that dynein may play a role in regulating gene expression through timely degradation of receptor signalling, modulating kinase and signal pathway activity. Multiple genes show biological relevance to motor neuron and muscle health, and we are investigating these cell types directly to further understand how pathological changes in gene expression arising from impaired dynein function may contribute to the progression of MND.
cm783@sussex.ac.uk and disease duration. Neurodegeneration in the ALS spinal cord is accompanied by a large glial response from activated microglia and astrocytes. However, it is unclear whether this inflammatory response influences the course of disease, nor whether it is under genetic regulation. Objectives: We set out to investigate how clinical factors influence gene expression in post-mortem spinal cords, and whether common genetic variants that modulate ALS risk can act through particular genes and/or cell-types. Methods: The New York Genome Center ALS Consortium cohort contains 380 spinal cord RNA-seq samples from 203 individuals. This includes 154 mostly sporadic ALS patients and 49 non-neurological disease controls. All donors have matched whole-genome sequencing. The dataset contains cervical, thoracic and lumbar spinal cord segments. We performed differential gene expression between ALS cases and controls in each spinal cord section, testing differentially expressed genes for enrichment with cell-types and pathways. We generated gene co-expression networks using the ALS cases to identify modules of genes that correlated with clinical traits. We then used the genotype data to map cis-regulated expression and splicing quantitative trait loci (eQTL/sQTL) across all donors in the cohort of European ancestry, followed by colocalization analysis between our QTLs and the latest ALS GWAS. Results: We identified thousands of differentially expressed genes in the ALS spinal cord as compared to controls. Celltype deconvolution and gene-set enrichment analyses showed a decrease in oligodendrocytes, accompanied by an increase in microglia and astrocyte markers, including markers of glial activation. Co-expression analyses within ALS cases found several modules enriched with microglia and astrocyte genes. We found microglia modules to be negatively correlated with disease duration, but not with age of disease onset nor site of onset. Reduced microglial activation and/or proliferation in ALS cases with slower disease progression validates multiple previous findings from mouse models and smaller patient cohorts. Our QTL mapping discovered 9492 and 5627 genes with genetically regulated expression and splicing respectively, across the three spinal cord sections. Colocalization suggested candidate ALS risk genes for several genome-wide significant and nominally significant ALS GWAS loci, including C9orf72, ATXN3, ACSL5 and FNBP1. Our co-expression analysis suggests that certain risk genes act within specific cell types, such as FNBP1 through oligodendrocytes and ACSL5 through activated microglia. Discussion: In summary, we characterize transcriptomic changes in ALS spinal cord, hypothesize mechanisms for GWAS loci, and provide a transcriptomic and genetic resource for the ALS community.
jack.humphrey@mssm.edu GEN-20 Genetic analysis identifies pathways and cell types involved in amyotrophic lateral sclerosis disease risk Department of Neurology, Johns Hopkins University, Baltimore, MD, USA Background/Introduction: Polygenic risk plays a central role in amyotrophic lateral sclerosis (ALS) risk. Genome-wide association studies (GWAS) offer a method to explore this type of risk. Furthermore, minor effect variants are likely to be biologically related, converging on the molecular mechanisms underlying disease risk. Methods: To address this gap, we systematically applied polygenic risk score analysis to a genomic dataset involving 78,500 individuals to distinguish the cellular processes and the cell types relevant to the disease process for ALS. To identify genes that mediate risk via changes in expression level, we performed a two-sample Mendelian randomization using gene expression from brain and blood. Results: Several biological processes involved in neuron projection morphogenesis, membrane trafficking, and signal transduction mediated by ribonucleotides were identified as drivers of ALS risk, specifically in GABAergic interneurons and oligodendrocytes. Using two-sample Mendelian randomization, we also nominate six differentially expressed genes in the blood that were significantly associated with the risk of developing ALS (ATG16L2, ACSL5, MAP1LC3A, MAPKAPK3, PLXNB2, and SCFD1). Summary: In aggregate, we show the convergence of many genetic etiologies onto a smaller number of final common pathways and cell types. Our work highlights the power of using genetic data to predict functional outcomes. By extension, we show the potential of this approach to identify new pathways that could be explored further as targets for therapy. ( Background: Repeat expansions have been proved to be the cause and risk factor of amyotrophic lateral sclerosis (ALS), such as tandem nucleotide repeat in exon or intron of C9orf72 and ATXN2 (1,2). However, there are few research about the association between repeat expansion in the AR and NOP56 genes and ALS (3), especially in China. Objectives: To evaluate the frequency of C9orf72 and ATXN2 nucleotide repeat mutations in ALS patients, and test the hypothesis whether repeat expansion in the NOP56 and AR genes are risk factors for ALS in mainland China. Methods: In this study, 736 patients with ALS and several hundred healthy controls from mainland China were recruited. Polymerase chain reaction (PCR) and Repeatprimed PCR (RP-PCR) were performed to determine the different nucleotide repeat size within the C9orf72, ATNX2, AR, and NOP56 genes. Results: The GGGGCC hexanucleotide repeat expansion in C9orf72 was observed in six ALS patients (0.8%, 6/736), in none of the controls (0/365). The patients with GGGGCC repeat expansion had a shorter median survival than those with the normal length (p ¼ 0.006). With regard to the CAG repeat in ATXN2, we identified the intermediate (27-34) CAG repeat was associated with ALS (p ¼ 0.031); and except for the family history (p ¼ 0.044), there was no significant difference in other clinical characteristics between the groups with and without intermediate CAG repeat (p > 0.05). Meanwhile, we observed that NOP56 repeat lengths (>8 repeats) were associated with an increased risk for ALS (p ¼ 0.041), but the phenotypes in the two groups were similar (p > 0.05). Finally, we found no association between CAG repeat length in the AR gene and ALS. Discussion: The results demonstrated that in our ALS cohort, the pathogenetic C9ORF72 GGGGCC repeat is very rare, while the intermediate-length size of CAG repeats in the ATXN2 gene is more frequent which increase the risk of ALS but have no effect on disease phenotypes; and the large size of GGCCTG repeats expansion in NOP56 may be a risk factor for ALS as well. Background: Although a growing number of variants and genes have been associated with MND, it is difficult to efficiently extract meaningful biological information from bulk genetic findings of MND. Methods: To obtain a complete list of genes and variants relevant to MND, systematic searches were performed in the PubMed database prior to 30 August 2020, and then, MNDrelated publications were reviewed manually. Sixty-three genomic data sources were integrated to annotate each variant and gene. Results: We developed a comprehensive, one-stop, online database named Gene4MND by integrating five classes of genetic data (rare variants, copy-number variants, associated risk variants from genome-wide association studies (GWAS), differentially expressed genes and differential DNA methylation of genes) from published MND-related studies and 63 popular genomic data sources. Furthermore, Gene4MND utilizes a scoring system to prioritize genes and provides an analytic platform for MND. Gene ontology and protein-protein interaction analyses based on prioritized genes revealed that most prioritized genes were involved in immune-related pathways. To identify high-confidence genes by their relevance to MND, gene prioritization (except for C9orf72 and ATXN2) was conducted by combining the different types of genetic evidence. Discussion: With the development of sequencing technologies, a growing number of genetic insights into MND have been revealed. Nevertheless, the pathophysiology of MND remains largely unknown, and the scattered results create a major challenge for interpreting close relationships between genotypes and phenotypes. Therefore, we constructed a comprehensive genetic resource for MND, Gene4MND, to systematically integrate multiple types of data from various studies and to obtain meaningful biological information from genetic findings of MND.

liuzhen0968@csu.edu.cn
Research Institute, Sydney, Australia; 3 Sydney Medical School, University of Sydney, Sydney, Australia; 4 Molecular Medicine Laboratory, Concord Hospital, Sydney, Australia; 5 Department of Clinical Medicine, Sydney, Australia Background: Gene mutations remain the only known cause of Amyotrophic lateral sclerosis (ALS), and with over 30 ALS genes identified, ALS is a genetically heterogeneous disease. Approximately 10% of cases are familial, while the remaining cases occur apparently sporadically. Despite heritability estimates of 40-60% for all forms of ALS, only two-thirds of familial and $10% of sporadic cases can be explained by known ALS gene mutations. As such, almost 90% of ALS cases remain with no known genetic predisposition to disease. Although almost all ALS gene mutations identified thus far have been small nucleotide level changes, the most common known cause is pathogenic expansion of a short tandem repeat within the C9orf72 gene. This is a complex form of genomic variation which spans thousands of nucleotides and is therefore comparable in size to structural variants (SVs). SVs are complex genomic alterations affecting long stretches of DNA and have previously been implicated in other neurodegenerative conditions including Parkinson's disease and spinocerebellar ataxia. Taken together, this suggests that novel SVs are likely to account for a percentage of ALS cases. The study of SVs in ALS has so far been precluded due to technical hindrances but is now feasible with the advent of sophisticated bioinformatics tools and sequencing technologies. Objectives: We aim to develop an innovative discovery pipeline for novel SVs contributing to the cause of ALS which combines our unique dataset of whole-genome sequencing (WGS) data from 850 ALS cases, together with cutting edge bioinformatics tools and complimentary sequencing strategies. Methods: Our pipeline incorporates several bioinformatics tools including Lumpy, Manta, MetaSV, Duphold, ReciprocalOverlapAnnotator, AnnotSV and Samplot which are first applied to WGS data to generate a high confidence SV call set for each individual. Customised R filtering is then implemented to perform cohort specific analyses and remove population-based SVs. Identified SVs then undergo wet-lab validation employing a combination of traditional PCR, repeat-primed PCR, TaqMan assays and Bionano Genomics optical genome mapping.
Results: A small ALS family negative for known ALS gene mutations has been analysed through this pipeline. After initially identifying >7000 SVs across this family, bioinformatics filtering narrowed the pathogenic candidates to just eight SVs. While our wet-lab validation strategy remains under development, validation efforts have so far excluded a further three SVs. Thus, just five SVs remain under consideration as the potential cause of ALS within this family. Discussion: The identification of novel SVs in ALS will not only implicate novel ALS genes, but also highlight SV as an important contributor to ALS pathogenesis. ALS relevant SVs will be prime targets for much needed therapeutic development, particularly given their effects on gene expression and the recent promise shown by gene silencing therapies in ALS clinical trials. emily.mccann@mq.edu.au GEN-24 Structural variants in SNCA may increase propensity for protein aggregation in amyotrophic lateral sclerosis B. Roberts 1,2 , A. Akkari 1,3,4,5 and L. Flynn 1,3 1 Perron Institute for Neurological and Translational Sciences, Perth, Australia; 2 The University of Western Australia, Perth, Australia; 3 Centre for Molecular Medicine and Innovative Therapeutics, Perth, Australia; 4 Duke University, Division of Neurology, Durham, USA; 5 Black Swan Pharmaceuticals, Durham, USA Background: Alpha-synuclein (aSyn) is a hallmark for disease in neurodegenerative disorders, such as Parkinson's disease (PD), with mutations in the SNCA gene that encodes aSyn, facilitating the development of Lewy bodies. Recently, polymorphic variants, or structural variants (SV), within the SNCA locus have been observed to constitute a significant risk factor for developing idiopathic PD, modulating gene expression via translation modification and splicing mechanisms (1). A growing body of evidence now suggests that aSyn may be a key contributor to the development of ALS, with aSyn positive Lewy bodies co-aggregating alongside ALS disease hallmarks such as SOD1 (2) and TDP-43 (3). These data provide reason to believe that the SVs associated with idiopathic PD development may implicate aSyn expression and subsequent disease risk in ALS. Objective: The aim of this study was to examine the expression of SNCA and presence of SNCA SVs in an ALS cohort to determine whether SVs may explain abnormalities in expression and, therefore, implicate overall aSyn protein expression in ALS disease risk or progression. Methods: SNCA and aSyn expression were evaluated in sporadic ALS olfactory neurosphere-derived cells, and PCR and sequencing techniques were used to determine patient genotypes relative to SV risk alleles. Results: A 4-fold increase in SNCA expression was observed when comparing average SNCA expression in ALS patients compared to healthy controls (p ¼ 0.171), with RNA from one patient displaying over a 10-fold increase in SNCA expression to that of the average expression in control cells. Immunocytochemistry revealed a significantly higher number of aSyn-dense fibrils in a high SNCA expressing patient cell line (1.75 ± 0.29%) compared to low expressing cell lines (0.16 ± 0.1%; p < 0.0001), with no fibrils present in control cell lines. Furthermore, analysis of patient genotypes revealed SV risk alleles from two out of three SVs studied exclusively present in the high SNCA expressing patient cell line, with no risk alleles present in control lines. Discussion: SNCA expression in ALS patient cells indicates that aSyn may indeed play a pathological role in ALS, with patient lines expressing higher levels of the gene transcript and greater numbers of aSyn-dense fibrils than controls. Furthermore, SVs in the SNCA gene that have been previously associated with Lewy body pathology in idiopathic PD, are found to be concomitant with the pathological overexpression of aSyn in ALS and may be useful in identifying ALS patients that could benefit from an aSyn targeted therapy. 21493468@student.uwa.edu.au