Abstracts from the 57th European Society of Human Genetics (ESHG) Conference: Oral Presentations (2025)

PL1.1 ESHG Award Lecture

Angus Clarke

As is always true, my career has been shaped by circumstances; in my case, these included the constraints of the 1980s molecular genetics lab. My involvement with genetics research began through exposure to four key conditions: (i) Xp21 (Duchenne and Becker) muscular dystrophy; (ii) X-linked hypohidrotic ectodermal dysplasia (XHED); (iii) Rett syndrome, and (iv) Huntington’s disease (HD), which (along with myotonic dystrophy) was the air we breathed in Cardiff in the age of gene mapping and cloning as it was pursued within the framework of a long-term relationship with each family and with a commitment to their social and medical welfare, enabled by a multidisciplinary genetic counselling team. I must mention Professor Peter Harper, and also Professor John Burn, as key professional influences, who provided great support and guidance.

I had already been involved with Duchenne muscular dystrophy through work in the Cardiff Muscle Clinic, in my earlier role in paediatrics. This led to an interest in newborn screening for Duchenne, a social science evaluation of its impact on families, and an interest in consent. Work with Rett syndrome led to a search for other families with more than one affected person, and that led to a close association with the UK family support group, a focus on the needs of those affected by rare diseases, and how to provide effective support without causing damage from the ‘hype’ of inflated claims for new treatments. My work with HD began more recently, when I took over the Cardiff HD predictive testing clinic from Peter Harper and then took part in projects with the UK HD predictive testing consortium, with its strong interest in genetic counselling.

However, it is perhaps my work with XHED that best illustrates my developing interests in the ethical and social aspects of human genetics. In my project on XHED, from summer 1986 to the end of 1987, I acquired: (i) blood samples from many family members across the UK for a linkage study; (ii) phenotypic information about many patients; (iii) in-depth knowledge of the motorways and housing estates of mainland Britain; and (iv) I heard the stories of many members of XHED families. This contributed to mapping the EDA gene, identifying the spectrum of pathogenic variants, and to my returning some years later with an audio recorder to listen to the stories more systematically. These studies led me to an interest in stigmatisation and how the effects of this on family life are often shaped by the mode of inheritance of the condition.

With others - Heather Skirton, Clara Gaff, Marion McAllister and Nicki Taverner - I have worked to promote a model of genetic counselling practice and to develop the Cardiff University MSc course in Genetic (and Genomic) Counselling.

Clinical genetics trainees today face different circumstances from the 1980s but i would encourage them to follow the new paths of engagement with disease support groups, the genetic counselling process, and the ethical and social issues that emerge in the context of human genetics research and service provision. Be enthusiastic in developing treatments, where feasible, but be scrupulously honest about the limitations of what is achievable. Avoid hype, raising expectations beyond what is realistic, to avoid the destructiveness of dashed hopes.

PL2.1 Advancing biology through single cell phenomics

Bart Deplancke

In my presentation, I will discuss recent advancements from our laboratory in single cell phenomic analysis technologies. These include: 1) IRIS for ‘Interconnecting a Robotic Image of a cell to its ScRNA-seq profile’. This is a novel, so far unpublished technology that pairs the transcriptomic fingerprint of a single cell with its high-resolution image. I will present proof-of-concept examples showhing how IRIS provides an unprecedented opportunity to i) leverage machine learning approaches to uncover the molecular determinants underlying single cell phenotypes and vice versa, contributing to an integrated view of cellular function and structure, and ii) transform clinical practices through paired single cell imaging and transcriptomics; 2) Live-seq: a methodology for extracting single cell transcriptomes without compromising cell viability, allowing temporal analyses of cellular responses and bridging the gap between transcriptomic profiling and functional behavior over time; and 3) (if time permits) scTF-seq: an approach to quantify transcriptomic changes based on transcription factor dose in single cells, providing insights into cell fate determination and the underlying heterogeneity in cellular reprogramming. These breakthroughs significantly enhance our ability to unravel complex biological processes at the single-cell level with precision and temporal resolution. As such, they may have broad implications for gene regulation, cellular reprogramming, and developmental biology, opening new avenues for research and therapeutic strategies.

PL3.1 Zero Childhood Cancer National Precision Medicine Program: Improving outcomes for children with high risk cancer cancer utilising comprehensive, integrated multiomic profiling

Vanessa Tyrrell ^1;2, Chelsea Mayoh^2;3, Paul Ekert^2;3;4, Mark Cowley^2;3, David Ziegler^2;3;5, Marie Wong^2;3, Paulette Barahona², Mark Pinese^2;3, Michelle Haber^2;3, Rob Salomon², Eliza Courtney^2;5, Richard Lock^2;3, Noemi Fuentes-Bolanos^2;3;5, Glenn Marshall^2;3;5, Meera Warby⁵, Toby Trahair^2;3;5, Dong Anh Khuong Quang^6;7, Claire Wakefield³, Katherine Tucker^3;5;8, Kristine Barlow-Stewart^2;9, Bhavna Padhye¹⁰, Loretta Lau^2;3;5

¹UNSW Sydney, School of Clinical Medicine, UNSW Medicine and Health, Sydney, Australia; ²Children’s Cancer Institute, Randwick, Australia; ³UNSW Sydney, School of Clinical Medicine, UNSW Medicine & Health, Sydney, Australia; ⁴Peter MacCallum Cancer Centre, Melbourne, Australia; ⁵Sydney Children’s Hospital, Randwick, Kids Cancer Centre, Randwick, Australia; ⁶The Royal Children’s Hospital Melbourne, Children’s Cancer Centre, Parkville, Australia; ⁷Murdoch Children’s Research Institute, Parkville, Australia; ⁸Prince Of Wales Hospital, Hereditary Cancer Clinic, Randwick, Australia; ⁹University of Sydney School of Medicine, Camperdown, Australia; ¹⁰The Children’s Hospital at Westmead, Cancer Centre for Children, Westmead, Australia

Consortium: Zero Childhood Cancer National Precision Medicine Program

Background/Objectives: Precision medicine is revolutionising the way cancers are detected, characterised, and treated. The Zero Childhood Cancer program (ZERO) represents one of the world’s largest and most comprehensive high-risk (<30% survival) paediatric cancer cohorts, employing integrated multiomic profiling to identify precision guided treatment (PGT) approaches, to improve outcomes and minimise long-term side effects in survivors.

Methods: ZERO performs integrated somatic whole genome, whole transcriptome and methylation profiling, and germline whole genome analyses, to identify genomic aberrations, and determine potential genomically guided treatment options for each child. These are discussed at a national multidisciplinary tumour board, and a clinical research report is issued to treating clinicians.

Results: Since September 2017, over 1500 children with cancer have been enrolled. In the first 250 patients with high-risk cancer, the molecular basis for the cancer was identified in 94% of patients, with 70% being potentially actionable (Wong et al, Nature Medicine, 2020). Recent analyses of 384 consecutively enrolled high-risk child cancer patients, with >18mths of follow up, demonstrated that those receiving a precision-guided treatment, informed by comprehensive molecular profiling, led to improved two year progression free survival when compared to unguided novel therapies or standard-of-care cytotoxic therapies (27% vs 11%).

Conclusion: These findings demonstrate the power of multiomic profiling to improve outcomes for children with high-risk cancer. ZERO is now open to all Australian children diagnosed with cancer, aiming to demonstrate utility of this multiomic precision medicine model, irrespective of type, stage, or risk.

Grants: Medical Research Future Fund (Australia), Minderoo Foundation

Conflict of Interest: None declared

PL3.2 All by All of Us: common and rare variant association testing in 250,000 whole genomes across diverse ancestry groups

Konrad Karczewski ^1;2, Wenhan Lu², Robert Carroll³, Ying Wang^1;2, Riley Grant², Matthew Solomonson², Megan He³, Michael Lyons³, Wei Zhou^1;2, Anjene Musick⁴, Daniel King², Alicia Martin^1;2, Dan Roden³, Benjamin Neale^1;2

¹Massachusetts General Hospital, Boston, United States; ²Broad Institute, Cambridge, United States; ³Vanderbilt University Medical Center, Nashville, United States; ⁴National Institutes of Health, Bethesda, United States

Background/Objectives: Whole genome sequencing of population biobanks provide opportunities for comprehensive phenome- and genome-wide (“all x all”) association analyses, including rare burden and common variant testing of coding and non-coding variation, across many human phenotypes. Additionally, the increased diversity of the All of Us cohort, in which 80% of the 245,400 participants are underrepresented in genomic research, enhances gene discovery, and multi-phenotype analysis identifies pleiotropy relevant to human disease.

Methods: We built a comprehensive multi-ancestry phenome- and genome-wide analysis framework, using a generalized mixed model framework to perform common variant association tests and rare variant burden tests, followed by meta-analysis.

Results: We applied this framework to 3,348 anthropometric, biomarker, disease, and medication phenotypes derived from electronic health records and surveys across 6 genetic ancestry groups for each quantitative trait and binary trait with over 200 cases. We find thousands of associations for rare variant burden analysis and common variant associations enabled by this diverse cohort. We show consistency with previous efforts such as UK Biobank, and demonstrate how multi-ancestry approaches improve gene discovery.

Conclusion: We present a roadmap for iterative releases of the data up to 1 million individuals that will empower more well-powered association tests for diseases with less than 1% prevalence. We release the full dataset of summary statistics to the public, via bulk data downloads as well as an interactive public-facing browser.

Grants: NIH grants OT2OD035404 and OT2OD002750

Conflict of Interest: Konrad Karczewski Tome, Vor, Nurture Genomics, Wenhan Lu: None declared, Robert Carroll: None declared, Ying Wang: None declared, Riley Grant: None declared, Matthew Solomonson: None declared, Megan He: None declared, Michael Lyons: None declared, Wei Zhou: None declared, Anjene Musick: None declared, Daniel King: None declared, Alicia Martin: None declared, Dan Roden: None declared, Benjamin Neale Deep Genomics, Neumora

PL3.3 A phenome-wide association study of methylated GC-rich repeats identifies a GCC repeat expansion in AFF3 as a significant cause of intellectual disability

Bharati Jadhav¹, Paras Garg¹, Joke van Vugt², Kristina Ibañez³, Delia Gagliardi^3;4, William Lee¹, Mariya Shadrina¹, Tom Mokveld⁵, Egor Dolzhenko⁵, Alejandro Martin Trujillo¹, Scott Gies¹, Mafalda Barbosa¹, Miten Jain⁶, Henry Houlden⁴, Benedict Paten⁶, Jan Veldink², Arianna Tucci³, Andrew Sharp ¹

¹Icahn School of Medicine at Mount Sinai, Department of Genetics and Genomic Sciences and Mindich Child Health and Development Institute, New York, United States; ²Utrecht University, Department of Neurology, UMC Utrecht Brain Center, Utrecht, Netherlands; ³Queen Mary University of London, London, United Kingdom; ⁴University College London, Department of Neuromuscular Diseases, London, United Kingdom; ⁵Pacific Biosciences, Menlo Park, United States; ⁶University of California, Santa Cruz, United States

Consortium: Genomics England Research Consortium, Project MinE ALS Sequencing Consortium

GC-rich tandem repeat expansions (TREs) are often associated with DNA hypermethylation, gene silencing, and folate-sensitive fragile sites and underlie several congenital and late-onset disorders, including fragile X and ALS. Through a combination of DNA methylation profiling and tandem repeat genotyping, we identified 24 methylated TREs. We investigated their effects on human traits using PheWAS in 168,641 individuals from the UK Biobank, identifying 156 significant TRE:trait associations (10% FDR) involving 17 TREs. Of these, a GCC expansion in the promoter of AFF3 was linked with a 2.4-fold reduced probability of completing secondary education, an effect size comparable to several recurrent pathogenic microdeletions. Given its strong negative influence on educational attainment, we hypothesized that this AFF3 expansion might represent the pathogenic mutation in some patients with neurodevelopmental disorders. In support of this, in a cohort of 6,371 probands with neurodevelopmental problems of unknown etiology from the 100,000 Genomes Project, we observed a significant enrichment of AFF3 expansions compared to controls. Finally, we identified SNVs tagging AFF3 expansion; many of these are the same SNVs reported at this locus in prior GWAS for educational attainment. Our observations, therefore, indicate that a subset of GWAS signals in the genome are likely driven by underlying TREs that preferentially occur on specific founder haplotypes. With a population prevalence that is at least 5-fold higher than the TRE causing fragile X syndrome and 3- to 8-fold higher than recurrent microdeletions of 1q21.1, 15q13.3, and 16p11.2, AFF3 expansions therefore represent a significant cause of neurodevelopmental delay.

Conflict of Interest: None declared

PL3.4 In-silico prioritization of open chromatin sequences increases the detection of activating variants

Kilian Salomon ¹, Max Schubach¹, Mohan Dash¹, Chengyu Deng², Nadav Ahituv², Jay Shendure³, Martin Kircher¹

¹Berlin Institute of Health at Charité (BIH), Computational Genome Biology, Berlin, Germany; ²University of California San Francisco, Department of Bioengineering and Therapeutic Sciences, San Francisco, United States; ³University of Washington, Department of Genome Sciences, Seattle, United States

Consortium: NIH Impact of Genomic Variation on Function consortium (IGVF)

Background/Objectives: A substantial reservoir of variants associated with pathological phenotypes resides in non-coding sequences, especially proximal and distal regulatory sequences. As part of the NIH Impact of Genomic Variation on Function (IGVF) consortium, we investigate functional consequences of variation using Massively Parallel Reporter Assays (MPRAs).

Methods: For this purpose, we tested over 28,000 candidate cis-regulatory regions (cCREs) from the proximity (50kb) of 526 genes associated with neurological, cardiac or clinically actionable diseases as well as a random gene set. Within these cCREs, we included >46,000 variants from gnomAD with an allele frequency (AF) of at least 1% and samples of lower frequency variants. Specifically, we ranked sequences based on the Enformer sequence model (Avsec Ž et al. 2021) to select 70% potentially activating, 15% repressing, and 15% random variants.

Results: We measured variant and cCRE expression in neural progenitor cells derived from WTC11. In an initial analysis, we observed 150 significant variant effects, including 81 activating effects. This enrichment of expression-activating variants is only present in the computationally prioritized set, but not in variants above the AF > 1% threshold.

Conclusion: Previous work of saturation mutagenesis MPRAs showed that it is more likely to observe variants with negative effects (Kircher M et al. 2019). Our large set of significant activating variant effects implies that Enformer prioritized these effects successfully. Further investigations are required to test the robustness of in-silico tools to support MPRA designs or whether rare variants more frequently introduce activating binding sites.

Grants: IGVF (1UM1HG011966)

Conflict of Interest: None declared

PL3.5 Understanding the genetic complexity of puberty timing across the allele frequency spectrum.

Katherine Kentistou¹

¹MRC Epidemiology Unit, University of Cambridge, School of Clinical Medicine, Cambridge, United Kingdom

Consortium: the ReproGen Consortium

Background/Objectives: Pubertal onset is highly polygenic and has been linked to later life disease, including obesity, type 2 diabetes, and hormone-sensitive cancers.

Methods: We performed a GWAS meta-analysis of age at menarche including data from 44 cohorts of European and East-Asian ancestry (n ~ 800,000) and rare exome variant associations (n∼220,000). We developed a GWAS causal-gene identification method, based on omics data and enhancer maps.

Results: Our multi-ancestry GWAS identified 1,080 associations with age at menarche. Women at the top/bottom centiles of polygenic risk exhibited a ∼11/∼14-fold higher risk of delayed/precocious puberty. We identified 665 high-confidence menarche genes supported by multiple functional predictors. These were enriched for genes underlying the embryonic migration of GnRH neurons in RNAseq data. We identified an uncharacterized G-protein coupled receptor, which we demonstrate can amplify canonical MC3R signaling in vitro and genetically. Using data from a longitudinal growth cohort we showed that 45% of the menarche signals exert a primary effect on childhood growth. We identified 6 genes harbouring rare loss-of-function variants that associate with menarche, three of which were novel; KDM4C, PDE10A, and ZNF483. The impact of polygenic risk was abolished by rare variants in ZNF483, which affected the protein’s zinc finger domains and thus its ability to correctly bind to its downstream targets.

Conclusion: These findings extend our understanding of the biological complexity of pubertal timing, highlight body size dependent and independent mechanisms, and demonstrate how common genetic variants can influence the clinical extremes of pubertal development.

Grants: Medical Research Council, MC_UU_12015/2.

Conflict of Interest: None declared

PL3.6 Decoding ALG13-CDG: multi-omics profiling of ALG13-CDG brain organoids reveals distinct glycosylation defects impacting axon growth, neuronal migration, synaptic plasticity, and seizure susceptibility.

Rameen Shah¹, Silvia Radenkovic², Rohit Budhraja³, Graeme Preston¹, Akhilesh Pandey³, Steven Sloan⁴, Eva Morava¹, Tamas Kozicz ¹

¹, Genetics and Genomic Sciences, New York, United States; ², Clinical Genomics, Rochester, United States; ³, Department of Laboratory Medicine and Pathology, Rochester, United States; ⁴, Human Genetics, Atlanta, United States

Background/Objectives: Asparagine-linked glycosylation 13 (ALG13) is an X-linked congenital disorder of glycosylation (CDG). Interestingly, unlike most other N-linked CDGs, investigations on ALG13-CDG patients’ blood and fibroblast samples have not revealed defects in glycosylation. ALG13-CDG presents with neurological symptoms such as seizures, intellectual disability, central hypotonia, and developmental delays. These observations have led us to hypothesize that ALG13-CDG might primarily affect the brain.

Methods: To test this hypothesis, we developed a brain model of ALG13-CDG by generating induced pluripotent stem cells from fibroblasts and differentiating them into cortical brain organoids (BOs). At 77-92 days of differentiation, we conducted comprehensive multiomics profiling on these BOs, including single-cell RNA-seq, proteomics, glycoproteomic, and metabolomics analyses.

Results: Our investigation unveiled a significant reduction in the glycosylation of several proteins crucial for axon growth, neuronal migration, and synaptic plasticity. We observed a parallel reduction in glycosylation for proteins implicated in seizures due to lipid accumulation and disturbances in calcium transport. Our proteomics studies revealed alterations in the expression of genes associated with neuronal migration and nucleotide synthesis. By employing single-cell RNA-seq, we elucidated distinct changes in gene expression among GABAergic neurons, glutamatergic neurons, and radial glia, with significant changes in genes important for neuronal migration, axon guidance, calcium ion concentration, and lipid metabolism. Our metabolomics investigations highlighted an elevation in GlcNAc levels and downregulation in metabolites associated with nucleotide synthesis.

Conclusion: These multi-omics studies on ALG13-deficient BOs provide valuable insights into specific brain-related disturbances inherent to ALG13-CDG and offer promising avenues for potential therapeutic interventions.

Conflict of Interest: None declared

C01.1 Identification of microbiota components correlated with host lifestyle, molecular, biochemical, immunophenotypic measurements and genotype in a deeply phenotyped Sardinian cohort

Maria Antonietta Diana ^1;2, Maristella Pitzalis¹, Marco Masala¹, Andrea Maschio¹, Fabio Busonero¹, Monia Lobina¹, Marcella Devoto^1;3, Myriam Gorospe⁴, David Schlessinger⁴, Valeria Orrù¹, Edoardo Fiorillo¹, Federico Santoni^1;5, Francesco Cucca^1;2, Mauro Pala¹

¹C.N.R. - Istituto di Ricerca Genetica e Biomedica, Monserrato, Italy; ²University of Sassari, Scienze Biomediche, Sassari, Italy; ³Sapienza University of Rome, Medicina Traslazione e di Precisione, Roma, Italy; ⁴National Institute On Aging, Baltimore, United States; ⁵Lausanne University Hospital, Lausanne, Switzerland

Background/Objectives: Existing literature already explored the association between microbiota and host traits, but suffers from limitations such as small sample sizes, few covariates and reliance on non-whole-genome methods. We aim to overcome these by analyzing metagenomic data in a large cohort. Our goal is to correlate microbiota with diverse host measurements, including genetics, lifestyle, and disease markers.

Methods: We leverage whole-genome shotgun sequencing of fecal microbiota from 2,700 ProgeNIA cohort volunteers where high-resolution genotypes and phenotypes were available from previous projects. Measured phenotypes include blood markers, anthropometric traits, disease status, lifestyle factors (e.g., alcohol consumption, coffee intake, smoking), immunophenotypic data, leukocyte transcriptional profiling, and high-resolution genetic characterization.

Results: We have identified more than 100 taxonomic levels displaying significant correlations with factors such as alcohol consumption, smoking, Proton Pump Inhibitors (PPI), BMI, anti-diabetics intake, cancer, diabetes, glycemia, sex and age. We conducted a Quantitative Trait Loci (QTL) GWAS on 5566 taxa using REGENIE obtaining 330 taxa with at least one significant association with the phenotypic traits. 54 of those associations colocalize with sentinel variants linked to other traits and diseases such as Lymphocyte percentage of white cells, Diastolic blood pressure, Gamma glutamyl transferase levels, Asthma and Lung adenocarcinoma among others. Causality has been inspected with one and two-steps Mendelian randomization methods.

Conclusion: Our study, empowered by a significant sample size and broad phenotyping, enabled the discovery of novel causal interactions between specific taxa and several factors increasing our knowledge about the mutual influence of microbiota on human health and disease.

Conflict of Interest: None declared

C01.2 Multi-trait genetic colocalization of urine protein levels provide insights into renal protein handling and related clinical outcomes.

Oleg Borisov ¹, Stefan Haug¹, Nora Scherer¹, Sara Monteiro-Martins¹, Yong Li¹, Anna Köttgen¹

¹Institute of Genetic Epidemiology, Medical Center – University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg, Germany

Background/Objectives: Genetic association studies of protein levels provide a valuable link between molecular processes and human phenotypes including diseases. We generated urine proteomics data and integrated it with multiomics datasets and phenome-wide association studies to elucidate genetic mechanisms of protein handling in the kidney and its connection to diseases.

Methods: We performed genetic colocalization analyses (“coloc” package) to assess whether urine protein levels measured in the German Chronic Kidney Disease study (N > 1000) share common genetic architecture with orthogonal omics and clinical datasets. We colocalized urine protein genetic associations with multi-tissue transcriptomics, plasma proteomics, and clinical outcome datasets (UK Biobank and FinnGen).

Results: We performed 3.5 million colocalization tests in more than 100 genetic loci and identified 9,000 pairs of traits with strong colocalization evidence (posterior probability H4 > 80%). The majority of loci (>90%) showed evidence of multi-trait colocalization (up to 1000 colocalizing traits per protein). One example was Prostate Stem Cell Antigen (PSCA) – where cis associations (P < 1e-300) with urine PSCA levels colocalized with PSCA expression in kidney tissues and urogenital clinical manifestations, including bladder cancer and urinary tract infections.

Conclusion: We report colocalization results between urine protein associations and phenome-wide association studies including molecular and clinical traits. The generated resource will facilitate a deeper understanding of genetic factors influencing protein handling by the kidney and will help to reveal circulating disease biomarkers which are also informative when quantified in urine.

Grants: This work was supported by the German Research Foundation (DFG) project ID 431984000 (SFB 1453).

Conflict of Interest: None declared

C01.3 Single-cell eQTL mapping in healthy and IBD-afflicted terminal ileum offers novel insights into disease susceptibility

Tobi Alegbe ¹, Monika Krzak¹, Mennatallah Ghouraba¹, Michelle Strickland¹, Marcus Tutert¹, Bradley Harris¹, Matiss Ozols¹, Lucia Ramirez-Navarro¹, Jason Skelton¹, Jasmin Ostermayer¹, May Xueqi Hu¹, Noor Wana¹, Vivek Iyer¹, Cristina Cotobal-Martin¹, Miles Parkes², Gareth-Rhys Jones³, Chris Wallace⁴, Rebecca McIntyre¹, Tim Raine², Carl Anderson¹

¹Wellcome Sanger Institute, Hinxton, United Kingdom; ²Addenbrooke’s Hospital, United Kingdom; ³University of Edinburgh, United Kingdom; ⁴MRC Biostatistics Unit, Cambridge, United Kingdom

Background/Objectives: Inflammatory bowel diseases (IBD) are complex polygenic diseases of the gastrointestinal tract with over 320 genetic susceptibility loci associated. For many loci, the associated variants are non-coding making the causal genes and cell types unclear. Expression quantitative trait loci (eQTL) can help link genes to GWAS variants. However, existing efforts have been hampered by the limited resolution afforded by bulk RNA sequencing.

Methods: We generated single-cell RNA-sequencing data from terminal ileal biopsies of 152 non-IBD and 88 Crohn’s-afflicted individuals. Our dataset consisted of over 1 million single-cell transcriptomes, characterising 49 cell types. We performed pseudobulk sc-eQTL mapping finding 6,853 eGenes (genes with an eQTL, FDR < 0.05), 44% of which were undetectable without cell type annotation.

Results: We reproduced eQTL effects from similar studies (GTEX, CEDAR, and OneK1K) but also 15% of our eGenes were only present in our dataset. We were able to map sc-eQTL interactions with our phenotypes such as inflammation finding 253 eGenes with an interaction effect. Colocalisation with CD, UC and IBD GWAS resulted in 22%, 20% and 17% of GWAS loci, respectively, colocalising (PPH4 > 0.75). Amongst these was a novel colocalisation, WRAP73 (PPH4 = 0.97), which has previously been shown to be differentially methylated in IBD.

Conclusion: We demonstrate the value of mapping eQTLs with tissue-derived single-cell data. As our sample size grows, we will gain more insight into IBD biology bringing us a step closer to novel, genetically motivated therapeutics.

Grants: Wellcome Trust [206194 and 108413/A/15/D], The Crohn’s Colitis Foundation [612986 and 997266] and Open Targets [OTAR2057].

Conflict of Interest: Tobi Alegbe: None declared, Monika Krzak: None declared, Mennatallah Ghouraba: None declared, Michelle Strickland: None declared, Marcus Tutert: None declared, Bradley Harris: None declared, Matiss Ozols: None declared, Lucia Ramirez-Navarro: None declared, Jason Skelton: None declared, Jasmin Ostermayer: None declared, May Xueqi Hu: None declared, Noor Wana: None declared, Vivek Iyer: None declared, Cristina Cotobal-Martin: None declared, Miles Parkes: None declared, Gareth-Rhys Jones: None declared, Chris Wallace Part-time employee for GlaxoSmithKline, Rebecca McIntyre Full-time employee of Relation Therapeutics, Tim Raine Has received research/educational grants and/or speaker/consultation fees from Abbvie, Arena, Aslan, AstraZeneca, Boehringer-Ingelheim, BMS, Celgene, Ferring, Galapagos, Gilead, GSK, Heptares, LabGenius, Janssen, Mylan, MSD, Novartis, Pfizer, Sandoz, Takeda and UCB, Carl Anderson Has received consultancy or speaker fees from Genomics plc, BridgeBio and GSK

C01.4 C-STEM: a fast and powerful method for robust trans-eQTL mapping in multi context studies

Lena Krockenberger ¹, Mike Thompson², Noah Zaitlen^3;4;5, Xuanyao Liu⁶, Brunilda Balliu^7;8

¹University of California, Los Angeles, Bioinformatics Interdepartmental Graduate Program, Los Angeles, United States; ²Centre for Genomic Regulation (CRG), Systems Biology Department, Barcelona, Spain; ³University of California, Los Angeles, Department of Computer Science, Samueli School of Engineering, Los Angeles, United States; ⁴University of California, Los Angeles, Department of Neurology, David Geffen School of Medicine, Los Angeles, United States; ⁵University of California, Los Angeles, Department of Computational Medicine, David Geffen School of Medicine, Los Angeles, United States; ⁶The University of Chicago, Section of Genetic Medicine, Chicago, United States; ⁷University of California, Los Angeles, Department of Pathology and Laboratory Medicine, David Geffen School of Medicine, Los Angeles, United States; ⁸University of California, Los Angeles, Department of Biostatistics, Fielding School of Public Health, Los Angeles, United States

Background/Objectives: Most disease-associated genetic variants lie outside exons, fueling extensive research into genetic mechanisms governing transcriptional regulation. While progress has been made in identifying cis effects on gene expression, detecting high-quality distal associations (trans-eQTLs) remains challenging. Existing methods for trans-eQTL mapping overlook the inherent intra-individual correlation of gene expression found in multi context studies with repeated sampling. These oversights can significantly diminish the power to detect trans-eQTLs.

Methods: We introduce C-STEM, a fast and powerful method for multi-context trans-eQTL mapping. C-STEM accounts for intra-individual correlation by decomposing gene expression into shared and context-specific components, builds cross-validated cis-genetic predictors (CVGP) for each component, and constructs a final predictor using both components. C-STEM tests for association between all CVGP-trans gene pairs, reducing the number of tests and enhancing power. We utilize hierarchical testing to control FDR across and within contexts, boosting power when significant CVGP-trans gene pairs exist in multiple contexts.

Results: Simulations demonstrate C-STEM’s increased power in detecting trans-gene regulation compared to other methods. Applied to bulk RNA-seq data from the GTEx consortium (N = 948) and single-cell RNA-Seq data from the CLUES (N = 234) and OneK1K (N = 937) cohorts, C-STEM identifies 89% of trans-eQTLs mapped by existing approaches, while providing a 65% increase in the number of trans-gene regulations identified. Existing approaches overestimate context specificity of trans-eQTL effects; 12% of trans-eQTLs appear unique to a single context using existing methods, compared to only 6% using C-STEM.

Conclusion: C-STEM enables construction of multi-context trans-eQTL maps, enhancing understanding of cross-context gene regulatory networks underlying complex human traits.

Grants: NIH U01HG012079

Conflict of Interest: None declared

C01.5 ProteoVAE: guided disentangled Variational Autoencoder for the multivariate analysis of the blood plasma proteome

Michela Carlotta Massi ¹, Nathaniel Robert Nethercott², Solène Cadiou¹, Francesca Ieva^1;2, Emanuele Di Angelantonio^1;3

¹Human Technopole, Italy; ²Politecnico di Milano, Milano, Italy; ³University of Cambridge, United Kingdom

Background/Objectives: Understanding the genetic basis of variation in plasma protein levels, essential for human biology and drug targeting, remains a challenge due to the complexity of the plasma proteome and limitations in existing pQTL studies, which fail to capture shared regulatory patterns and face a high multiplicity burden. This study introduces a novel Deep Learning-based method to perform multivariate association tests, i.e. latentQTL analyses, to explore the genetic determinants of protein level variability through interdependent protein sets.

Methods: We present proteoVAE, a disentangled Variational AutoEncoder (VAE) that maps protein level profiles into lower-dimensional spaces, where each dimension represents an independent latent factor driving variation in the plasma proteome, devoid of technical confounders. We also propose a novel method to associate protein sets to each dimension and study factors’ functional meaning; Finally, we exploit these latent factors to perform Genome-Wide Association tests, uncovering latentQTLs.

Results: Applied to 54,219 UKBiobank participants’ plasma proteomic profiles from The Pharma Proteomics Project, proteoVAE efficiently mapped 2,923 protein levels into low-dimensional latent spaces (from 10 to 128 dimensions). We associate protein-sets to each latent factor, revealing significant enrichment in intra-group protein correlation. Notably, each protein-set represented meaningful and disentangled biological pathways. latentQTL results replicated known pQTL signals, while uncovering new associations with the identified pathways.

Conclusion: proteoVAE offers an innovative, interpretable framework for linking genetic factors to functionally coherent protein groups, simplifying the genetic architecture of the plasma proteome, reducing the burden of multiple testing and identifying new therapeutic targets and drug repurposing opportunities.

Conflict of Interest: None declared

C01.6 Genetic prediction of UK Biobank proteomics data for the OmicsPred portal

Douglas Loesch ¹, Yu Xu², Henry Taylor^2;3, Samuel Lambert², Scott C. Ritchie², Xilin Jiang², Carles Foguet Coll², Benjamin B. Sun⁴, Heiko Runz⁴, Christopher D. Whelan⁵, Slavé Petrovski¹, Abhishek Nag¹, Dirk Paul¹, Michael Inouye²

¹AstraZeneca, Centre for Genomics Research, Cambridge, United Kingdom; ²University of Cambridge, British Heart Foundation Cardiovascular Epidemiology Unit, Department of Public Health and Primary Care, Cambridge, United Kingdom; ³National Human Genome Research Institute (NHGRI), National Center for Precision Health Research, Bethesda, United States; ⁴Biogen, Cambridge, United States; ⁵Janssen Pharmaceuticals, Boston, United States

Background/Objectives: Multi-omic analyses are a powerful tool for uncovering the molecular underpinnings of disease. Genetic prediction of omics data can extend omics information into biobank-scale studies where financial restraints have typically imposed significant hurdles.

Methods: We have expanded the OmicsPred portal, a repository of omics genetic scores, to include proteomics data from the UK Biobank (N = 54,219), dramatically increasing both the sample size and the number of Olink-characterized proteins over currently available scores. We trained genetic scores for proteins using Bayesian ridge regression and imputed genotype data on a training set of 34,557 participants, followed by testing the scores on the remaining participants. We then performed phenome-wide association studies (PheWAS) with the genetic scores using Cox regression and the time to diagnosis.

Results: Out of the 2,612 models we trained, 1,712 had an R² of at least 0.01 in the withheld set and 166 had an R² of at least 0.3. To evaluate score portability, we stratified the withheld set by genetically predicted ancestry and compared the variance explained (R²) in the European ancestry subset with four other ancestry groups using the slope of the regression line (0.88, 087, 0.71, and 0.44 for Admixed American, South Asian, East Asian, and African, respectively).

Conclusion: To demonstrate the utility of our scores, we performed PheWAS in the UK Biobank and All of Us, encompassing over 900,000 participants from diverse backgrounds. Other potential applications include causal inference and pathway-aware genetic risk prediction. Genetic scores will be made available on the OmicsPred portal (https://www.omicspred.org/).

Grants:

Conflict of Interest: Douglas Loesch Employee of AstraZeneca, Yu Xu: None declared, Henry Taylor: None declared, Samuel Lambert: None declared, Scott C. Ritchie: None declared, Xilin Jiang: None declared, Carles Foguet Coll: None declared, Benjamin B. Sun May own stock or interests in Biogen., Employee of Biogen., Heiko Runz Holds stock in Biogen, Employee of Biogen, Christopher D. Whelan May hold stock in Janssen Pharmaceuticals or Johnson and Johnson, Employee of Janssen Pharmaceuticals, a Johnson and Johnson company, Slavé Petrovski May own stocks or interests in AstraZeneca, Employee of AstraZeneca, Abhishek Nag May own AstraZeneca stocks or interests., Employee of AstraZeneca, Dirk Paul May own stock or interests in AstraZeneca, Employee of AstraZeneca, Michael Inouye Research collaborations with Nightingale Health and Pfizer., Grants include British Heart Foundation (RG/18/13/33946), NIHR Cambridge Biomedical Research Centre (BRC-1215-20014; NIHR203312), Cambridge BHF Centre of Research Excellence (RE/18/1/34212), BHF Chair Award (CH/12/2/29428) and the Health Data Research UK (Molecules to Health Records programme), Trustee of the Public Heath Genomics Foundation, member of the Scientific Advisory Board of Open Targets

C02.1 Mapping the prevalence and origin of (mosaic) chromosomal abnormalities in human blastocysts: findings and potential value for PGT

Machteld Baetens ¹, Lisa De Witte², Kelly Tilleman³, Frauke Vanden Meerschaut³, Sandra Janssens¹, Ariane Van Tongerloo¹, Virginie Szymczak¹, Dominic Stoop³, Annelies Dheedene¹, Sofie Symoens¹, Björn Menten⁴

¹Ghent University Hospital, Center for Medical Genetics, Ghent; ²Ghent University, Center for Medical Genetics, Ghent, Belgium; ³Ghent University Hospital, Department for Reproductive Medicine, Ghent; ⁴Ghent University Hospital - Ghent University, Center for Medical Genetics, Ghent

Background/Objectives: Diagnosing (mosaic) chromosomal aneuploidy in blastocysts relies mostly on copy number analysis of single trophectoderm (TE) biopsies, with no cell-division information. Whereas meiotic errors generally affect all embryonic cells, mitotic errors lead to embryo mosaicism, linked with healthy live birth, albeit with lower implantation and higher miscarriage rates than euploid embryos. Combining copy number and cell-division origin analysis, mitotic errors could more reliably be identified, resulting in more transferable embryos.

Methods: TE biopsies of 1,479 embryos were processed by GENType (De Witte et al., 2022) for copy number analysis and comprehensive single nucleotide polymorphism (SNP) analysis.

Results: 68% of the biopsies were euploid, others presented with whole-chromosome aneuploidy (22%), segmental aneuploidy (7%) or both (2%). The overall incidence of whole-chromosome mosaicism was 8%. Origin analysis revealed that uniform aneuploidies predominantly occurred from meiotic errors (97%) in the oocyte, with a higher prevalence in acrocentric and smaller chromosomes (P < 0.0001) and advancing female age (P < 0.0001). Conversely, mosaic aneuploidy originated from mitotic errors (92%), on both parental alleles (P > 0.5), distributed across all chromosomes (P > 0.5), with no relation to maternal age (P > 0.5).

These findings demonstrate high concordance (96%) between cell-division origin and copy number. Interestingly, 8% of aneuploidies with an intermediate copy number resulted from a meiotic error, while 3% with a uniform aneuploidy were of mitotic origin.

Conclusion: Our study provides more insight into the underlying mechanisms of (mosaic) chromosomal abnormalities at the blastocyst stage and demonstrates that integrating cell-division origin information improves diagnosing true mosaicism.

Grants: Research Foundation Flanders (FWO; 1S74619N)

Conflict of Interest: None declared

C02.2 Clinical utility of preimplantation genetic testing for aneuploidy origin (PGT-AO) to enhance embryo selection for monogenic disorders: a retrospective pan-country integrative genomic and clinical study

Rick Essers ^1;2, Sajjad Biglari^1;2, Anouk Janssen^1;2, Servi Stevens^1;2, Jeroen Meekels¹, Bart de Koning¹, Jos C.M.F. Dreessen¹, Merryn Macville^1;2, Sonja de Munnik¹, Malou Heijligers¹, Ron van Golde^1;2;3, Marianne L van Buul-van Zwet⁴, Math H. E. Pieters⁴, Koen van Zomeren⁵, Jannie van Echten-Arends⁵, Alwin Derijck^6;7, Sebastiaan Mastenbroek^6;7, Arthur van den Wijngaard^1;2, Edith Coonen^1;2;3, Aimee D.C. Paulussen^1;2, Masoud Zamani Esteki^1;2

¹Maastricht University Medical Center MUMC + , Department of Clinical Genetics, Maastricht, Netherlands; ²GROW Research Institute for Oncology and Reproduction, Department of Genetics and Cell Biology, Maastricht, Netherlands; ³Maastricht University Medical Center MUMC + , Department of Obstetrics and Gynecology, Maastricht, Netherlands; ⁴University Medical Center (UMC) Utrecht, IVF Center, Utrecht, Netherlands; ⁵University Medical Center Groningen, Section of Reproductive Medicine, Department of Obstetrics and Gynecology, Groningen, Netherlands; ⁶Amsterdam UMC location University of Amsterdam, Center for Reproductive Medicine, Amsterdam, Netherlands; ⁷Amsterdam Reproduction and Development Research Institute, Amsterdam, Netherlands

Background/Objectives: Randomized controlled trials revealed that PGT-A does not increase live birth rates. PGT-A overlooks the mosaic nature and segregational origin of aneuploidies, causing potentially viable mosaic embryos to not be transferred. Our novel PGT- A approach, termed PGT-AO, provides a nuanced understanding of early embryonic development in clinical practice and potentially leads to viable mosaic embryo transfer.

Methods: This retrospective study includes 816 embryos from 323 couples counseled by clinical geneticists at University Medical Centers in Maastricht, Amsterdam, Utrecht, and Groningen. We examined 284 embryos with genome haplarithmisis-based PGT-AO to assess if (mosaic) aneuploidies and segregational origin would have altered transfer decisions and subsequent clinical outcomes.

Results: Of transferred embryos, 26.7% (48/180) resulted in live-births. PGT-AO shows that 14.6% (7/48) carried abnormalities, including embryos with mosaic mitotic trisomies (n = 2), mosaic monosomies (n = 2), and mosaic duplications (n = 2) and a deletion (n = 1). Of the transferred embryos, 73.3% (132/180) did not result in live-births, and 36.4% (48/132) contained an aberration. A total of 35 trisomies were found in 41.7% (20/48), including meiotic (n = 9), mitotic (n = 5), meiotic and mitotic origins (n = 1), and 5 were indeterminable.

Conclusion: Our preliminary results show PGT-AO can avoid transferring embryos with aneuploidies of meiotic origin and non-mosaic status, while potentially considering embryos with aneuploidies of mosaic and mitotic origin for transfer. This increases pregnancy rate per embryo transfer and minimizes transfers with poor outcomes, potentially avoiding termination of aneuploid pregnancies.

Grants:

• EVA (KP111513), MUMC+

• Horizon Europe (NESTOR, 101120075), European Commission

• Horizon 2020 Innovation (ERIN, EU952516)

Conflict of Interest: None declared

C02.3 The potential of long-read whole genome sequencing based preimplantation genetic testing

Yan Zhao ¹, Greet Peeters¹, Olga Tšuiko¹, Eftychia Dimitriadou¹, Erika Souche¹, Tatjana Jatsenko¹, Joris Robert Vermeesch¹

¹KU Leuven, Department of Human Genetics, Leuven, Belgium

Background/Objectives: Haplotype linkage analysis is commonly employed to enable preimplantation genetic testing for monogenic disorders (PGT-M). To haplotype, DNA from close relatives need to be accessible. However, close relatives are not always available, and this requirement can delay and increase costs for PGT-M. To overcome the need for relatives, we hypothesized that long-read whole genome sequencing (lrWGS) of parental samples and trophectoderm/single blastomere biopsies could enable genome-wide haplotype-base PGT-M analysis.

Methods: We used cell lines and available sequencing data from the Genome in a Bottle (GIAB) Ashkenazi trio. For the parents, publicly available lrWGS data were downloaded. For the son, multi-cell (10 cells) and single-cell samples were collected and sequenced to simulate trophectoderm and single blastomere biopsies, respectively. Variant calling and phasing was performed for all individuals, with a pedigree-based analysis strategy employed specifically for the son to improve accuracy. PGT-M was simulated by comparing haplotypes of the son with parental haplotypes.

Results: Leveraging phased haplotypes of the trio, parental haplotypes inherited by the son were inferred across ~ 99% of protein-coding gene regions for both multi-cell and single-cell samples. Furthermore, haplotype based PGT-M exhibited a diagnosis accuracy of ~99% for SNV loci within protein-coding gene regions for both multi-cell and single-cell samples.

Conclusion: Our study demonstrates that lrWGS-based PGT provides an alternative to current methods, eliminating the reliance on family members for phasing.

Grants: Marie Skłodowska-Curie grant agreement no. 813707 (MATER).

Conflict of Interest: None declared

C02.4 Pre-implantation genetic testing by polygenic risk scores selects different embryos across score construction methods with randomness

Shinichi Namba ^1;2;3, Masato Akiyama⁴, Haruka Hamanoue⁵, Kazuto Kato², Minae Kawashima⁶, Itaru Kushima⁷, Koichi Matsuda¹, Masahiro Nakatochi⁷, Soichi Ogishima⁸, Kyuto Sonehara^1;2;3, Ken Suzuki², Atsushi Takata³, Gen Tamiya^3;8, Chizu Tanikawa¹, Ken Yamamoto², Natsuko Yamamoto², Norio Ozaki⁷, Yukinori Okada^1;2;3

¹The University of Tokyo; ²Osaka University; ³RIKEN Center; ⁴Kyushu University; ⁵Yokohama City University Hospital; ⁶ROID-DS; ⁷Nagoya University; ⁸Tohoku University

Consortium: The BioBank Japan Project

Background/Objectives: Private enterprises offer pre-implantation genetic testing by polygenic risk scores (PRS) to select embryos with ‘desirable’ potential, prior to clinical implementation of PRS for disease prevention. Here, we question the testing’s validity; selected embryos can differ across PRS methods and replicates of the same method.

Methods: Using six popular methods, we constructed PRS for adult height, a well-known target of mating preferences. Using biobank resources, we virtually simulated mating process and generated ten embryo genotypes per mate-pair to evaluate the concordance of selected embryos across methods. We repeated the simulation for type 2 diabetes and saturated summary statistics that we simulated with a sample size larger than any studies to date (N = 1×10⁷).

Results: In any combination of PRS methods, the same embryo was selected as the top-ranked one with only a 30.0% chance for adult height (median; range, 20.4–41.6%). The top-ranked embryo by a particular method was lowest-ranked by at least one other method with a median 5.9% chance (3.4–8.0%), and vice versa (median 5.2% [4.0–7.8%]). This inconsistency was not affected by assortative mating. We observed consistent results for type 2 diabetes and the simulated phenotype, with the latter indicating that the inconsistency cannot be solved in the current or near-future sample sizes. Furthermore, simply repeating the same method produced a different embryo ranking.

Conclusion: The selected embryo substantially relied on the choice of PRS method and the random fluctuation in PRS construction, raising serious ethical concerns about selecting embryos and consequently discarding other embryos in an unreliable way.

Grants: None.

Conflict of Interest: None declared

C02.5 Spatiotemporal transcriptomic atlas of mouse placentation

Yanting Wu^1;2;3;4, Kaizhen Su ⁵, Ying Zhang^6;7, Langchao Liang^6;8;9, Fei Wang⁷, Ling Gao², Qiutong Zheng², Cheng Li², siyue chen², Yunfei Su², Yiting Mao², Simeng Zhu⁵, Chaochao Chai^6;8;9, Qing Lan⁷, Man Zhai⁷, Xin Jin⁷, Jinglan Zhang^2;3;4, Xun Xu^7;10, Yu Zhang², Ya Gao⁷, Hefeng Huang^1;2;3;4;5;11

¹Zhejiang University School of Medicine, Key Laboratory of Reproductive Genetics (Ministry of Education), Hangzhou, China; ²Fudan University, Obstetrics and Gynecology Hospital, Institute of Reproduction and Development, Shanghai, China; ³Chinese Academy of Medical Sciences, Research Units of Embryo Original Diseases, Shanghai, China; ⁴Shanghai Key Laboratory of Reproduction and Development, Shanghai, China; ⁵Shanghai Jiao Tong University, International Peace Maternity and Child Health Hospital, School of Medicine, Shanghai, China; ⁶BGI Research, Wuhan, China; ⁷BGI Research, Shenzhen, China; ⁸BGI Research, Qingdao, China; ⁹University of Chinese Academy of Sciences, College of Life Sciences, Beijing, China; ¹⁰Guangdong Provincial Key Laboratory of Genome Read and Write, Shenzhen, China; ¹¹Zhejiang University School of Medicine, Department of Reproductive Endocrinology, Women’s Hospital, Hangzhou, China

Background/Objectives: The placenta, a transient yet vital organ for gestational support, undergoes intricate morphological and functional transformations throughout gestation. Disruptions in placentation are always associated with adverse perinatal outcomes and birth defects. Establishing a spatiotemporal transcriptome atlas of mouse placentation across different stages is crucial for providing valuable research resources and enhancing our understanding of placentation.

Methods: We collected 13 mouse uterine segments, each containing an embryo and its placenta, at 6 different time points spanning from E7.5 to E14.5. We then performed Stereo-seq and utilized the spatial transcriptome data to generate an atlas for placentation.

Results: Stereo-seq was used to construct a Mouse Placentation Spatiotemporal Transcriptomic Atlas (MPSTA). Focusing on the labyrinth region, spatiotemporal ligand-receptor dynamics unveiled pivotal modulators of trophoblast development and placental angiogenesis. At the invasion front, we identified interface-specific transcription factor regulons involved in gestational maintenance. Our analysis also indicated that a maternal high-fat diet preferentially impacts this interface, exacerbating inflammatory responses and perturbing angiogenic homeostasis. Finally, we discovered a compartmentalized expression pattern of certain imprinted genes, with paternal alleles predominantly expressed in the labyrinth and maternal genes in the interface and decidua. We also observed some of the heart defects-associated genes exhibited higher expression in the placenta than in the foetal heart, especially in syncytiotrophoblasts and glycogen trophoblasts across E10.5-E14.5, providing research clues for the mechanisms of placenta-heart axis.

Conclusion: Our findings furnish a comprehensive, spatially resolved atlas, offering invaluable insights and benchmarks for future explorations into placental development.

Grants: NSFC (82088102); National Key Research and Development Program of China (2021YFC2700701).

Conflict of Interest: Yanting Wu National Key Research and Development Program of China (No. 2021YFC2700701), Full time, Kaizhen Su: None declared, Ying Zhang Full time, Langchao Liang: None declared, Fei Wang: None declared, Ling Gao Full time, Qiutong Zheng: None declared, Cheng Li Full time, siyue chen: None declared, Yunfei Su: None declared, Yiting Mao: None declared, Simeng Zhu: None declared, Chaochao Chai: None declared, Qing Lan Full time, Man Zhai Full time, Xin Jin Full time, Jinglan Zhang Full time, Xun Xu Full time, Yu Zhang Full time, Ya Gao Full time, Hefeng Huang National Natural

Science Foundation of China (No. 82088102), Full time

C02.6 Comparison of seven carrier screening panels: results from a large multicenter study with more than 24000 patients (virtual)

Mónica Fabbro ¹, Filippo Zambelli², Denny Sakkas³, Irene Miguel Escalada⁴, Juanjo Guillen Quilez⁵, Sebastián Menazzi¹, Micaela Galain¹, Eloisa Caviglia¹, Yannina Diaz Cano¹, Sergio Papier¹, Karinna Lattes⁶, Aline Lorenzon⁷, Jose Buratini⁸, Frederikke Lindenberg⁹, Mina Popovic², Cecilia Fernandez¹

¹Cegyr - Eugin Group, Novagen Genetics Laboratory, Buenos Aires, Argentina; ²Eugin Group, Research and Development, Barcelona, Spain; ³Boston IVF Fertility Clinic, IVF Laboratory, Boston, United States; ⁴Clínica Eugin, Research and Development, Barcelona, Spain; ⁵Clínica Eugin, Medical, Barcelona, Spain; ⁶Center for Infertility and Human Reproduction CIRH - Eugin Group, Research and Development, Barcelona, Spain; ⁷Huntington Medicina Reprodutiva - Eugin Group, Research and Development, São Paulo, Brazil; ⁸Biogenesi-Istituti Clinici Zucchi - Eugin Group, Research and Development, Monza, Italy; ⁹Copenhaguen Fertility Center - Eugin Group, Research and Development, Copenhaguen, Denmark

Background/Objectives: Carrier screening (CS) tests have evolved with diverse methodological approaches and variability in gene inclusion. The lack of consensus on the most suitable genetic panel for reproductive medicine presents a challenge in clinical applications. This study aimed to conduct a comprehensive review and comparison of various CS panels utilized by different providers.

Methods: This study included 8 IVF centers from 6 countries. Results from 24038 individuals that were tested using one of 7 different CS panels (designated A to G) were included. Panels were evaluated based on included genes, methodology, carrier rate (CR), and most frequently detected diseases (MFDD). Manual curation of reported MFDD variants was conducted, and panels were compared to the 113 genes recommended by the ACMG.

Results: Panels included a different number of genes: 283(A), 299(B), 302(C,D), 310(E), 484(F), and 546(G). Providers used different methodologies: genotyping by Array (C) or NGS (B,G), full-exon NGS (D,F) or a combination (A,E). Significant (p < 0.05) CR variations were observed: 58.8%(A), 69.4%(B), 44.9%(C), 79.4%(D), 81.7%(E), 79.2%(F), and 88%(G). Observed MFDD and carrier frequencies for common diseases like Cystic fibrosis (CFTR), GJB2 nonsyndromic hearing loss, Congenital adrenal hyperplasia (CYP21A2), Familial mediterranean fever (MEFV), and Spinal muscular atrophy (SMN1) displayed notable variations. Importantly, curation of variants led to significant changes in disease CRs. Panels only included 65 to 79 ACMG genes.

Conclusion: Panel variability affects carriers and disorders detected, impacting the number of at-risk couples to be identified. Curation of reported variants should be considered. Healthcare professionals should take this into consideration when implementing CS.

Grants:

Conflict of Interest: Mónica Fabbro Full-time employee at Novagen, Filippo Zambelli Full-time employee at Eugin Clinic, Denny Sakkas Full-time employee at Boston IVF, Irene Miguel Escalada Full-time employee at Eugin Clinic, Juanjo Guillen Quilez Full-time employee at Eugin Clinic, Sebastián Menazzi Full-time employee at Novagen, Micaela Galain Full-time employee at Novagen, Eloisa Caviglia Full-time employee at Novagen, Yannina Diaz Cano Full-time employee at Novagen, Sergio Papier Shareholder of CEGYR, Karinna Lattes Full-time employee at CIRH, Aline Lorenzon Full-time employee at Huntungton, Jose Buratini Full-time employee at Biogenesi-Istituti Clinici Zucchi - Eugin Group, Frederikke Lindenberg Full-time employee at Copenhaguen Fertility Center - Eugin Group, Mina Popovic Full-time employee at Clínica Eugin, Cecilia Fernandez Full-time employee at Novagen

C03.1 Copy number analysis from genome sequencing data of 11,754 rare disease parent-child trios: a model for identifying autosomal recessive human gene knockouts including a novel gene for autosomal recessive retinopathy

Eric Olinger ^1;2, Ian Wilson³, Sarah Orr², Miguel Barroso-Gil², Ruxandra Neatu², Denize Atan^4;5, John Sayer^2;6

¹Cliniques universitaires Saint-Luc (UCLouvain), Center for Human Genetics, Bruxelles, Belgium; ²Newcastle University, Translational and Clinical Research Institute, Newcastle, United Kingdom; ³Newcastle University, Biosciences Institute, Newcastle, United Kingdom; ⁴Bristol Eye Hospital, Bristol, United Kingdom; ⁵Bristol Medical School, Translational Health Sciences, Bristol, United Kingdom; ⁶Newcastle Upon Tyne Hospitals NHS Foundation Trust, Renal Services, Newcastle

Consortium: Genomics England Research Consortium

Background: In parent-child trios with genome sequencing data, we investigated inherited biallelic deletions to identify known and novel genetic disorders.

Methods: We developed a CNV analysis pipeline based on autosomal genome sequencing read depth of Genomics England 100,000 Genomes Project data from 11,754 parent-child trios and additional 18,875 non-trios. A control cohort of 15,440 cancer patients provided independent deletion frequencies.

Results: Autosomal recessive (AR) modelling detected 34 distinct rare deletions that were homozygous in the proband and heterozygous in both parents. These inherited biallelic deletions were only detected in 52 trios. These “knockout” regions included 37 genes among them 8 with an OMIM AR annotation. Deletions of NPHP1, followed by OTOA, both within segmental duplications, were the only recurrent findings explaining phenotypes in a total of 10 and 3 patients, respectively. Recurrent heterozygous NPHP1 deletions were detected in 0.3%-0.5% of controls. We reviewed “knockout” patients for the remaining 29 genes without disease associations and identified SLC66A1 as a likely novel cause for AR rod-cone dystrophy in 4 families.

Conclusion: A tailored CNV analysis of genome sequencing trio data shows that biallelic inherited gene deletions are rare, with NPHP1 biallelic deletions causing nephronophthisis the leading finding. We propose SLC66A1 as a novel cause for AR retinopathy.

Grants: Kidney Research UK Grants Paed_RP_001_20180925, ST_001_20171120; LifeArc, MRC (MR/Y007808/1); Northern Counties Kidney Research Fund (20/01); Swiss National Science Foundation Grants P2ZHP3_195181 and P500PB_206851

Conflict of Interest: None declared

C03.2 Mechanisms of misexpression-associated structural variants in whole blood

Thomas Vanderstichele ¹, Katie L Burnham¹, Niek de Klein¹, Manuel Tardaguila², Brittany Howell¹, Klaudia Walter¹, Kousik Kundu¹, Wanseon Lee¹, Jonas Koeppel¹, Alex Tokolyi¹, Elodie Persyn³, Artika Nath⁴, Jonathan Marten³, Slavé Petrovski⁵, David Roberts⁶, Emanuele Di Angelantonio³, John Danesh³, Alix Berton⁷, Adam Platt⁸, Adam Butterworth³, Nicole Soranzo², Leopold Parts¹, Michael Inouye³, Dirk Paul⁵, Emma Davenport¹

¹Wellcome Sanger Institute, Hinxton, United Kingdom; ²Human Technopole, Milan, Italy; ³University of Cambridge, British Heart Foundation Cardiovascular Epidemiology Unit, Cambridge, United Kingdom; ⁴Baker Heart and Diabetes Institute, Cambridge Baker Systems Genomics Initiative, Melbourne, Australia; ⁵AstraZeneca, Centre for Genomics Research, Cambridge, United Kingdom; ⁶University of Oxford, Radcliffe Department of Medicine, Oxford, United Kingdom; ⁷AstraZeneca, Translational Science and Experimental Medicine, Gothenburg, Sweden; ⁸AstraZeneca, Translational Science and Experimental Medicine, Cambridge, United Kingdom

Consortium: INTERVAL consortium

Background: Gene misexpression is the unexpected transcription of a gene in a context where it is usually inactive. In this study, we aimed to establish the genetic mechanisms by which gene misexpression occurs in healthy individuals.

Methods: Using 2,804 paired RNA and whole-genome sequencing samples (INTERVAL cohort), we identified 105 rare structural variants (SVs) associated with consistent gene misexpression. To understand the underlying mechanisms we combined STAR-fusion, Salmon, and regulatory annotations, and visualised their effects using Sashimi and coverage plots.

Results: Overall, we identified mechanisms for 42% of misexpression events with a nearby associated SV. First, we identified 12 deletions and 5 tandem duplications associated with highly aberrant read counts and coverage across the intergenic regions upstream of the misexpressed gene. We concluded that these variants resulted in transcriptional readthrough leading to downstream gene misexpression. Interestingly, 2 deletion carriers also showed evidence of intergenic splicing involving the misexpressed genes RTP1 and OTP. Next, we identified 7 deletions and 3 tandem duplications that created fusion transcripts, including a duplication associated with misexpression of myosin heavy chain 1, a gene normally expressed exclusively in skeletal muscle. Finally, we identified a partial gene inversion leading to transcript-specific misexpression of ROPN1B. Surprisingly, we did not observe misexpression occurring via 3D chromatin architecture rearrangements as reported in certain rare diseases.

Conclusion: We have identified putative mechanisms for a subset of misexpression-associated SVs, including transcriptional readthrough, transcript fusions and gene inversion, thus increasing understanding of the impact of rare genetic variants on gene expression.

Grants: BB/V509425/1

Conflict of Interest: Thomas Vanderstichele Has received PhD studentship funding from AstraZeneca, Katie L Burnham: None declared, Niek de Klein: None declared, Manuel Tardaguila: None declared, Brittany Howell: None declared, Klaudia Walter: None declared, Kousik Kundu Stockholder of AstraZeneca, Current employee at AstraZeneca, Wanseon Lee: None declared, Jonas Koeppel: None declared, Alex Tokolyi: None declared, Elodie Persyn: None declared, Artika Nath: None declared, Jonathan Marten Completed this work while employed by the University of Cambridge, but is now an employee of Genomics plc, Slavé Petrovski Stockholder of AstraZeneca, Current employee at AstraZeneca, David Roberts Employee of NHS Blood and Transplant, Emanuele Di Angelantonio: None declared, John Danesh Received multiple grants from academic, charitable and industry sources outside of the submitted work, Scientific advisory boards for AstraZeneca, Novartis, and UK Biobank, Alix Berton Currently an employee of Bayer AG, Research and Early Development Precision Medicine, Research & Development, Pharmaceutical Division, Wuppertal, DE, Adam Platt Stockholder of AstraZeneca, Current employee at AstraZeneca, Adam Butterworth: None declared, Nicole Soranzo: None declared, Leopold Parts: None declared, Michael Inouye Research collaboration with AstraZeneca which is unrelated to this study, Trustee of the Public Health Genomics (PHG) Foundation and a member of the Scientific Advisory Board of Open Targets, Dirk Paul Stockholder of AstraZeneca, Current employee at AstraZeneca, Emma Davenport: None declared

C03.3 Optical genome mapping identifies hidden structural variants in previously undiagnosed rare disease cases in Solve-RD

Kornelia Neveling¹, Ronald van Beek¹, Eveline Kamping¹, Marloes Steehouwer¹, Bart van der Sanden ¹, Lydia Sagath¹, Laura Batlle-Masó², Roger Colobran², Lisenka Vissers¹, Alexander Hoischen^1;3

¹Radboud university medical center, Department of Human Genetics, Research Institute for Medical Innovation, Nijmegen, Netherlands; ²Vall d’Hebron University Hospital (HUVH), Autonomous University of Barcelona (UAB), Vall Hebron Research Institute (VHIR), Barcelona, Spain; ³Radboud university medical center, Department of Internal Medicine; Radboud Expertise Center for Immunodeficiency and Autoinflammation and Radboud Center for Infectious Disease (RCI), Nijmegen, Netherlands

Consortium: Solve-RD DITF-ITHACA, Solve-RD DITF-Euro-NMD, Solve-RD DITF-RND, Solve-RD ERN-RITA, Solve-RD DITF-EpiCARE, SOLVE-RD DATF on behalf of the Solve RD OGM working group and the Solve-RD consortium

Background/Objectives: Optical genome mapping (OGM) allows genome-wide detection of structural variants (SV) as small as 500bp. Here, we systematically applied OGM to identify disease-causing SVs in rare disease (RD) patients that remained undiagnosed after standard-of-care analysis in the respective RD-expert centers.

Methods: 159 individuals from 83 families, with 98 affected individuals have been successfully analyzed using OGM. These included 40 singletons, 11 duos, 26 parent-child trios with expected de novo origin of disease, and 6 other family constellations. The largest groups of samples were selected by clinical experts of ERN-ITHACA (52 families, of which 25 trios), followed by 13 NMD-, 8 RND-, 6 RITA-, and 4 EpiCare- families.

Results: Ongoing data analysis revealed 6 de novo SVs in 26 trios (SV mutation rate: ~1/4 genomes), confirming an expected low SV mutation rate, but also implying every de novo SV as a potential disease candidate. Additionally, we focused on rare (segregating) SVs in singleton cases (30-50/genome). In total, we identified (candidate) disease-causing SVs in >10% of index cases. This includes SVs in well-established disease genes (e.g. SERPING1 in two hereditary angioedema families), as well as rare inversions (n = 1), complex chromosomal aberrations (n = 1) or overlapping non-coding SVs (n = 5) each suggesting novel disease mechanisms. These SVs were usually missed by the initial test modalities or were identified in parallel by Solve-RDs short- and long-read genome sequencing efforts.

Conclusion: This large-scale pan-European study successfully demonstrates the identification of so far hidden SVs in RD by OGM.

Grants: Solve-RD received EU Horizon 2020 funding (779257).

Conflict of Interest: Kornelia Neveling The Solve-RD project has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No. 779257, Ronald van Beek The Solve-RD project has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No. 779257, Eveline Kamping The Solve-RD project has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No. 779257, Marloes Steehouwer The Solve-RD project has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No. 779257, Bart van der Sanden The Solve-RD project has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No. 779257, Lydia Sagath The Solve-RD project has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No. 779257, Laura Batlle-Masó The Solve-RD project has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No. 779257, Roger Colobran The Solve-RD project has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No. 779257, Lisenka Vissers The Solve-RD project has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No. 779257, Alexander Hoischen The Solve-RD project has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No. 779257

C03.4 Unraveling undiagnosed rare disease families by HiFi long-read genome sequencing

Lydia Sagath ¹, Wouter Steyaert¹, german demidov², Nick Zomer¹, Marloes Steehouwer¹, Ronny Derks¹, Marjan Weiss¹, Amber den Ouden¹, Simone Van Den Heuvel¹, Hilde Swinkels¹, Kornelia Neveling¹, Stephan Ossowski², Lisenka Vissers¹, Christian Gilissen¹, Alexander Hoischen^1;3

¹Radboud University Medical Center, Department of Human Genetics, Research Institute for Medical Innovation, Nijmegen, Netherlands; ²Universitätsklinikum Tübingen – Institut für Medizinische Genetik und angewandte Genomik, Tübingen, Germany; ³Radboud University Medical Center, Department of Internal Medicine; Radboud Expertise Center for Immunodeficiency and Autoinflammation and Radboud Center for Infectious Disease (RCI), Nijmegen, Netherlands

Consortium: Solve-RD long-read sequencing working group, Solve-RD DITF-ITHACA, Solve-RD DITF-Euro-NMD, Solve-RD DITF-RND, Solve-RD DITF-EpiCARE, Solve-RD DATF, and the Solve-RD consortium

Background/Objectives: Solve-RD aims to identify the disease-causing genetic variant in previously undiagnosed rare disease (RD) families; we present one of the largest long-read sequencing (LRS) studies to date. We evaluated the effectiveness of LRS in resolving genetic causes in RD families lacking a clear molecular diagnosis despite previous testing.

Methods: We studied 293 individuals from 114 genetically undiagnosed RD families selected by European Rare Disease Networks experts. The 93 families in the ‘unsolved’ cohort were affected by rare neurological, neuromuscular, or epilepsy disorders, and 21 families by clinically recognizable ‘unsolvable syndromes’ for which genetic causes remain unknown. Using 10x coverage HiFi-LRS we detected small variants, SNVs, structural variants (SVs), and short tandem repeats (STRs), with de novo analysis in trios.

Results: Variant analysis in known disease genes resulted in novel genetic diagnoses in ten families with neurological or neuromuscular disorders, involving de novo and inherited SVs, SNVs, and STRs in DMD, NOP56, DAB1, RFC1, SPAST, and TTN. Additionally, we found potential disease-causing SVs in three families (affecting MCF2/FGF13, REEP1, and PSMA3). In an ‘unsolvable syndromes’ patient, we identified a known disease-causing SNV in TUBA1A, re-classifying the patient’s diagnosis.

Conclusion: While we did not identify a common genetic cause in any of the ‘unsolvable syndromes’, we identified causal genetic variants in 10% of families from the ‘unsolved’ cohort, and candidate variants in an additional 5%. Our study shows the potential and effectiveness of even modest-coverage LRS in solving RDs.

Grants: European Union’s Horizon 2020 (779257) and Sigrid Jusélius Stiftelse (220540).

Conflict of Interest: None declared

C03.5 Dosage imbalance of the chromatin remodeler CHD1L contributes to the 1q21.1-CNV associated mirrored neurodevelopmental phenotypes

Marianne Lemée ¹, Maria Nicla Loviglio¹, Tao Ye¹, Céline Keime¹, Peggy Tilly¹, Pernelle Klein¹, Chantal Weber¹, Olivia Wendling¹, Hugues Jacobs¹, Delphine Duteil¹, Juliette D Godin¹, Christophe Romier¹, Christelle Golzio¹

¹I.G.B.M.C. - Institut de génétique et de biologie moléculaire et cellulaire, Illkirch-Graffenstaden, France

Background/objectives: Deletions and duplications of the distal 1q21.1 region are associated with syndromic forms of autism (MIM612474, 612475). Variable phenotypes have been reported, including congenital heart defects, autism, schizophrenia, head circumference and height defects.

Methods: To elucidate which gene(s) are responsible for the 1q21.1 duplication/deletion-associated phenotypes, we performed gene manipulation in zebrafish and mice. We deciphered the function of candidate driver(s) by multi-omics approaches on human iPSC-derived neural progenitor cells (NPC) and cerebral organoids (CO).

Results: We modeled duplication of the 1q21.1 region by overexpressing the 8 genes in the region in zebrafish. We found that only overexpression of CHD1L led to macrocephaly and increased larval body length. Conversely, deletion of the CHD1L zebrafish ortholog resulted in microcephaly and decreased larval body length. These mirror phenotypes are also observed when Chd1l expression is modulated in a mammalian model, notably with a variation in the number of mature Tbr1-positive neurons in the mouse embryo. NPCs derived from control and CHD1L Knock-out isogenic hiPSCs were subjected to transcriptomic, cistromic and chromatin accessibility analyses. These approaches revealed that CHD1L regulates the expression levels and accessibility of genes involved in neuronal differentiation and synaptogenesis, including autism susceptibility genes such as UNC5D and DPP6. Strikingly, absence of CHD1L shifts the cellular identity from forebrain to retinal fate in CO. Finally, pathogenic missense and truncating variants of CHD1L were found in individuals with autism and/or height defects.

Conclusion: CHD1L is a major contributor of the 1q21.1 duplication/deletion-associated microcephaly and growth defects. CHD1L dosage is required during human brain regionalization.

Grants: ANR, FNSNF.

Conflict of Interest: None declared

C03.6 Multiomics and deep phenotyping in MECP2 Duplication Syndrome: Insight into disease severity, expression variability, and nucleic acid therapeutics

Davut Pehlivan ^1;2;3, Jesse Bentsson⁴, Sameer Bajikar^2;3, Christopher M. Grochowski^2;5, Ming Yin Lun⁴, Mira Gandhi², Angad Jolly², Holly Harris⁶, Bernhard Suter¹, Sukru Aras⁷, Melissa Ramocki⁸, Haowei Du², Ellen Wilkey⁴, Cemal Karakas⁹, Jesper Eisfeldt¹⁰, Maria Pettersson¹⁰, Shalini Jhangiani⁵, Lynn Liu¹¹, Marwan Shinawi¹², Virginia Kimonis¹³, Wojciech Wiszniewski¹⁴, Kyle Mckenzie¹⁵, Timo Roser¹⁶, Angela Morgante¹⁷, Alberto Cornier¹⁸, Ahmed Abdelmoity¹⁹, Donna Muzny⁵, Tadahiro Mitani²⁰, Kazuhiro Muramatsu²⁰, Shin Nabatame²¹, Daniel Glaze¹, Jawid Fatih², Richard Gibbs^2;5, Zhandong Liu^1;3, Anna Lindstrand¹⁰, Fritz Sedlazeck⁵, James Lupski^2;5, Huda Yahya Zoghbi^1;2;3, Claudia Carvalho⁴

¹Baylor College of Medicine, Department of Pediatrics Section of Neurology and Developmental Neuroscience, Houston, United States; ²Baylor College of Medicine, Molecular and Human Genetics, Houston, United States; ³Texas Children’s Hospital, Jan and Dan Duncan Neurological Research Institute, Houston, United States; ⁴Pacific Northwest Research Institute, Seattle, United States; ⁵Baylor College of Medicine, Human Genome Sequencing Center, Houston, United States; ⁶Baylor College of Medicine, Developmental and Behavioral Pediatrics, Houston; ⁷1250 Moursund St, Houston, United States; ⁸University Otolaryngology, Greenwich, United States; ⁹University of Louisville, Department of Pediatrics Division of Neurology, Louisville, United States; ¹⁰Karolinska Institute, Department of Molecular Medicine and Surgery, Sweden; ¹¹University of North Carolina, Department of Pediatrics - Neurology, Chapel Hill, United States; ¹²Washington University, Department of Genetics and Genomic Medicine, Town and Country, United States; ¹³University of California Irvine, Department of Pediatrics, Division of Genetics and Genomic Medicine, Orange, United States; ¹⁴Oregon Health Science University, Department of Molecular and Medical Genetics, Portland; ¹⁵University of Alberta, Department of Pediatrics, Alberta, Canada; ¹⁶Ludwig Maximilian University of Munich, Division of Pediatric Neurology, Developmental Medicine and Social Pediatrics, Department of Pediatrics, Munich, Germany; ¹⁷Institute of Biosciences, Department of Genetics and Evolutionary Biology, Sao Paolo, Brazil; ¹⁸San Jorge Children’s Hospital, Department of Genetics, San Juan, Puerto Rico; ¹⁹Children’s Mercy Kansas City, Division of Neurology, Department of Pediatrics, Kansas City, United States; ²⁰Jichi Medical University, Department of Pediatrics, Tochigi, Japan; ²¹Osaka University Graduate School of Medicine, Department of Pediatrics, Osaka, Japan

Background/Objectives: Copy number gains spanning MECP2 cause X-linked intellectual developmental disorder Lubs-type (MRXSL). It is a severe neurodevelopmental disorder with diverse clinical manifestations. Each individual has a unique genomic size of duplication and varying genomic rearrangement structures. Our hypothesis posits that these genomic rearrangement structures contribute to interindividual clinical variability and disease severity.

Methods: We conducted extensive genomic studies including custom-designed aCGH, optical genome mapping and short-read/long-read genome sequencing on 137 MRXSL individuals. We correlated genomic structures with transcriptomic and deep phenotyping analyses including Human Phenotype Ontology (HPO) semantic similarity scores.

Results: CNV sizes ranged from 64 kb to 16 Mb with the following structural distribution: 48% tandem duplications, 22% terminal, 20% Duplication-Triplication-Duplication (DUP-TRP/INV-DUP) structure and 10% other complex genomic structures. De novo events overwhelmingly occurred in the terminal duplication group (65%) compared to the tandem duplication (17%) and CGR groups. RNAseq data from lymphoblastoid cell lines revealed that the MECP2 transcript in MECP2 triplications is statistically different from all duplications, but not between other genomic structures. Genotype-phenotype analyses indicated a gradual worsening of phenotypic features, including overall survival, developmental levels, microcephaly, epilepsy, and genitourinary/eye abnormalities in the following order: tandem duplications, other CGRs, terminal duplications/translocations, and triplications encompassing MECP2. HPO analysis revealed specific patterns between different structures.

Conclusion: MECP2 copy number alterations alone do not solely influence the disease phenotype; the genomic rearrangement structure also plays crucial role in clinical manifestations and severity. Employing comprehensive analytical approach enhances our understanding of genomic disorder impact, guiding more effective therapeutic design.

Conflict of Interest: None declared

C04.1 The spliceosomal factor CRNKL1 is a novel disease gene required for brain development

Sankalita Ray¹, Rosie Sullivan¹, Mischa Ruegg¹, Gregory Gimenez², Julia Horsfield², Gemma Poke³, Sarah Duerinckx⁴, Muzhirah Haniffa⁵, Wei-Teik Keng⁵, Gaik-Siew Ch’ng⁶, David Parry⁷, Masamune Sakamoto⁸, Naomichi MATSUMOTO⁸, Emma Wakeling⁹, Ghaydda Mirzaa^10;11, Kimberly A. Aldinger^12;13, William Dobyns¹⁴, Marc Abramowicz¹⁵, Louise Bicknell ¹

¹University of Otago, Department of Biochemistry, Dunedin, New Zealand; ²University of Otago, Department of Pathoogy, Dunedin, New Zealand; ³Genetic Health Service New Zealand, Central Hub, Wellington, New Zealand; ⁴Université Libre de Bruxelles (ULB), Institut de Recherche Interdisciplinaire en Biologie Humaine et Moléculaire (IRIBHM), Brussels, Belgium; ⁵Hospital Kuala Lumpur, Department of Genetics, Kuala Lumpur, Malaysia; ⁶Penang Hospital, Department of Genetics, Penang, Malaysia; ⁷Institute of Genetics and Cancer, MRC Human Genetics Unit, Edinburgh, United Kingdom; ⁸Yokohama City University Graduate School of Medicine, Department of Human Genetics, Yokohama, Japan; ⁹Great Ormond Street Hospital for Children NHS Foundation Trust, North East Thames Regional Genetics Service, London, United Kingdom; ¹⁰Brotman Baty Institute for Precision Medicine, Seattle, United States; ¹¹University of Washington, Department of Pediatrics, Seattle, United States; ¹²Seattle Children’s Research Institute, Center for Integrative Brain Research, Seattle, United States; ¹³University of Washington, Division of Genetic Medicine, Department of Pediatrics, Seattle, United States; ¹⁴University of Minnesota, Department of Pediatrics, Division of Genetics and Metabolism, Minneapolis, United States; ¹⁵Université de Geneve, Department of Genetic Medicine and Development, Faculty of Medicine, Geneva, Switzerland

Background/Objectives: Splicing is a complex process, requiring for cell functioning and to create the transcriptomic diversity required for specialised functions in higher eukaryotes. Despite the ubiquitous nature, pathogenic variants in components associated with genetic disease often cause a tissue-specific phenotype, hinting at further complexities which are not yet fully understood.

Methods: Exome sequencing and clinical collaborations or Genematcher partners identified a patient cohort. In vitro and zebrafish experiments were performed to explore the pathogenicity of identified variants, including RNA-seq analysis for the effects on splicing.

Results: We have identified eight families with de novo missense variants in a spliceosomal component, CRNKL1. Significantly, seven individuals have one of two variants that all affect the same amino acid, while the eighth individual has a similar substitution affecting an amino acid in the same position in a neighbouring a-helix. All patients share a specific phenotype: profound pre- and post-natal microcephaly, with pontocerebellar hypoplasia, seizures and severe intellectual disability. Patient-derived fibroblasts show no obvious effects, with minimal splicing alterations and healthy growth. A zebrafish model demonstrated an obvious lack of brain development with a significant reduction in proliferating cells. RNA-seq analysis indicated widespread splicing changes, with altered expression of neuronal and cell cycle genes.

Conclusion: We have identified CRNKL1 as a novel disease gene and demonstrate the requirement for this protein in brain development and functioning. Our findings contribute to a growing disease cluster within the spliceosome which causes a severe brain-specific phenotype.

Grants: Neurological Foundation of New Zealand, Cure Kids New Zealand

Conflict of Interest: None declared

C04.2 A genetic link between neurodevelopmental disorders and activation of transposable elements: TRIM28 as a novel disease causal gene

Laura Castilla-Vallmanya ¹, Ninoslav Pandiloski¹, Carrie Davis-Hansson¹, Fereshteh Dorazehi¹, Christopher Douse¹, Susanna Balcells², Raquel Rabionet², Johan Jakobsson¹

¹Lund University, Stem Cell Center, Lund, Sweden; ²University of Barcelona, Genetics, Microbiology and Statistics, Barcelona, Spain

Background/Objectives: TRIM28 is an epigenetic co-repressor protein that acts as a master regulator of transposable elements (TEs). These mobile genetic sequences are normally epigenetically silenced to protect genome integrity. TRIM28 silences specific TEs during brain development in neural progenitors. Perturbations of this pathway in mice result in aberrant neurodevelopment and an inflammatory response in the adult brain. Here, we aim to provide functional proof of the pathogenicity of TRIM28 mutations in humans.

Methods: We identified heterozygous missense variants in TRIM28 in four individuals with moderate intellectual disability through whole exome sequencing. We generated CRISPR-edited induced pluripotent stem cells (iPSCs) lines carrying two of these mutations, which affect the PHD-Bromodomain, and a TRIM28 knock-down iPSC model using CRISPRi. We differentiated the TRIM28-mutant iPSCs into forebrain neural progenitors and unguided neural organoids. The transcriptional and epigenetic profile of our models was analyzed by RNA-seq and CUT&RUN histone mark profiling.

Results: We found that TRIM28-mutant iPSCs show a loss of the histone mark H3K9me3 coupled with overexpression of genes and TEs, predominantly endogenous retroviruses (ERVs), an abundant type of TE. These transcriptional changes highly overlap with those observed in the TRIM28-CRISPRi cells, suggesting a loss-of-function effect of the mutations. TRIM28-mutant differentiated neural progenitor cells also showed epigenetic and transcriptional changes related to abnormal neurodevelopment.

Conclusion: Our results show that TRIM28 mutations lead to epigenetic dysregulation causing aberrant TE and gene expression and could explain the neurodevelopmental phenotype of the patients.

Grants: Lindhés Advokatbyrå, Nilsson-Ehle, Vetenskapsrådet, Cancerfonden, Barncancerfonden, Hjärnfonden, Ministerio Ciencia, Innovación y Universidades.

Conflict of Interest: None declared

C04.3 Somatic Short-Tandem Repeats (STR) amplification in neurodevelopmental disorders with regression: the role of FAN1 and DNA repair pathways deficiency

Maria Cerminara ^1;2, Giovanni Spirito^3;4, Silvia Boeri⁵, Luca Pandolfini⁴, Giulia Rosti^2;6, Margherita Mancardi⁵, Livia Pisciotta^7;8, Marco Fontana^1;2, Alessandra Bianchi^1;2, Loretta Ferrera¹, Francesco Caroli¹, Marco Di Duca¹, Maria Teresa Divizia⁶, Elisa De Grandis^2;5, Sara Trova³, Diego Vozzi⁴, Remo Sanges^4;9, Lino Nobili^2;5, Federico Zara^1;2, Stefano Gustincich^4;9, Aldamaria Puliti^1;2

¹IRCCS Istituto Giannina Gaslini, Medical Genetics Unit, Genoa, Italy; ²University of Genoa, DiNOGMI Dipartimento di Neuroscienze, Riabilitazione, Oftalmologia, Genetica e Scienze Materno-Infantili, Genoa, Italy; ³CMP3vda, Aosta, Italy; ⁴Istituto Italiano di Tecnologia – IIT, Central RNA Laboratory, Genoa, Italy; ⁵IRCCS Istituto Giannina Gaslini, Child Neuropsychiatry Unit, ERN EpiCARE Centre, Genoa, Italy; ⁶IRCCS Istituto Giannina Gaslini, Clinical Genetics and Genomics Unit, Genoa, Italy; ⁷ASST Fatebenefratelli Sacco, Child Neuropsychiatry Unit, Milan, Italy; ⁸University of Milan, Milan, Italy; ⁹Scuola Internazionale Superiore di Studi Avanzati (SISSA), Area of Neuroscience, Trieste, Italy

Background/Objectives: Some patients with neurodevelopmental disorders (NDDs) exhibit regression, losing acquired developmental milestones. DNA repair genes, including FAN1, have been proposed as age-modifier of neurological/psychiatric diseases acting through STRs instability. In NDDs, this mechanism remains unexplored. We aim to advance our understanding on FAN1 and STRs amplification in regressive NDDs employing advanced integrated approaches.

Methods: We used multiple genomic technologies, including exome sequencing, short-read (SR-GS) and long-read genome sequencing (LR-GS) (Oxford Nanopore Technologies), to detect SNVs/indels and STR expantions from blood DNA, Mitomycin C (MMC)-induced cytotoxicity to evaluate DNA damage in lymphoblastoid cells, and capillary electrophoresis-based PCR assay to verify STRs amplification in possible target genes.

Results: Among 7 sporadic cases with heterozygous FAN1 loss-of-function variants, we selected three patients who showed regression. All three patients showed increased sensitivity to MMC-mediated cytotoxicity. SR-GS data indicated additional deleterious variants in other DNA-repair genes, and an increased burden of somatic SNV/indels compared to their parents. In one patient, SR-GS and LR-GS data showed increased amount of short-reads mapping inside repeats, and long-reads, supporting potentially deleterious STR expansions especially in brain-related Gene-Ontologies and NDD genes (e.g. SHANK2). This patient had a maternal FAN1 nonsense variant, and paternally inherited ERCC2 missense and SFR1 nonsense variants. These genes interact in DNA repair pathways via RAD51, central for homologous recombination.

Conclusion: Our study suggests FAN1 deficiency and simultaneous alterations in other DNA repair genes, can lead to somatic STRs expansion, and accumulation in NDD genes, potentially contributing to regressive phenotypes.

Grants: PRIN-20203P8C3X_001

Conflict of Interest: None declared

C04.4 Molecular and functional characterization of a novel tubulinopathy caused by recessive variants in the Tubulin Tyrosine Ligase gene

Valentina Serpieri ¹, Alessia Orsi¹, Satish Bodakuntla^2;3, Veronique Henriot^2;3, Roberta De Mori⁴, Monia Ginevrino⁵, Patrizia Accorsi⁶, Lorenzo Pinelli⁷, Abdulaziz Alsemari⁸, Mohanned Alsuhaibani⁹, Stefan T Arold¹⁰, Caroline Bénech¹¹, Tommaso Biagini¹², Renato Borgatti^13;14, Lucio Giordano⁶, Joseph Gleeson^15;16, Maria Magdalena Magiera^2;3, Tommaso Mazza¹², Sylvia Quemener^11;17;18, Rocio Rius^19;20, Romina Romaniello¹⁴, Silvia Tardivo⁴, Kevin Uguen^11;17;18, Sue White^19;20, Maha Zaki²¹, Carsten Janke^2;3, Enza Maria Valente²²

¹University of Pavia, Department of Molecular Medicine, Pavia, Italy; ²Institut Curie, Université PSL, CNRS UMR3348, Orsay, France; ³Université Paris-Saclay, CNRS UMR3348, Orsay, France; ⁴IRCCS Santa Lucia Foundation, Induced Pluripotent Stem Cells Unit, Rome, Italy; ⁵IRCCS Bambino Gesù Children’s Hospital, Laboratory of Medical Genetics, Translational Cytogenomics Reseach Unit, Rome, France; ⁶ASST-Spedali Civili of Brescia, Unit of Child Neurology and Psychiatry, Brescia, Italy; ⁷ASST-Spedali Civili of Brescia, Neuroradiology Unit, Pediatric Neuroradiology Section, Brescia, France; ⁸King Faisal Specialist Hospital and Research Centre, Department of Neurosciences, Riyadh, Saudi Arabia; ⁹King Faisal Specialist Hospital and Research Centre, Department of Radiology, Riyadh, Saudi Arabia; ¹⁰King Abdullah University of Science and Technology (KAUST), Computational Bioscience Research Centre (CBRC), Riyadh, Saudi Arabia; ¹¹Univ Brest, Inserm, EFS, UMR 1078, GGB, Brest, France; ¹²Fondazione IRCCS Casa Sollievo della Sofferenza, Bioinformatics Unit, San Giovanni Rotondo, Foggia, Italy; ¹³University of Pavia, Department of Brain and Behavioral Sciences, Pavia, Italy; ¹⁴IRCCS Mondino Foundation, Unit of Child Neurology and Psychiatry, Pavia, Italy; ¹⁵University of California, Department of Neurosciences, San diego, La Jolla, United States; ¹⁶Rady Children’s Institute for Genomic Medicine, Rady Children’s Hospital, San Diego, La Jolla, United States; ¹⁷CHU de Brest, Service de Génétique Médicale et Biologie de la Reproduction, Brest, France; ¹⁸Centre de Référence Déficiences Intellectuelles de causes rares, Brest, France; ¹⁹Murdoch Children’s Research Institute, Victorian Clinical Genetics Service, Melbourne, Australia; ²⁰University of Melbourne, Department of Paediatrics, Melbourne, Australia; ²¹Human Genetics and Genome Research Institute, National Research Centre, Clinical Genetics Department, Cairo, Egypt; ²²IRCCS Mondino Foundation, Neurogenetics Research Center, Pavia, Italy

Background/Objectives: Tubulinopathies are complex brain malformation syndromes caused by pathogenic heterozygous variants in genes encoding tubulin subunits. So far, genes implicated in post-translational modifications of tubulins have not been shown to cause tubulinopathy-related phenotypes. By exome sequencing and GeneMatcher search, we identified 8 patients from 5 unrelated families carrying biallelic pathogenic variants in TTL, a gene encoding tubulin tyrosine ligase. This is a ubiquitous enzyme that catalyzes the conversion of detyrosinated to tyrosinated α-tubulin. The phenotype is characterized by a complex neurological presentation (global developmental delay, intellectual disability, ataxia, spasticity, microcephaly) and a peculiar brain malformative pattern of the cerebral cortex, basal ganglia, commissures and posterior fossa, resembling a tubulinopathy.

Methods: Functional studies focused so far on TTL p.Cys338Tyr, detected in homozygosity in two sisters.

Results: Molecular dynamics simulation showed that mutant TTL protein is abnormally flexible, potentially leading to impaired interaction with α-tubulin. An in vitro enzymatic assay demonstrated almost abolished enzymatic activity. Moreover, TTL Cys338Tyr overexpression in mouse hippocampal neurons resulted in abnormally increased axonal length compared to controls. Preliminary studies in fibroblasts from the two sisters disclosed markedly reduced levels of the mutant protein, a significant reduction of primary cilia length and proportion of ciliated cells, an increased percentage of aberrant mitotic spindles, and abnormal cell shape indicating reduced polarization of lamellipodium extension.

Conclusion: Our preliminary work indicates that TTL loss of function causes a novel recessive tubulinopathy, suggesting that mutations in genes involved in tubulin post-translational modifications may account for a proportion of undiagnosed patients with tubulinopathies.

Grants:

Conflict of Interest: None declared

C04.5 Hypomorphic endocytic component REPS2 interferes with neuronal differentiation and causes X-linked cerebral palsy and cognitive impairment

Hao Hu¹

¹Guangzhou Women and Children’s Medical Center, Laboratory of Medical Systems Biology, Guangzhou, China

Background/Objectives: Clathrin-mediated endocytosis (CME) plays an important role in neuronal maturation and synaptic integration. Here, we identified hypomorphic variants of REPS2 on chromosome X, encoding a regulatory component of CME, from four unrelated male patients with defects in motor and cognition.

Methods: bioinformatics and omics technology; cell, zebrafish and mice models

Results: REPS2 was found to be ubiquitously expressed in multiple brain regions with enrichment in cortical neurons at different developmental stages. In zebrafish models, reps2-knockdown led to brain abnormalities and behavioral alteration, which could be effectively rescued by the human wild-type REPS2 transcript but not by transcripts with the hypomorphic variants. The mouse knockin (KI) models with a hypomorphic Reps2 variant showed impaired performances in motor, cognition, and social behaviors. The defective Reps2 hindered migration of cortical neuronal cells in embryos and interfered with cortical neurite development during adulthood. The defective Reps2 also impeded the formation of clathrin-coated vesicles and thus the progression of cortical neuronal endocytosis.Via single-cell transcriptomics and proteomics analysis, we identified a batch of critical genes affected by the Reps2 defect and confirmed their mutual interactions at cellular level. Furthermore, we found the deficit of dopamine in specific brain regions of the mice KI models, and ameliorated the behavioral abnormalities by supplementing L-dopa accordingly.

Conclusion: In this work, we established for the first time the phenotype–genotype correlation for Reps2 in brain development, and provided novel mechanistic and therapeutic insights into CME-dependent neuronal differentiation.

Grants: This work was supported by the National Natural Science Foundation of China (81974163, 82001121).

Conflict of Interest: None declared

C04.6 Association of rare functionally deficient SMPD1 mutations with Parkinson’s disease

Elaine Chew ¹, Zhehao Liu^2;3, Zheng Li³, Sun Ju Chung⁴, Michelle Lian¹, Moses Tandiono¹, Yue Jing Heng¹, Ebonne Ng⁵, Louis Tan⁶, Wee Ling Chng², Tiak Ju Tan³, Esther Peh⁷, Ying Swan Ho⁷, Xiao Yin Chen³, Erin Lim³, Chu Hua Chang¹, Jonavan Leong¹, Ting Xuan Peh³, Ling-Ling Chan⁸, Yinxia Chao⁵, Wing-Lok Au⁶, Kumar Prakash⁵, Jia Lun Lim⁹, Yi Wen Tay⁹, Vincent Mok¹⁰, Anne Chan¹⁰, Juei-Jueng Lin¹¹, Beomseok Jeon¹², Kyuyong Song¹³, Clement Tham¹⁴, Chi Pui Pang¹⁴, Jeeyun Ahn¹⁵, Kyu Hyung Park¹⁶, Janey L. Wiggs¹⁷, Tin Aung^2;18, Ai Huey Tan¹⁹, Azlina Ahmad Annuar⁹, Mary Makarious^20;21, Cornelis Blauwendraat²⁰, Mike Nalls^20;22, Laurie Robak^23;24, roy alcalay²⁵, ziv gan-or^26;27;28, Shen-Yang Lim¹⁹, chiea chuen khor^2;3, Eng-King Tan^2;6, Zhenxun Wang^2;3, Jia Nee Foo^1;3

¹Lee Kong Chian School of Medicine, Nanyang Technological University Singapore, Singapore, Singapore; ²Duke-National University of Singapore Medical School, Singapore, Singapore; ³Genome Institute of Singapore, Singapore, Singapore; ⁴Department of Neurology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea, Rep. of South; ⁵Department of Neurology, National Neuroscience Institute, Singapore General Hospital, Singapore, Singapore; ⁶Department of Neurology, National Neuroscience Institute, Singapore, Singapore; ⁷Bioprocessing Technology Institute, Singapore, Singapore; ⁸Department of Diagnostic Radiology, Singapore General Hospital,, Singapore, Singapore; ⁹Department of Biomedical Science, Faculty of Medicine, University of Malaya, Kuala Lumpur, Malaysia; ¹⁰The Chinese University of Hong Kong, Hong Kong, Hong Kong; ¹¹Department of Neurology, Chushang Show-Chwan Hospital, Nantou, Taiwan; ¹²Department of Neurology, Seoul National University Hospital, Seoul, Korea, Rep. of South; ¹³Department of Biochemistry and Molecular Biology, University of Ulsan College of Medicine, Seoul, Korea, Rep. of South; ¹⁴Department of Ophthalmology and Visual Sciences, Chinese University of Hong Kong, Hong Kong, Hong Kong; ¹⁵Department of Ophthalmology, Seoul Metropolitan Government Seoul National University Boramae Medical Center, Seoul, Korea, Rep. of South; ¹⁶Department of Ophthalmology, Seoul National University Hospital,, Seoul, Korea, Rep. of South; ¹⁷Massachusetts Eye and Ear Infirmary, Department of Ophthalmology, Harvard Medical School, Boston, United States; ¹⁸Singapore Eye Research Institute, Singapore, Singapore; ¹⁹Department of Medicine and the Mah Pooi Soo and Tan Chin Nam Centre for Parkinson’s and Related Disorders, Faculty of Medicine, University of Malaya, Kuala Lumpur, Malaysia; ²⁰Laboratory of Neurogenetics, National Institute on Aging, National Institutes of Health, Bethesda, United States; ²¹Department of Clinical and Movement Neurosciences, UCL Queen Square Institute of Neurology, London, United Kingdom; ²²Data Tecnica International LLC, Glen Echo, United States; ²³Department of Molecular and Human Genetics, Baylor College of Medicine, Texas, United States; ²⁴Jan and Dan Duncan Neurologic Research Institute, Texas Children’s Hospital, Houston, United States; ²⁵Tel Aviv Sourasky Medical Center, Tel Aviv, Israel; ²⁶The Neuro (Montréal Neurological Institute-Hospital), McGill University, Montréal, Canada; ²⁷Department of Human Genetics, McGill University, Montréal, Canada; ²⁸Department of Neurology and Neurosurgery, McGill University, Montréal, Canada

Background/Objectives: Parkinson’s disease is a progressive neurodegenerative disease that adversely affects balance, muscle control and movement. We seek to improve understanding of Parkinson’s disease pathogenesis mechanisms by gleaning mechanistic insights from rare protein-altering genetic variants associated with the disease.

Methods: Whole-exome sequencing was performed on 9,810 individuals (Discovery cohort: 4,298 patients and 5,512 controls) of East Asian ancestry. We tested gene-based association of rare protein-altering variants with Parkinson’s disease status and tested findings in another 11,227 individuals (Replication cohort: 5,585 patients and 5,642 controls) of Asian and European ancestry. Functional assays were conducted to delineate pathogenicity of observed rare protein-altering SMPD1 variants.

Results: Exome-wide significant association (P < 2.5×10^-6) of both GBA1 and SMPD1 were observed in the Discovery cohort and subsequently replicated. We established that rare protein-altering SMPD1 variants with less than 44% of normal acid sphingomyelinase activity were associated with Parkinson’s disease risk in both Discovery (odds ratio (OR) = 2.37, 95% CI = 1.68 - 3.35, P = 4.35 × 10^-7) and Replication cohorts (OR = 2.18, 95% CI = 1.69 - 2.81, P = 4.80 × 10^-10), thus resulting in highly significant association observed when all cohorts were meta-analyzed (OR = 2.24, 95% CI = 1.83 - 2.76, P = 1.25 × 10^-15).

Conclusion: Our findings show that rare functionally deficient SMPD1 variants are associated with increased Parkinson’s disease risk and emphasize the significance of sphingomyelin metabolism in Parkinson’s disease pathobiology. Our approach also underscored the utility of functional assays to delineate pathogenic variants in exome sequencing investigations.

Grant: NRF-NRFF2016-03, MOH-000559, MOE-MOET32020-0004

Conflict of Interest: None declared

C05.1 Characterising the contribution of rare protein-coding variants to prostate cancer risk and severity in 36,866 cases

Jonathan Mitchell ¹, Niedzica Camacho¹, Patrick R. Shea², Konrad Stopsack³, Vijay Joseph⁴, Oliver Burren¹, Kathryn Penney³, Jacob Berchuck⁵, Quanli Wang¹, David Goldstein², Athena Matakidou⁶, Carolina Haefliger¹, Ruth March⁷, Brian Dougherty⁸, Keren Carss¹, Slavé Petrovski¹, Vaidehi Jobanputra², Philip Kantoff⁴, Kenneth Offit⁴, Lorelei Mucci³, Mark Pomerantz⁵, Margarete Fabre¹

¹AstraZeneca, Centre for Genomics Research, Cambridge, United Kingdom; ²Columbia University, New York, United States; ³Harvard University, Cambridge, United States; ⁴Memorial Sloan Kettering Cancer Center, New York, United States; ⁵Dana-Farber Cancer Institute, Boston, United States; ⁶GSK, Stevenage, United Kingdom; ⁷AstaZeneca, Cambridge, United Kingdom; ⁸AstraZeneca, Watham, United States

Background/Objectives: The pathogenesis of prostate cancer has a strong heritable component. While rare coding germline variants in several genes have been identified as risk factors from candidate gene and linkage studies, the full spectrum of causal rare variants exome-wide remains largely unexplored. To address this, we meta-analysed five exome/genome sequenced cohorts in combination with imputed array data from a population with enriched low-frequency deleterious variants.

Methods: Firstly, we performed a gene-level collapsing analysis, utilising global biobanks, curated disease cohorts and AstraZeneca clinical trial participants with germline whole exome (WES) or whole genome sequencing (WGS) data (total 19,644 cases; 154,025 controls). Secondly, we performed an ExWAS, additionally incorporating imputed array data from the FinnGen cohort (total 33,326 cases; 273,642 controls).

Results: In the gene-level collapsing analysis we identified rare damaging variants as associated with the development of prostate cancer in the DNA damage response pathway genes BRCA2, ATM and CHEK2, and additionally SAMHD1. In the association analysis with prostate cancer severity, we discovered that rare damaging variants in AOX1 increased severity, as do rare damaging variants in BRCA2. At the single-variant level, we found rare non-synonymous variants significantly associated with increased risk of prostate cancer in three genes (HOXB13, CHEK2, BIK), and with decreased risk in four genes (ANO7, SPDL1, AR, TERT).

Conclusion: Our study clarifies the role of known and identifies novel genes associated with prostate cancer risk and severity, providing deeper insight into disease pathogenesis, with the potential for improved risk stratification and novel therapeutic approaches.

Grants: NA

Conflict of Interest: Jonathan Mitchell AstaZeneca, Niedzica Camacho AstraZeneca, Patrick R. Shea: None declared, Konrad Stopsack: None declared, Vijay Joseph: None declared, Oliver Burren AstraZeneca, Kathryn Penney: None declared, Jacob Berchuck: None declared, Quanli Wang AstraZeneca, David Goldstein Actiobio, Athena Matakidou GSK, Carolina Haefliger Debiopharm, Ruth March AstraZeneca, Brian Dougherty AstraZeneca, Keren Carss AstraZeneca, Slavé Petrovski AstraZeneca, Vaidehi Jobanputra: None declared, Philip Kantoff Convergent Therapeutics, Kenneth Offit: None declared, Lorelei Mucci: None declared, Mark Pomerantz: None declared, Margarete Fabre AstraZeneca

C05.2 Deciphering dual clinical entities associated with TP53 pathogenic variants: insights from 53,085 HBOC panels analysis in French laboratories

Edwige Kasper ¹, Flavie boulouard^2;3, BASSET NOEMIE⁴, Golmard Lisa⁵, Hela Sassi⁶, Emilie Bouvignies⁷, Maud Brauchaud⁷, Charbonnier Camille⁸, Nathalie Parodi⁷, marion rolain⁷, Juliette Albuisson⁹, Ayman AL SAATI¹⁰, Marie Bidart^11;12, celine bonnet^13;14, Ahmed Bouras¹⁵, Nadia Boutry-Kryza¹⁶, Virginie Bubien¹⁷, Adrien Buisson¹⁸, Laurent Castera^2;3, Chrystelle Colas⁵, Florence Coulet⁴, Capucine DELNATTE¹⁹, Valentin Derangère²⁰, Celine Garrec²¹, Marion Gauthier-Villars⁵, Mathilde Gay-Bellile²², Vincent Goussot²³, JESSICA LE GALL⁵, Mathis Lepage²², Anna Lokchine², Etienne ROULEAU⁶, Nicolas Sevenet^17;24, Dominique Stoppa-Lyonnet^5;25;26, Pierre VANDE PERRE²⁷, Dominique Vaur^2;3, Paul Vilquin²⁸, Gaelle Bougeard²⁹, Stéphanie Baert-Desurmont⁸, Théry Jean-Christophe³⁰, Claude Houdayer¹

¹Univ Rouen Normandie, Normandie Univ, Inserm U1245 and CHU Rouen, Department of Genetics, F-76000 Rouen, France; ²Laboratory of cancer biology and genetics, Centre François Baclesse, Caen, France; ³Inserm U1245, Cancer and Brain Genomics, Normandie Univ, UNICAEN, FHU G4 génomique, Rouen, France; ⁴Département de Génétique médicale APHP Sorbonne Université U.F. d’Onco-angiogénétique et Génomique des tumeurs solides Hôpitaux Universitaires Pitié Salpêtrière - Charles Foix, Paris; ⁵Department of Genetics, Institut Curie, Paris, France PSL University, Paris, France; ⁶Service de Génétique des Tumeurs, Institut Gustave Roussy, Villejuif; ⁷Univ Rouen Normandie, Normandie Univ, CHU Rouen, Department of Genetics, F-76000 Rouen, France; ⁸Univ Rouen Normandie, Normandie Univ, Inserm U1245 and CHU Rouen, Department of Biostatistics and CNRMAJ, F-76000 Rouen, France.; ⁹Plateforme de Transfert en Biologie Cancérologique, Centre Georges Francois Leclerc, 1 rue Pr Marion, 21000 DIJON; ¹⁰Oncogenetics Laboratory, Oncopole Claudius Regaud, IUCT-Oncopole, Toulouse, France et Université de Toulouse, Université Toulouse III-Paul Sabatier, Toulouse, France; ¹¹1-CHU Grenoble Alpes, Institut de Biologie et Pathologie, Laboratoire de Génétique Moléculaire : Maladies Héréditaires et Oncologie; ¹²2-Université Grenoble Alpes, Institut Albert Bonniot, INSERM 1209, CNRS UMR 5309, Equipe “Génétique, Epigénétique et Thérapies de l’Infertilité”; ¹³Genetics Laboratory, University Hospital of Nancy, Nancy, France;; ¹⁴INSERM UMR_S 1256, Nutrition, Genetics, and Environmental Risk Exposure (NGERE), Faculty of Medicine of Nancy, University of Lorraine, Nancy, France; ¹⁵Centre Léon Bérard, Laboratory of Constitutional Genetics for Frequent Cancer HCL-CLB, Lyon, France; ¹⁶Genetic Department, Hospices civils de Lyon, Bron, France; ¹⁷Unité d’Oncogénétique, Institut Bergonié, 229 cours de l’Argonne, 33076 Bordeaux cedex; ¹⁸Bio-pathology department, Centre Léon Bérard, Lyon, France; ¹⁹Institut de Cancérologie de l’Ouest, Unité d’Oncogénétique, Bd Professeur Jacques Monod, 44800 Saint Herblain, France (les 2 pour CD et juste le premier pour CG); ²⁰Plateforme de Transfert en Biologie Cancérologique (CGFL), Centre Georges Francois Leclerc, 1 rue Pr Marion, 21000 DIJON, INSERM UMR1231, Université de Bourgogne; ²¹Laboratoire de Génétique Moléculaire, Service de Génétique Médicale, Centre Hospitalier Universitaire (CHU) de Nantes, 9 quai Moncousu Nantes 44093, France); ²²Department of Oncogenetics, Centre Jean Perrin, INSERM U1240, Université Clermont Auvergne, Clermont-Ferrand 63000, France.; ²³Département de Biologie et Pathologie des Tumeurs, Centre Georges Francois Leclerc, 1 rue Pr Marion, 21000 DIJON; ²⁴INSERM U1312, BRIC, Université de Bordeaux, 146 rue Léo Saignat, 33076 Bordeaux cedex; ²⁵INSERM U830, Institut Curie, Paris, France; ²⁶Université Paris Cité, Paris, France; ²⁷Oncogenetics Laboratory, Oncopole Claudius Regaud, IUCT-Oncopole, Toulouse, France; ²⁸Service de Génomique des Tumeurs et Pharmacologie, AP-HP. Paris Cité Université, Hôpital Saint-Louis, Paris, France; ²⁹Univ Rouen Normandie, Inserm U1245, F-76000 Rouen, France; ³⁰Univ Rouen Normandie, Normandie Univ, Inserm U1245, Centre Henri Becquerel, F-76000 Rouen, France

Background/Objectives: Germline TP53 variants predispose to Li-Fraumeni syndrome (LFS), characterized by a wide tumor spectrum at an early age. Excess risk of early onset breast cancer (<31y) has motivated its inclusion in Hereditary Breast and Ovarian Cancer (HBOC) panels. Aim of this work is to provide new insights about TP53 in HBOC patients.

Methods: We collected personal and familial cancer history of 398 patients harboring a TP53 variant identified by 53,085 HBOC panels sequencing in 15 French laboratories.

Results: Heterozygous TP53 variant was identified in 0.44% of HBOC panels, evenly distributed between pathogenic variants (PV) and VUS. Breast cancers associated with TP53 were predominantly triple positive particularly Her2+ breast cancer (p < 0.0001), in situ cancer (p < 0.0001) or phyllode tumors (p < 0.0001). Interestingly, TP53 PV were identified across all ages in breast cancer patients with a median age of onset at 37y [22;82], without enrichment before 31y. We demonstrated that LOF variants were linked with HBOC phenotype, and missense variants, especially with dominant negative effect, with LFS phenotype (p = 0.0096). Patients with breast cancer harboring LOF variants displayed an earlier age of onset compared to missense (p = 0.0030). Surprisingly, we identified, in late onset cancer patients,TP53 hotspot PV usually identified in classic LFS which underlines variable penetrance.

Conclusion: This study suggests the existence of two phenotypic entities associated with TP53 PV: clinical LFS and TP53-related breast cancer. The type of TP53 variant and modifying factors, reflected in familial history, influence these phenotypes and both should be considered to define clinical follow-up of patients and relatives.

Conflict of Interest: None declared

C05.3 Estimating the frequency of Birt-Hogg-Dubé syndrome from genomics and population cohorts

Bryndis Yngvadottir ^1;2, Lucy Richman³, Avgi Andreou², Jessica Woodley^4;5, Anita Luharia^4;5, Derek Lim⁶, Eamonn Maher^2;7, Stefan Marciniak^1;3;8;9

¹University of Cambridge, Department of Respiratory Medicine, Cambridge, United Kingdom; ²University of Cambridge, Department of Medical Genetics, Cambridge, United Kingdom; ³Cambridge University Hospitals NHS Foundation Trust, Cambridge Biomedical Campus, Cambridge, United Kingdom; ⁴Central and South Genomic Laboratory Hub, Birmingham Women’s and Children’s NHS Foundation Trust, Birmingham, United Kingdom; ⁵West Midlands Regional Genomics Laboratory, Birmingham, United Kingdom; ⁶Birmingham Women’s and Children’s NHS Foundation Trust, Department of Clinical Genetics, Birmingham, United Kingdom; ⁷Aston University, Aston Medical School, Birmingham; ⁸University of Cambridge, Cambridge Institute for Medical Research (CIMR), Cambridge, United Kingdom; ⁹Royal Papworth Hospital, Cambridge, United Kingdom

Consortium: Genomics England Research Consortium

Genes & Health Research Team

Background/Objectives: Birt-Hogg-Dubé syndrome (BHDS) is an autosomal dominant disorder characterised by benign skin lesions, lung cysts, pneumothorax and an increased lifetime risk for kidney cancer. The syndrome is likely clinically underdiagnosed and its prevalence in the general population underestimated. Most affected families have pathogenic variants in the FLCN gene.

Methods: We extracted pathogenic loss-of-function FLCN variants from >550,000 exomes/genomes of participants in the 100,000 Genomes Project (100kGP), the UK Biobank (UKB) and East London Genes & Health (ELGH). We then calculated age-related risks of pneumothorax and kidney cancer in UKB and compared this to a clinical series of BHDS patients.

Results: The mutation frequency of loss-of-function FLCN variants was 1 in 2,714 in the 100kGP, 1 in 4,189 in UKB and 1 in 1,487 in ELGH. The lifetime risk of pneumothorax in FLCN mutation carriers in the population was substantial with no significant difference between the BHDS patient cohort and UKB participants (p = 0.2154) while the lifetime risk of kidney cancer was significantly lower in UKB compared to BHDS patients (p = 0.0005).

Conclusion: Our frequency estimates for pathogenic FLCN variants are at least ~48 times higher than previous clinical estimates of BHDS frequency (1 in 200,000). These findings highlight the importance of clinical context in managing individuals with FLCN mutations.

Grants: Myrovlytis Trust, Cambridge BRC

This research was made possible through access to data and findings in the National Genomic Research Library via the Genomics England Research Environment.

This research has been conducted using the UK Biobank Resource under application number 95547.

Conflict of Interest: None declared

C05.4 Effect of a Polygenic Risk Score for colorectal cancer incidence in patients with early-onset, familial, or hereditary colorectal cancer

Isabel Spier^1;2;3, Hannah Klinkhammer ^4;5, Claudia Perne¹, Per Hoffmann¹, Reinhard Büttner⁶, Friederike David¹, Evelin Schröck^3;7;8;9, Silke Redler¹⁰, Marcus Franke^3;7;8, Gabriela Möslein¹¹, Matthias Kloor^12;13, Börge Schmidt¹⁴, Annette Erle^1;2;3, Deepak Vangala^15;16, Huu Phuc Nguyen¹⁶, Verena Steinke-Lange^3;17;18, Elke Holinski-Feder^3;17;18, Andreas Forstner^1;19, Robert Hüneburg^2;3;20, Carlo Maj²¹, Christoph Engel²², Andreas Mayr⁴, Stefan Aretz^1;2;3

¹Institute of Human Genetics, Medical Faculty, University of Bonn, Bonn, Germany; ²National Center for Hereditary Tumour Syndromes, University Hospital Bonn, Bonn, Germany; ³European Reference Network on Genetic Tumour Risk Syndromes (ERN GENTURIS); ⁴Institute for Medical Biometry, Informatics and Epidemiology, Medical Faculty, University of Bonn, Bonn, Germany; ⁵Institute for Genomic Statistics and Bioinformatics, Medical Faculty, University of Bonn, Bonn, Germany; ⁶Institute of Pathology, University Hospital Cologne, Cologne, Germany; ⁷Institute for Clinical Genetics, Faculty of Medicine and University Hospital Carl Gustav Carus, TU Dresden, Dresden, Germany; ⁸National Center for Tumor Diseases Dresden (NCT/UCC), Germany: German Cancer Research Center (DKFZ), Heidelberg, Germany; ⁹German Cancer Consortium (DKTK), Dresden, Germany; ¹⁰Institute of Human Genetics, Medical Faculty and University Hospital Düsseldorf, Heinrich Heine University, Düsseldorf, Germany; ¹¹Center for Hereditary Tumors, Bethesda Hospital, Duisburg, Germany; ¹²Department of Applied Tumor Biology, Institute of Pathology, Heidelberg University Hospital, Heidelberg, Germany; ¹³Clinical Cooperation Unit Applied Tumor Biology, German Cancer Research Center Deutsches Krebsforschungszentrum (DKFZ), Heidelberg, Germany; ¹⁴Institute for Medical Informatics, Biometry and Epidemiology (IMIBE), University Hospital Essen, Essen, Germany; ¹⁵Gemeinschaftspraxis für Hämatologie und Onkologie, Ruhr-University Comprehensive Cancer Center, Dortmund, Germany; ¹⁶Department of Human Genetics, Ruhr-University Bochum, Bochum, Germany; ¹⁷Medizinische Klinik und Poliklinik IV, Klinikum der Universität München, Munich, Germany; ¹⁸MGZ - Medical Genetics Center, Munich, Germany; ¹⁹Institute of Neuroscience and Medicine (INM-1), Research Center Jülich, Jülich, Germany; ²⁰Department of Internal Medicine I, University Hospital Bonn, Bonn, Germany; ²¹Center for Human Genetics, University of Marburg, Marburg, Germany; ²²Institute for Medical Informatics, Statistics and Epidemiology (IMISE), University of Leipzig, Leipzig, Germany

Consortium: German Consortium for Familial Intestinal Cancer

Background/Objectives: The study investigates how a polygenic risk score (PRS) affects colorectal cancer (CRC) risk in individuals of different, clinically or molecularly defined risk groups, using a large cohort from the German Consortium for Familial Intestinal Cancer.

Methods: A total of 1,839 European-descendant individuals were stratified according to low, intermediate, or high PRS for the following risk groups: (i) Lynch syndrome [LS; n = 679 with CRC and n = 422 CRC-free carriers]; (ii) sporadic early-onset CRC [SEO-CRC; n = 518], (iii) CRC patients with positive family history [F-CRC; n = 220], and (iv) MSI/dMMR CRC [n = 144] vs. pMMR CRC [n = 485]. Logistic regression and Cox proportional hazards models were applied to estimate odds ratios and to compute lifetime incidences in comparison to 3,119 population based controls and sporadic late-onset CRC patients (≥60 years) [SLO-CRC; n = 97]).

Results: In all risk groups, CRC individuals showed significantly increased PRS compared to population controls (p < 0.02). The SEO-CRC and F-CRC risk groups both showed a two times increased CRC risk in the high PRS category (p < 0.001) corresponding to lifetime incidences of 23% and 13%, respectively. PRS did not differ significantly between SEO-CRC and SLO-CRC cases or between LS CRC patients and CRC-free LS carriers. Finally, non-LS individuals with MSI/dMMR tumors had a significantly lower PRS than individuals with pMMR tumors (p < 0.001), but no difference compared to LS CRC individuals.

Conclusion: Preliminary findings suggest PRS effectively stratifies CRC risk in specific groups, but larger studies are needed, particularly for SEO and SLO CRC cohorts, to further explore a PRS-based CRC risk stratification.

Grants:

Conflict of Interest: None declared

C05.5 Personalized peptide vaccination among 173 patients with glioblastoma

Pauline Latzer ¹, Henning Zelba¹, Florian Battke², Annekathrin Reinhardt¹, Borong Shao¹, Saskia Biskup¹

¹Zentrum für Humangenetik Tübingen; ²CeGaT GmbH

Background/Objectives: Current treatment outcome of patients with glioblastoma (GBM) remains poor. Following standard therapy, recurrence is universal with limited survival.

Methods: GBM tumors from 173 patients were analysed for somatic mutations to generate a personalized peptide vaccine targeting tumor-specific neoepitopes. In agreement with their treating physician, patients added a personalized peptide vaccine to their treatment as an individual healing attempt. Patients were monitored from October 2015 until August 2023. We retrospectively evaluated their clinical courses and the results of their immune monitoring data.

Results: Among all patients, including 70 treated prior to progression (primary) and 103 treated after progression (recurrent), the median overall survival from first diagnosis was 31.9 months (95% CI: 25.0-36.5). Adverse events were infrequent and were predominantly grade 1 or 2. An immune response to at least one of the vaccinated peptides was detected in blood samples of 87 of 99 (88%) monitored patients. T-cell responses to vaccinated neoepitope peptides were durable in most patients. Significantly prolonged survival was observed for patients with multiple vaccine-induced immune responses (53 months) compared to those with no/low induced responses (27 months; P = 0.03).

Conclusion: Altogether, our results highlight that the application of personalized neoantigen-targeting peptide vaccine is feasible and represents a promising potential treatment option for GBM patients. This is the largest real-world observation involving GBM patients safely treated with a personalized peptide vaccine to date. This real-world observation will be translated into a clinical trial entitled NeoGBM to evaluate the specific contribution of the neoantigen vaccine.

Grants: none.

Conflict of Interest: None declared

C05.6 Comprehensive guidelines for the diagnosis, counselling, surveillance and clinical management of people with constitutional mismatch repair deficiency: a join effort from ERN GENTURIS and the European consortium Care for CMMRD

Chrystelle Colas¹, Léa Guerrini-Rousseau², Manon Suerink³, Richard Gallon⁴, Christian Kratz⁵, Eloise Ayuso⁶, Laurence Brugières², Katharina Wimmer ⁷

¹Institut Curie, France; ²Gustave Roussy Cancer Center, Villejuif, France; ³Leiden University Medical Centre, Leiden, Netherlands; ⁴Newcastle University, Translational and Clinical Research Institute, Newcastle, United Kingdom; ⁵Hannover Medical School, Pediatric Hematology and Oncology, Hannover, Germany; ⁶Patient representative, Paris, France; ⁷Medical University Innsbruck, Institute of Human Genetics, Innsbruck, Austria

Consortium: ERN Genturis CMMRD Guideline Group

Background/Objectives: Constitutional mismatch repair deficiency (CMMRD) firstly described 25 years ago confers an extraordinarily high and lifelong cancer risk, including hematologic, brain, and gastrointestinal tract malignancies, and is associated with several non-neoplastic features. Our understanding of this condition has improved and novel assays to assist CMMRD diagnosis have been developed. Surveillance protocols need adjustment taking into account recent observational prospective studies assessing their effectiveness. Response to immune checkpoint inhibitors and the effectiveness and toxicity of other treatments have been described. An update and collation of the different guidelines on diagnosis and clinical management of CMMRD into one comprehensive guideline was needed.

Methods: Seventy-two expert members of the European Reference Network GENTURIS and/or the European care for CMMRD consortium and one patient representative developed recommendations for CMMRD diagnosis, genetic counselling, surveillance, quality of life, and clinical management based on a thorough literature review and a modified Delphi process.

Results: Recommendations for the diagnosis of CMMRD provide indication criteria and strategies for testing, and define criteria for diagnosis. Recommendations for surveillance cover each CMMRD-associated tumour type and contain information on starting age, frequency, and surveillance modality. Recommendations for clinical management cover cancer treatment, management of benign tumours or non-neoplastic features, and chemoprevention. Recommendations also address genetic counselling and quality of life.

Conclusion: Based on existing guidelines and all currently available data, we present 82 recommendations to improve and standardise the care of CMMRD patients in Europe. These recommendations are not meant to be prescriptive and may be adjusted based on individual decisions.

Conflict of Interest: Chrystelle Colas honoraria or consultation fees from AstraZeneca, Léa Guerrini-Rousseau: None declared, Manon Suerink: None declared, Richard Gallon grants/research support from Cancer Research UK Catalyst and UK National Health Service, Christian Kratz: None declared, Eloise Ayuso: None declared, Laurence Brugières honoraria or consultation fees from ESAI and TAKEDA, Katharina Wimmer: None declared

C06.1 Liquid Biospy identifies causes of cutaneous mosaicism and hemihypertrophy

Maximilian Witzel ¹, Thomas Keßler¹, Ariane Hallermayr¹, Angela Abicht¹, Teresa Neuhann¹

¹MGZ München, München, Germany

Background/Objectives: Mosaicism of skin with or without overgrowth remains a challenge in the pediatric and genetic practice. Sampling errors of blood DNA, skin biopsy, cultured fibroblast can hamper a correct diagnosis. Liquid biopsy offers a noninvasive tool, sampling whole body cfDNA.

Methods: Duplex Sequencing technology for detection of low frequency variants in plasma was applied in children with mosaic conditions. The multi-gene panel included AKT3, BRAF, CCND2, EPHB4, FGFR1, GNA11, GNAQ, GNAS, HRAS, KRAS, MAP2K1, MTOR, NRAS, PIK3CA, PIK3R2, RASA1, SMO and others (>20 genes).

Results: We identified relevant pathogenic variants with suballelic VAF in 6 patients: patient (P1) with the clinical diagnosis Klippel-Trenaunay syndrome (AKT1 gene, NM_001382430.1:c.49G>A, p.(Glu17Lys), VAF 0.6%; class 5), patient (P2) with cavernoma of basal ganglia and prominent vessels of the eye sclera (PIK3CA gene, NM_006218.4:c.277C>T, p.(Arg93Trp), VAF 0.14%, class 4), patient (P3) with hemihypertrophy, Naevus flammaeus and syndaktyly (PIK3CA gene, NM_006218.4:c.316G>T, p.(Gly106Cys), VAF 2%, class 4), patient (P4) with hemihypertrophy, hyperpigmentation (KRAS gene, NM_004985.5, c.35G>A p.(Gly12Asp), VAF 1%, class 5), patient (P5) with diffuse capillary malformation with overgrowth (PIK3CA gene, NM_006218.4:c.353G>A p.(Gly118Asp) class 4/5, VAF 5.1%) and patient (P6) with premature puberty, asymmetrical breast development, subcutaneous lipoma, diffuse hyperpigmentation and café-au-lait spots (PIK3CA gene, NM_006218.4:c.1093G>A p.(Glu365Lys), VAF 0.12%, class 4/5). Additional NGS sequencing (panel, exom or trio exome) of blood and skin fibroblast had remained unconclusive in all patients.

Conclusion: Liquid biopsy is noninvasive, precise and significantly increases diagnostic yield in patients with mosaic diseases with profound therapeutical implications, e.g. targeted therapy, cancer surveillance, counseling.

Conflict of Interest: Maximilian Witzel full time employed by MGZ München, Thomas Keßler full time employed by MGZ München, Ariane Hallermayr full time employed by MGZ München, Angela Abicht full time employed by MGZ München, Teresa Neuhann full time employed by MGZ München

C06.2 A novel form of Ehlers-Danlos syndrome with prominent vascular features in humans and mice caused by a heterozygous THBS2 mutation

Noam Hadar ^1;2, Omri Porgador^1;2, Idan Cohen², Hilla Levi², Vadim Dolgin^1;2, Yuval Yogev^1;2, Sufa Sued-Hendrickson^1;2, Ilan Shelef³, Elena Didkovsky⁴, Ohad Shmuel Birk^1;2;5, Marina Eskin-Shwartz^1;5;6

¹The Morris Kahn Laboratory of Human Genetics, Faculty of Health Sciences, Ben-Gurion University, Beer Sheva, Israel; ²The Shraga Segal Department of Microbiology, Immunology, and Genetics, Faculty of Health Sciences, Ben-Gurion University, Beer Sheva, Israel; ³Department of Radiology, Soroka Medical Center, and Faculty of Health Sciences, Ben-Gurion University, Beer Sheva, Israel; ⁴Department of Pathology, Rabin Medical Center, Petah-Tikva, and Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel; ⁵Genetics Institute, Soroka Medical Center, Beer Sheva, Israel; ⁶Faculty of Health Sciences, Ben-Gurion University, Beer Sheva, Israel

Background/Objectives: Ehlers-Danlos syndromes (EDS) are a group of connective tissue disorders caused by collagen and collagen-interacting gene mutations. We studied a novel distinct type of dominantly inherited EDS with vascular features.

Methods: Clinical and histopathological phenotyping and genetic human studies followed by generation and analyses of CRISPR/Cas9 knock-in (KI) mice.

Results: We studied a three-generation pedigree displaying an apparently autosomal dominant phenotype of frequent joint dislocations, joint hypermobility, atrophic scarring, prolonged bleeding time, and age-related aortic dilatation and rupture. Light microscopy revealed highly disorganized collagen fibers in the reticular dermis and transmission electron microscopy (TEM) displayed irregularly shaped fibroblasts and endothelial cells, with abundant amorphous extracellular matrix (ECM) substance, especially near blood vessels. Genetic analysis demonstrated a disease-causing heterozygous variant in THBS2 (NM_003247.5:c.2686T>C, p.Cys896Arg). We generated knock-in (KI) mice harboring this mutation in the mouse ortholog. Heterozygous THBS2^Cys896Arg KI mice exhibited clinical features similar to the human disease, as evidenced by comparable morphologic, histologic, and TEM findings, and prolonged bleeding times.

Conclusion: We delineate a new form of EDS with vascular features caused by a dominant THBS2 missense mutation. The phenotype observed in heterozygous THBS2^Cys896Arg KI mice mirrors that in the human disease and that of Thbs2 homozygous null-mutant mice. THBS2 encodes Thrombospondin-2, a secreted homotrimeric matricellular protein that directly interacts with Matrix Metalloproteinase 2 (MMP2), an ECM-shaping protein, and mediates its clearance. THBS2 loss-of-function hinders MMP2 clearance, causing an increase in MMP2-mediated proteoglycan cleavage, which leads to ECM abnormalities as seen in the human and mouse disease.

Conflict of Interest: None declared

C06.3 Generation of alpl zebrafish models for detailed pathogenic analysis of the rare disease hypophosphatasia.

Regina Hark ¹, Lukas Hekel¹, Angela Borst¹, Barbara Ohlebusch^1;2, Franz Jakob³, Eva Klopocki¹, Stephanie Graser³, Daniel Liedtke¹

¹Institute of Human Genetics, Biocenter, University of Würzburg, Würzburg, Germany; ²Zoological Institute, University of Cologne, Cologne, Germany; ³Bernhard-Heine-Centrum for Locomotion Research, Department of Musculoskeletal Tissue Regeneration, Orthopaedic Hospital König-Ludwig-Haus, University of Würzburg, Würzburg, Germany

Background/Objectives: Hypophosphatasia (HPP), caused by mutations in the gene ALPL, is a rare metabolic disorder resulting in osteogenic and neuronal symptoms. Established human in vitro and mouse models entail experimental limitations, which could be resolved by using zebrafish (Danio rerio) as a supplemental HPP model for in vivo analyses. Our work aims to establish and investigate transgenic zebrafish alpl knockout lines and thereby, provide insights into developmental deficits that have been poorly elucidated in patients so far.

Methods: We generated different alpl knockout zebrafish models using the CRISPR/Cas9 system and characterized these using macroscopic, microscopic, and molecular approaches (Sanger sequencing, in-situ hybridization, immunostaining, bone/cartilage staining, in-vivo observations).

Results: Successful alpl knockout was confirmed in these lines. Homozygous zebrafish develop a lethal phenotype at larval stage. Early developmental abnormalities were observed, including progressive body axis curving, lack of swimming orientation, jaw deformations and decreased mineralization of bone and teeth. Furthermore, immunostaining revealed altered axonal growth within the telencephalon, the dorsal root ganglia and in sensory neurons.

Conclusion: We successfully generated novel alpl knockout zebrafish lines, which display altered craniofacial and neuronal development, resembling severe HPP forms. The detailed characterization of these in vivo models will help to gain a deeper understanding of the associated neurological phenotype, which is currently neglected in HPP patients.

Grants: The project is funded by DFG grant nr. 397519724.

Conflict of Interest: Regina Hark: None declared, Lukas Hekel: None declared, Angela Borst: None declared, Barbara Ohlebusch: None declared, Franz Jakob: None declared, Eva Klopocki: None declared, Stephanie Graser: None declared, Daniel Liedtke DFG research grant

C06.4 Human embryonic single-cell RNA sequencing data identifies potential involvement of epithelial cells in non-syndromic cleft lip with/without cleft palate

Anna Siewert ¹, Elisabeth Mangold¹, Kerstin Ludwig¹

¹Institute of Human Genetics, University of Bonn, School of Medicine & University Hospital Bonn

Background/Objectives: Non-syndromic cleft lip with/without cleft palate (nsCL/P) is one of the most common birth defects. Although over 45 risk loci have been identified, the function of the risk alleles at these loci and affected cell types are mostly unknown. Here, we identified nsCL/P-related cell types based on candidate genes from GWAS loci and generated co-expression networks to suggest novel candidate genes based on joint expression patterns.

Methods: We re-analyzed published single-cell RNA sequencing (scRNA-seq) data from unaffected human embryos (4-6 weeks) using Seurat. We investigated the association of cell types with the expression of nsCL/P GWAS candidate genes using single-cell disease relevance scores (scDRS) and generated co-expression networks from these cell types using hdWGCNA.

Results: We found that nsCL/P candidate genes are significantly associated with the epithelium (p < 0.0001), which furthermore showed significant within-cell type scDRS heterogeneity (p = 0.007). The marker genes of the associated epithelial cells comprised known nsCL/P candidate genes (e.g. IRF6, TFAP2A, ESRP1). Within the epithelium, we identified 18 co-expression modules, three of which we prioritized based on their nsCL/P candidate genes (e.g. IRF6, TFAP2A, ARHGAP29, etc. in one module) and which include interesting new candidate genes (e.g. CALD1, PDGFC).

Conclusion: Our results suggest the existence of a specific subset of epithelial cells involved in nsCL/P development. For some nsCL/P candidate genes, genetic interaction has already been shown (e.g. IRF6 and TFAP2A). Our co-expression networks, however, indicate further genetic associations and interactions between known and novel candidate genes that should be investigated.

Grants: DFG: LU-1944/3-1

Conflict of Interest: None declared

C06.5 Understanding the molecular bases of a new Chondrodysplasia with Multiple Dislocations case.

Miriam Villegas Villarroel ¹, Céline Huber¹, Genevieve Baujat², Adeline Bonnard³, Corinne Collet^1;2, Valérie Cormier-Daire^1;2

¹Université de Paris Cité, INSERM UMR1163, Imagine Institut, Molecular and physiopathological bases of osteochondrodysplasia, Paris, France; ²AP-HP, Necker Hospital, Service de médecine génomique des maladies rares, Centre de référence pour les maladies osseuses constitutionnelles, Paris, France; ³Robert-Debré Hospital, Genetics, Paris, France

Background/Objectives: Chondrodysplasia with multiple dislocations (CLM) form a group of rare disorders mainly characterized by joint laxity and large joint dislocations. Most of these disorders have been associated with pathogenic variants in genes encoding proteins involved in the proteoglycan biosynthesis. Our aim was to confirm the pathogenicity of variants identified in a novel gene in a patient with CLM.

Methods: Whole Exome Sequencing as well as protein expression and GAG production through DMMB assay in patient fibroblasts were performed. Using a specific DDR1 inhibitor (DDR1-IN-1), we further studied the consequences in human chondrocytes and osteoblast differentiation and mineralization through ALP and alizarin red staining.

Results: We identified a homozygous nonsense mutation (exon 13, c.1825C>T) in the DDR1 gene (Discoidin Domain Receptor Family, Member 1; MIM #600408) in one patient with hyperlaxity, multiple dislocations, and cerebellar hypoplasia.

We demonstrated in patient fibroblasts an important decrease in DDR1 mRNA, absence of the protein and a GAG production defect. Using DDR1-IN-1, we found an impairment of the osteoblast differentiation and mineralization, and a defect in WNT signaling pathway. We also identified an alteration in the IHH signaling pathway in chondrocytes.

Conclusion: Our results suggest that DDR1 is a new gene involved in the CLM group, through a proteoglycan biosynthesis defect.

Grants: FRM and ANR

Conflict of Interest: None declared

C06.6 RHOA-related mosaic ectodermal dysplasia: a large international cohort study and natural history data

Elodie Javey ^1;2, Bénédicte DEMEER³, Evgenia Sklirou⁴, David Geneviève⁵, Laurence Perrin⁶, Didier Bessis⁷, Odile Boute⁸, jean-luc alessandri⁹, Marta Spodenkiewicz¹⁰, JACQUEMONT Marie-Line⁹, Julie Servel¹¹, Tiffany Busa¹², Bertille Bonniaud¹³, LYSE RUAUD¹⁴, BOURRAT Emmanuelle¹⁴, Juliette Dupont¹⁵, Erica Gerkes¹⁶, Roelineke J Lunsing¹⁷, Gareth Baynam¹⁸, Seema Lalani¹⁹, Ian Glass²⁰, Ashley Lahr⁴, Aimee Allworth²⁰, Haley Streff¹⁹, Irene Valenzuela Palafoll²¹, Marta Codina²¹, Philippe Kerschen²², sophie scheidecker¹, Elise Schaefer²³, Audrey Schalk¹, Nicolas Chassaing²⁴, Alice Goldenberg²⁵, Daphné Lehalle²⁶, Diana Rodriguez²⁷, jehanne martel²⁸, Pierre Vabres^28;29, Paul Kuentz³⁰, Juliette Piard³⁰, Arthur Sorlin³¹

¹CHU de Strasbourg - Hôpital Civil, Laboratoire de diagnostic génétique, Strasbourg, France; ²CHU de Besançon, Centre de Génétique Humaine, Besançon, France; ³CHU Amiens-Picardie, Centre d’activité de Génétique clinique et Oncogénétique, Amiens, France; ⁴UPMC Children’s Hospital of Pittsburgh, Pittsburgh, United States; ⁵Hospital Center University De Montpellier, Département Génétique Médicale, Maladies rares et médecine personnalisée, Montpellier, France; ⁶Hospital Robert Debré Ap-Hp, Département de génétique, Paris, France; ⁷Hospital Center University De Montpellier, Service Dermatologie, Montpellier, France; ⁸Hospital Center University De Lille, Service de génétique clinique, Lille, France; ⁹Hospital Center Félix Guyon, Saint-Denis, Reunion; ¹⁰C.h.u. De Saint-Pierre, Service de Génétique, Saint-Pierre, Reunion; ¹¹C.h.u. De Saint-Pierre, Plateforme de coordination des Maladies Rares Réunion-Mayotte, Saint-Pierre, Reunion; ¹²Marseille University Hospital Timone, Département de génétique médicale, Marseille, France; ¹³Chu Dijon, Dijon, France; ¹⁴Hospital Robert Debré Ap-Hp, Service de dermatologie, Paris, France; ¹⁵Hospital de Santa Maria - Urgências Centrais, Serviço de Genética, Lisboa, Portugal; ¹⁶University Medical Center Groningen, Department of clinical genetics, Groningen, Netherlands; ¹⁷University Medical Center Groningen, Department of Child Neurology, Groningen, Netherlands; ¹⁸King Edward Memorial Hospital, Subiaco, Australia; ¹⁹Baylor College of Medicine, Houston, United States; ²⁰Seattle Children’s Hospital, Seattle, United States; ²¹Vall d’Hebron University Hospital, Medical genetics department, Barcelona, Spain; ²²Centre Hospitalier de Luxembourg, Service de Neurologie, Luxembourg, Luxembourg; ²³Igma Institute Genetic Medical D’alsace, Service de génétique médicale, Strasbourg, France; ²⁴Hospital Paule De Viguier, Service de Génétique Médicale, Toulouse, France; ²⁵Hospital Center University De Rouen, Service de génétique, Rouen, France; ²⁶Hospital Armand Trousseau Ap-Hp, Service de Génétique, Paris, France; ²⁷Hospital Armand Trousseau Ap-Hp, Service de Neuropédiatrie, Paris, France; ²⁸Chu Dijon, Centre de référence Maladies Rares Génétiques à Expression Cutanée (MAGEC), Dijon, France; ²⁹Chu Dijon, Service de dermatologie, Dijon, France; ³⁰CHU de Besançon, Génétique biologique constitutionnelle et cytogénétique, Besançon, France; ³¹Laboratoire national de sante, National Center of Genetics, Dudelange, Luxembourg

Background/Objectives: RHOA-related mosaic ectodermal dysplasia (MIM:618727) was first described in 2019 in 7 patients exhibiting syndromic pigmentary mosaicism, resulting from a postzygotic variation in the RHOA gene. Three subsequent articles described the phenotype of six other patients.

Clinical features included linear hypopigmentation, hemihypotrophy with facial asymmetry, scarring alopecia, acral, ocular and teeth anomalies. Hearing impairment was occasionally associated. Brain imaging revealed consistent morphological abnormalities, described as non progressive comprising leukoencephalopathy, ventricular dilation and dilation of Virchow-Robin spaces, without cognitive impairment.

Identification of a RHOA mosaic pathogenic variation is based on high-depth sequencing using DNA extracted from a biopsy of affected skin.

The aim of our work is to define and extend the clinical phenotype associated with pathogenic RHOA mosaic variants and to elucidate a possible progressive brain deterioration.

Methods: We established an international collaboration to collect detailed clinical and molecular data, brain images and photographs from patients with RHOA-related mosaic ectodermal dysplasia.

Results: We report a clinical and molecular study of 18 new and 13 previously published patients, including the first fetal case. Our results show new frequent clinical signs, digestive anomalies (67% of the patients) and neurodevelopmental disorders (40%). Other newly described clinical manifestations are focal epilepsy, nystagmus and corpus callosum abnormalities. Brain MRI shows progressive worsening in 70% of cases.

Conclusion: RHOA-related mosaic ectodermal dysplasia is a clinically recognizable entity, with a possible neurodevelopmental, cerebral and digestive involvement. Diagnosing and monitoring older patients could help better anticipate the progression of cerebral involvement.

Grants: none

Conflict of Interest: None declared

C07.1 Deciphering 5’UTR variants in ENG revealed for the first time a uCUG-creating variant in patients with Hereditary Hemorrhagic Telangiectasia and associated with Endoglin deficiency.

Omar Soukarieh ^1;2, Clemence deiber², Caroline Meguerditchian², Carole Proust², Maud Tusseau³, Alexandre Guilhem³, Shirine Mohamed⁴, Béatrice Jaspard-Vinassa¹, Aurélie Goyenvalle⁵, Sophie Dupuis-Girod³, David-Alexandre Tregouët²

¹Inserm U1034, Pessac, France; ²Bordeaux Population Health, Bordeaux, France; ³Hospices Civils De Lyon, Lyon, France; ⁴Chru Brabois Umj, Vandœuvre-lès-Nancy, France; ⁵Paris-Saclay University, Orsay, France

Background/Objectives: Upstream Open Reading Frames (upORFs) located in the 5′UTR of coding genes are key regulators of translation. Their alteration by non-coding 5’UTR variants can be pathogenic. Recently, we characterized 5 single nucleotide variants (SNVs) creating AUGs (uAUG-SNVs) in the 5′UTR of ENG causing Hereditary Hemorrhagic Telangiectasia (HHT). The created uAUGs are all in frame with the same stop codon located in the coding sequence (ustop-c.125). We here extend our analysis to all SNVs altering upORFs in the 5’UTR of ENG.

Methods: We in silico mutated all positions of the 5’UTR of ENG (n = 909 SNVs) and annotated those creating canonical and non-canonical upstream translation initiation sites (uTIS), and/or creating/deleting stop codons, by using an updated version of the MORFEE tool. We also performed a functional assay based on the expression of constructs containing the wild-type and mutant ENG transcript in human cells on a selection of SNVs to assess their effect on ENG protein.

Results: MORFEE annotated 328 variants creating or modifying upORFs, among which 8 additional variants create uAUGs and 102 create non-canonical uTIS in frame with the ustop-c.125. We demonstrated that 6/8 of uAUG-SNVs decrease ENG levels. Moreover, one of these 6 variants and one uCUG-SNV (c.-76C > T) have been identified in 2 unrelated French HHT patients. This latter, reported in ClinVar as variant of unkown significance, moderately decreases ENG levels in our functional assay.

Conclusion: Our results contributed to the molecular diagnosis of HHT. Our strategy could be applied in any other gene implicated in human diseases.

Grants: Lefoulon Delalande, ANR.

Conflict of Interest: None declared

C07.2 Somatic loss-of-function PIK3R1 and activating non-hotspot PIK3CA mutations associated with Capillary Malformation with Dilated Veins (CMDV)

Martina De Bortoli¹, Angela Queisser¹, Van Cuong Pham², Anne Dompmartin-Blanchere³, Raphaël Helaers¹, Simon Boutry^1;4, Cathy Claus⁵, An-Katrien De Roo^5;6;7, Frank Hammer⁸, Pascal Brouillard¹, Salim Abdelilah-Seyfried², Laurence Boon^1;5, Miikka Vikkula ^1;9;10

¹De Duve Institute UCLouvain, Human Molecular Genetics, Woluwe-Saint-Lambert, Belgium; ²Institute of Biochemistry and Biology, Postdam, Germany; ³Caen-Normandy University, Department of Dermatology, Caen, France; ⁴Interuniversity Institute Of Bioinformatics, Brussels, Belgium; ⁵Center for Vascular Anomalies Saint-Luc UCLouvain, Division of Pathology, Brussels, Belgium; ⁶University Clinics Saint-Luc, Service d’anatomopathologie, Brussels, Belgium; ⁷Institute of Experimental and Clinical Research UCLouvain, Brussels, Belgium; ⁸University Clinics Saint-Luc, Department of Medical Imaging, Brussels, Belgium; ⁹Cliniques universitaires Saint-Luc (UCLouvain), Center for Vascular Anomalies, Bruxelles, Belgium; ¹⁰WEL Research Institute, WELBIO department, Wavre, Belgium

Consortium: Revamp - Leducq network

V.A. cure ITN network

Background/Objectives: Common capillary malformations (CMs) are red vascular skin lesions, most commonly associated with somatic activating GNAQ or GNA11 mutations. We focused on CMs lacking such a mutation to identify novel genetic causes.

Methods: We used targeted next-generation-sequencing on 82 lesions. Bioinformatic analysis was performed with Highlander to pinpoint likely pathogenic variants. Primary cells were isolated from two lesions and PI3K-AKT-mTOR and RAS-RAF-MAPK signaling were assessed by western-blot. Zfish was used to model lack of PIK3R1 and effects of identified likely pathogenic variants.

Results: We identified 9 somatic pathogenic variants in PIK3R1 and PIK3CA, encoding for the regulatory and catalytic subunits of the PI3K kinase. Re-characterization of the lesions unraveled a common phenotype: a pale Capillary Malformation associated with visible Dilated Veins (CMDV). Primary-endothelial cells from PIK3R1-mutated lesions unveiled an abnormal increase in AKT phosphorylation, effectively reduced by mTOR, AKT and PIK3CA-inhibitors. Endothelium-specific expression of PIK3R1 mutants in zebrafish embryos resulted in abnormal development of the posterior capillary-venous plexus.

Conclusion: CMDV emerges as a clinical entity associated with somatic pathogenic variants in PIK3R1 or PIK3CA (non-hotspot). The activated AKT signaling can be effectively reversed by PI3K-pathway inhibitors. The zebrafish model holds promise as a valuable tool for future drug screens aimed at developing patient-tailored treatments.

Grants: Fonds de la Recherche Scientifique - FNRS Grants T.0240.23, P.C005.22, T.0146.16 and P.C013.20; Fund Generet managed by the King Baudouin Foundation (Grant 2018-J1810250-211305); Walloon Region through the FRFS-WELBIO strategic research programme (WELBIO-CR-2019C-06); European Union’s Horizon 2020 research and innovation programme under grant agreement No 874708 (Theralymph).

Conflict of Interest: None declared

C07.3 SHOX2-dependent arrhythmias: Mechanistic insights from human iPSC-derived cardiomyocytes

Sandra Hoffmann ¹, Kristin Rädecke¹, Annette Löwen¹, Simon A. Sumer¹, Micha Böckers¹, Fangfang Zhang², Birgit Campbell², Amelie Paasche³, Christian Goetz³, Sarah Huecker⁴, Adrian Chang⁵, Volker Eckstein⁶, Tatjana Dorn², Christoph Dieterich⁵, Stefan Kirsch⁴, Christian Schaaf¹, Constanze Schmidt³, Alessandra Moretti², Gudrun Rappold¹

¹Institute of Human Genetics, Heidelberg, Germany; ²Klinikum rechts der Isar der Technischen Universität München, First Department of Medicine, Cardiology, München, Germany; ³University Hospital Heidelberg, Heidelberg Centre for Heart Rhythm Disorders, Heidelberg, Germany; ⁴Fraunhofer-Institute for Toxicology and Experimental Medicine, Division of Personalized Tumor Therapy, Regensburg, Germany; ⁵Klaus Tschira Institute for Integrative Computational Cardiology, Section of Bioinformatics and Systems Cardiology, Heidelberg, Germany; ⁶University Hospital Heidelberg, FACS Core Facility, Department of Medicine V, Heidelberg, Germany

Background/Objectives: The orchestrated contraction of cardiac muscle relies on regulatory networks, where transcription factors guide downstream effectors for proper cardiac conduction. The transcription factor SHOX2 plays a pivotal role in the development and function of the sinoatrial node, emerging as a key pacemaker gene linked to conduction-related diseases. With the identification of SHOX2 variants in patients with early-onset atrial fibrillation and sinus node dysfunction, these variants have undergone functional characterization in various animal model systems.

Methods: To simulate SHOX2-related conduction dysfunction in a human system, iPSCs were generated from two different atrial fibrillation SHOX2-deficient patients, along with isogenic control lines. Differentiation of these cells into nodal and atrial cardiomyocytes enabled a comprehensive analysis using gene expression profiling, scRNA sequencing, and electrical phenotyping to uncover variant-specific changes contributing to the disease phenotype.

Results: Our data revealed deregulated gene expression patterns and impaired differentiation capability in patient cells. Furthermore, action potential characteristics in the patient lines showed changes attributed to differential ion channel expression, particularly deregulation of Na⁺ and K⁺ channels, unveiling novel disease mechanisms and therapeutic targets.

Conclusion: The SHOX2 human iPSC model mirrors the phenotypes observed in patients with atrial fibrillation and Shox2 animal models. This allowed us to gain valuable insights into the underlying molecular mechanisms that are affected by SHOX2 variant carriers and influence the occurrence and progression of conduction disorders.

Grants: Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) – 493726818 (to S.H.) and the Ministerium für Ernährung, ländlicher Raum und Verbraucherschutz, Baden Württemberg – 16(34)-0802/0415E (to S.H.).

Conflict of Interest: None declared

C07.4 Utility of genome sequencing and group-enrichment to support splice variant interpretation in Marfan syndrome

Alistair Pagnamenta ¹, Susan Walker², Dave Bunyan³, N Simon Thomas³, John Holloway⁴, Htoo Wai⁴, Dimitris Vavoulis¹, Chris Kershaw⁵, Siddharth Banka⁵, Ed Blair⁶, Dahlia HopMeier⁷, Stephanie L. Curtis⁸, Catherine Fielden⁹, Julie Evans⁹, Rebecca Whittington⁹, Sarah Smithson¹⁰, Helen Cox¹¹, Paul Clift¹², Edward Sanders¹³, Meriel McEntagart¹⁴, Matina Prapa¹⁴, Suzanne Alsters¹⁴, Deborah Morris-Rosendahl¹⁴, John Dean¹⁵, Vivienne McConnell¹⁶, Abhijit Dixit¹⁷, Claire Turner¹⁸, Sian Ellard¹⁸, Emma Baple¹⁸, Diana Baralle⁴, ian berry⁹, Jenny Taylor¹

¹University of Oxford, Centre for Human Genetics, Oxford, United Kingdom; ²Genomics England, London, United Kingdom; ³Salisbury NHS Foundation Trust, Wessex Regional Genetics Laboratory, Salisbury, United Kingdom; ⁴University of Southampton, Human Development and Health, Faculty of Medicine, Southampton, United Kingdom; ⁵Manchester University Hospitals NHS Foundation Trust, Manchester Centre for Genomic Medicine, Manchester, United Kingdom; ⁶Oxford University Hospitals NHS Foundation Trust, Oxford Centre for Genomic Medicine, Oxford, United Kingdom; ⁷Chapel Allerton Hospital, Leeds, United Kingdom; ⁸University Hospitals Bristol and Weston NHS Foundation Trust, Bristol Heart Institute, Bristol, United Kingdom; ⁹North Bristol NHS Trust, Bristol Genetics Laboratory, Bristol, United Kingdom; ¹⁰University Hospitals Bristol NHS Foundation Trust, Department of Clinical Genetics, Bristol, United Kingdom; ¹¹Birmingham Women’s and Children’s Hospital, West Midlands Clinical Genetics Service, Birmingham, United Kingdom; ¹²Queen Elizabeth Hospital, Grown Up Congenital Heart Disease Unit, Birmingham, United Kingdom; ¹³University of Oxford, MRC Weatherall Institute of Molecular Medicine, Oxford, United Kingdom; ¹⁴St George’s University Hospitals NHS Foundation Trust, SW Thames Centre for Genomic Medicine, London, United Kingdom; ¹⁵NHS Grampian, Department of Medical Genetics, Aberdeen, United Kingdom; ¹⁶Belfast Health and Social Care Trust, Regional Genetics Service, Belfast, United Kingdom; ¹⁷Nottingham University Hospitals NHS Trust, Department of Clinical Genetics, Nottingham, United Kingdom; ¹⁸Royal Devon University Healthcare NHS Trust, Department of Clinical Genetics, Exeter, United Kingdom

Background/Objectives: Marfan syndrome (MFS) is a connective tissue disorder associated with skeletal abnormalities, lens dislocation and Hereditary Thoracic Aortopathy (HTA). Pathogenic/likely pathogenic FBN1 variants are found in >90% of affected individuals and as the gene comprises 66 exons, splicing variants represent a significant fraction of these.

Methods: To better quantify the impact of aberrant FBN1 splicing in MFS we performed a systematic analysis of ultra-rare variants using short-read data from the 100k genomes project.

Results: Utilising aggregate data for 78,195 individuals, 13,864 singleton SNVs in FBN1 were identified of which 21 were predicted to impact splicing (SpliceAI>0.5). Incidence in individuals recruited under HTA (9/703) was significantly elevated compared to that seen in non-HTA participants (12/77,492; OR = 84, p = 9.7x10-14), a finding which prompted several variants to be re-assessed. Although RT-PCRs for an individual harbouring c.1961-3T > G previously showed no splicing effect, repeating this with intronic primers detected a 133bp exon extension, (p.Asp654AlafsTer17). Another variant (c.-182 + 1G > A) predicted exon1 extension and creates an uAUG. Examples of SNVs that simultaneously alter transcription/translation are rare. Additional analyses uncovered a further 14 families harbouring 11 different FBN1 splice variants. 14/20 variants lay beyond the +/-8 splice regions and 4 families harboured the previously reported recurrent c.7820-2034A > G.

Conclusion: Together with reports of cryptic SVs disrupting FBN1, our findings highlight the importance of incorporating the analysis of FBN1 introns into first-line MFS testing. The enrichment seen is an exemplar that we hope will stimulate further discussions about how statistical support for specific variant classes should be considered as part of interpretation frameworks.

Grants: MRC (MR/W01761X/1)

Conflict of Interest: Alistair Pagnamenta: None declared, Susan Walker Employed by Genomics England, Dave Bunyan: None declared, N Simon Thomas: None declared, John Holloway: None declared, Htoo Wai: None declared, Dimitris Vavoulis: None declared, Chris Kershaw: None declared, Siddharth Banka: None declared, Ed Blair: None declared, Dahlia HopMeier: None declared, Stephanie L. Curtis: None declared, Catherine Fielden: None declared, Julie Evans: None declared, Rebecca Whittington: None declared, Sarah Smithson: None declared, Helen Cox: None declared, Paul Clift: None declared, Edward Sanders: None declared, Meriel McEntagart: None declared, Matina Prapa: None declared, Suzanne Alsters: None declared, Deborah Morris-Rosendahl: None declared, John Dean: None declared, Vivienne McConnell: None declared, Abhijit Dixit: None declared, Claire Turner: None declared, Sian Ellard: None declared, Emma Baple: None declared, Diana Baralle: None declared, ian berry: None declared, Jenny Taylor: None declared

C07.5 A proteome-wide Mendelian randomization study identifies potentially causal associations with ischemic stroke

Lazaros Belbasis ¹, Jemma Hopewell¹, Cornelia M. van Dujin¹

¹University of Oxford, Nuffield Department of Population Health (NDPH), Oxford, United Kingdom

Consortium: N/A

Background/Objectives: Identification of novel drug targets for ischemic stroke (IS) is important to tackle its increasing global burden. Proteins participate in multiple biological processes and serve as drug targets. We aim to identify proteins associated with IS using a Mendelian randomisation approach.

Methods: We identified the lead cis protein quantitative trait locus (pQTL) for each plasma protein in the two largest proteo-genomic studies in populations of European ancestry (Olink platform in the UK Biobank and SomaScan platform in the deCODE Health Study). We retrieved the genetic effect of the selected pQTLs on IS from the latest GWAS from GIGASTROKE consortium. We used the Wald ratio to estimate the causal effects of proteins on IS and assessed the confounding by linkage disequilibrium through co-localisation.

Results: We tested 2,671 proteins for an association with IS, and 18 were statistically significant at a 5% false discovery rate. Sixteen proteins showed evidence of co-localisation with IS. ABO, F11, FURIN, KNG1, LRP4, FGF5, TMPRSS5, and WARS were associated with a higher IS risk. GRK5, MMP12, MMUT, CEP85, KIAA0319, PROCR, SH3BGRL3 and TMEM106B were associated with lower IS risk.

Conclusion: We provided a comprehensive analysis of the effect of plasma proteome on IS, with two of the identified proteins (F11 and PROCR) being tested in clinical trials. Our additional work considering IS subtypes and vascular risk factors further elucidates the potentially causal role of proteins in stroke aetiology and its mediating mechanisms.

Grants: The study is supported by an Oxford BHF–CRE Intermediate Transition Research Fellowship.

Conflict of Interest: Lazaros Belbasis Lazaros Belbasis is supported by an NDPH Early Career Research Fellowship (until July 2024) and a BHF-CRE Intermediate Transition Research Fellowship after August 2024., Jemma Hopewell: None declared, Cornelia M. van Dujin: None declared

C07.6 Enhancing 10-year cardiovascular disease prediction using polygenic risk scores and metabolomic biomarker scores

Scott C. Ritchie ¹, Xilin Jiang¹, Lisa Pennells¹, Yu Xu¹, Claire Coffey¹, Yang Liu¹, Praveen Surendran¹, Savita Karthikeyan¹, Samuel Lambert¹, John Danesh¹, Adam Butterworth¹, Angela Wood¹, Stephen Kaptoge¹, Emanuele Di Angelantonio¹, Michael Inouye¹

¹University of Cambridge, Public Health and Primary Care, Cambridge, United Kingdom

Background/Objectives: Polygenic risk scores (PRS) and metabolomic biomarker scores have shown promise for improving cardiovascular disease (CVD) prediction, but have not yet been evaluated in the context of European-recommended clinical risk factor prediction models (SCORE2) and guidelines for statin initiation for 10-year CVD prevention.

Methods: PRS were compared and combined with metabolomics scores in 168,517 UK Biobank participants (5,096 non-fatal and fatal CVD over 10-years of follow-up) 40–69 years of age without previous CVD, diabetes, or statin treatment. Improvement over SCORE2 in risk discrimination was assessed using Harrel’s C-index and stratification at clinically relevant risk thresholds using categorical net reclassification. Population modelling was subsequently applied to estimate the impact on 10-year CVD prevention if applied at scale.

Results: Risk discrimination of SCORE2 (C-index: 0.718) was similarly improved by both PRSs (ΔC-index 0.009 [0.007–0.012]) and metabolomics biomarker scores (ΔC-index: 0.011 [0.009–0.014]); capturing orthogonal information that combined yielded ΔC-index 0.019 (0.016–0.022) for 10-year CVD. Concomitant improvements in risk stratification following guidelines for statin initiation were observed, with PRSs and NMR scores yielding a combined improvement in net case reclassification of 13.04% (11.67–14.41%). Applied in population modelling, the number of CVD events prevented per 100,000 screened increased from 201 to 370 (ΔCVD_prevented: 170 [158–182]) while essentially maintaining the number of statins prescribed per CVD event prevented.

Conclusion: Combining PRSs and metabolic biomarker scores with SCORE2 moderately enhances prediction of first-onset CVD, and could have substantial population health benefit.

Grants: British Heart Foundation (RG/18/13/33946)

Conflict of Interest: Scott C. Ritchie: None declared, Xilin Jiang: None declared, Lisa Pennells: None declared, Yu Xu: None declared, Claire Coffey: None declared, Yang Liu: None declared, Praveen Surendran During the course of this project Dr. Praveen Surendran became a full-time employee of GSK Plc. All significant contributions to this study were made prior to this role and GSK Plc had no input to the study., Savita Karthikeyan: None declared, Samuel Lambert: None declared, John Danesh Professor John Danesh has received multiple grants from academic, charitable and industry sources outside of the submitted work., Professor John Danesh serves on scientific advisory boards for AstraZeneca, Novartis, and UK Biobank., Adam Butterworth Professor Adam S. Butterworth reports institutional grants from AstraZeneca, Bayer, Biogen, BioMarin, Bioverativ, Novartis, Regeneron and Sanofi., Angela Wood: None declared, Stephen Kaptoge: None declared, Emanuele Di Angelantonio: None declared, Michael Inouye: None declared

C08.1 Perspectives of preimplantation genetic testing users in Belgium on the ethics of PGT for polygenic conditions

Maria Siermann ^1;2, Joris Robert Vermeesch³, Taneli Raivio², Arne Vanhie^4;5, Karen Peeraer⁶, Olga Tsuiko^3;7, Pascal Borry¹

¹KU Leuven, Department of Public Health and Primary Care, Centre for Biomedical Ethics and Law, Leuven, Belgium; ²University of Helsinki, Faculty of Medicine, Department of Physiology, Helsinki, Finland; ³KU Leuven, Department of Human Genetics, Leuven, Belgium; ⁴University Hospital Leuven, Department of Gynaecology and Obstetrics, Leuven University Fertility Centre, Leuven, Belgium; ⁵KU Leuven, Group Biomedical Sciences, Department of Development and Regeneration, Endometriosis and Reproductive Medicine (LEERM), Laboratory of Endometrium, Leuven, Belgium; ⁶KU Leuven, Department of Development and Regeneration, Leuven, Belgium; ⁷University Hospital Leuven, Center for Human Genetics, Reproductive Genetics Unit, Leuven, Belgium

Background/Objectives: Preimplantation genetic testing using polygenic risk scores (PGT-P) is an ethically controversial technology that aims to screen embryos for risk of developing polygenic conditions. It is crucial to investigate ethical considerations and the perspectives of PGT users regarding ethics of embryo selection procedures.

Methods: We performed a qualitative interview study with PGT-M/SR users in Belgium, consisting of 18 interviews (10 couples, 8 women, n = 28). The aim was to get insights into the viewpoints of PGT-M/SR users on the ethics of PGT-P.

Results: Participants emphasized the physical, psychological and practical difficulties of their PGT-M/SR trajectories. Because of these burdens and the probabilistic nature of PGT-P, most participants did not find PGT-P an attractive option. They were largely negative about PGT-P without indication and about knowing risk information. PGT-P was seen as introducing unnecessary worries, complex choices and additional responsibilities. Many participants felt that PGT-P was going too far and that not everything should be controlled. If PGT-P would be implemented, participants generally wanted it to be limited to serious familial genetic conditions and decision-making to be left to healthcare professionals.

Conclusion: Our findings demonstrate that, from the perspective of those with experience with PGT-M/SR, PGT-P is seen as distinctly different from PGT-M/SR and as ethically contentious. It is important that these viewpoints are considered with regards to potential PGT-P implementation and guidelines.

Grants: This project has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 813707.

Conflict of Interest: None declared

C08.2 “This is going to change everything”: A mixed-methods study examining outcomes of reproductive genetic carrier screening within the first 12-months post-result

Erin Tutty ^1;2, Alison Archibald^1;2;3, Nigel Laing^4;5;6, Edwin Kirk^7;8;9, Belinda McClaren^1;2, Sharon Lewis^2;10, Martin Delatycki^2;3;11

¹Murdoch Children’s Research Institute, Genomics in Society, Parkville, Australia; ²The University of Melbourne, Department of Paediatrics, Parkville, Australia; ³Victorian Clinical Genetics Services, Parkville, Australia; ⁴Harry Perkins Institute of Medical Research, Nedlands, Australia; ⁵University of Western Australia, School of Biomedical Sciences, Crawley, Australia; ⁶PathWest Laboratory Medicine, Department of Diagnostic Genomics, Nedlands, Australia; ⁷Sydney Children’s Hospital, Centre for Clinical Genetics, Randwick, Australia; ⁸NSW Health Pathology Randwick, Genomics Laboratory, Randwick, Australia; ⁹University of New South Wales, School of Clinical Medicine, Sydney, Australia; ¹⁰Murdoch Children’s Research Institute, Reproductive Epidemiology, Parkville, Australia; ¹¹Murdoch Children’s Research Institute, Bruce Lefroy Centre, Parkville, Australia

Background/Objectives: Responsibly implementing reproductive genetic carrier screening (RGCS) involves promoting reproductive autonomy and psychological adaption to results. To inform implementation, this study explored psychosocial and reproductive outcomes from an Australian project (‘Mackenzie’s Mission’) at 12-months post-result.

Methods: Mackenzie’s Mission offered RGCS to >10,000 reproductive couples, 9107 had RGCS. A survey captured psychosocial (reproductive confidence, anxiety, and regret) and reproductive outcomes. Data were analysed descriptively. Participants with an increased chance result were interviewed. Data were analysed using reflexive thematic analysis.

Results: 4984 individuals completed the survey, including 116 with an increased chance (representing 96 couples, response = 55.0%). Most had high reproductive confidence and minimal regret. Anxiety was elevated for increased chance participants (median = 38.3, scores 40-80 = clinically meaningful). 53.1% of increased chance couples felt results impacted the timing/method of conception. 30/96 were pregnant at result disclosure; 16 had prenatal diagnosis. 30 conceived post-result; 17 used pre-implantation genetic testing. Interviewees (N = 19) felt their result “change[d] everything” about their reproductive plans. Reproductive decision-making was an evolving “journey”. Making a decision increased reproductive confidence and aided psychological adaptation. Emotions came in “stages” based on participants’ reproductive pathway. For example, those using pre-implantation genetic testing expressed anxious anticipation to conceive. Most felt “grateful” for the result.

Conclusion: By 12-months post-result, people have found value in their result given the reproductive confidence and autonomy it provides. Raised anxiety while people with an increased chance make and enact reproductive decisions suggests the need for longitudinal care to promote long-term well-being.

Grants: Medical Research Future Fund (GHFM73390), Research Training Program scholarship.

Conflict of Interest: Erin Tutty: None declared, Alison Archibald Employed by VCGS, a not-for-profit provider of genetic testing., Nigel Laing Employed by PathWest, a genetic testing provider, Edwin Kirk Employed by NSW Health Pathology, a service that provides reproductive genetic carrier screening on a fee-for-service basis, and employed part time by Genea, a fertility company that refers patients for reproductive genetic carrier screening., Belinda McClaren: None declared, Sharon Lewis: None declared, Martin Delatycki Employed by VCGS, a not-for-profit provider of genetic testing.

C08.3 Defining next steps in the implementation of polygenic scores in Australia’s cancer genetics clinics: Professionals’ views on determinants, strategies and needs

Rebecca Purvis ^1;2, Mary-Anne Young^3;4, Natalie Taylor⁵, Paul James^1;2, Laura Forrest^1;2

¹Peter MacCallum Cancer Centre and the Royal Melbourne Hospital, Parkville Familial Cancer Centre, Melbourne, Australia; ²The University of Melbourne, Sir Peter MacCallum Department of Oncology, Melbourne, Australia; ³Garvan Institute of Medical Research, Clinical Translation and Engagement Platform, Sydney, Australia; ⁴UNSW Medicine and Health, St Vincent Clinical Campus, School of Clinical Medicine, Sydney, Australia; ⁵University of New South Wales, School of Population Health, Sydney, Australia

Background/Objectives: Polygenic scores (PGS) capture a proportion of the genomic liability for cancer. However, there are significant evidence gaps regarding clinical implementation, with scant evidence regarding determinants and strategies. The study objective was to explore genetic healthcare professionals’ views on the clinical implementation of PGS in cancer genetics clinics, including expected barriers, enablers, and priorities.

Methods: We conducted multi-modal, semi-structured interviews with genetic healthcare professionals purposively sampled using professional networks across Australia. A process map was used as a co-design tool to encourage engagement. Deductive content analyses were conducted, informed by the Consolidated Framework for Implementation Research 2022 and the Theoretical Domains Framework.

Results: 27 professionals participated across sector roles. Participants were majority female (85.2%) with 14.6 years of experience on average (range 0.5-35 years). Participants raised concerns regarding compatibility with current care pathways, the relative advantage, cost-effectiveness and complexity of PGS, capability of deliverers and equity of access. Having a defined model of care supported by national guidelines, standardised reporting, stakeholder education and resources were top implementation priorities. Whilst professionals highlighted key outcomes, such as quick turnaround times, as indicators of implementation success, few discussed defined implementation strategies.

Conclusion: Overall, professionals were optimistic about the clinical implementation of PGS. There was complementation between professionals‘ implicit understanding of determinants and implementation frameworks, however, the steps towards clinical implementation in the sector were poorly defined. Leaders can utilitse prioritised determinants and outcome indicators to direct research resourcing and better define future implementation outcomes.

Grants: Peter Mac Foundation Lester Peters Award and an Australian Government RTP Scholarship.

Conflict of Interest: None declared

C08.4 Feedback from two major national studies involving over 2500 individuals on the issue of secondary findings from exome or genome sequencing

Eléonore Viora-Dupont ^1;2, Françoise Robert³, Benedicte Gerard⁴, Aline Chassagne^5;6, Aurore pelissier^6;7, Marion Bouctot⁸, Marie-Laure Humbert⁸, Dominique Salvi⁷, Elodie Gauthier^2;6, Stephanie Staraci^9;10, Nicolas Meunier-Beillard^6;8, Amandine Cadénes³, Myrtille Spetchian¹¹, Baurand Amandine², Caroline Sawka², Geoffrey Bertolone², Delphine Heron¹¹, Boris Keren¹¹, Damien Sanlaville³, Gaetan Lesca^3;12, Christel Thauvin-Robinet^1;2, Christophe Philippe¹, Sylvie Odent^13;14, Dominique Bonneau¹⁵, Didier LACOMBE¹⁶, Roseline Caumes¹⁷, Patrick Edery^3;18, Sabine Sigaudy¹⁹, David Geneviève²⁰, Bertrand Isidor^21;22, Gael Nicolas^23;24, Julien Thevenon²⁵, Catherine Lejeune^6;8, Christine Peyron^6;7, Hélène Dollfus⁴, Marcela Gargiulo^26;27, Christine Binquet^6;8, Laurence Faivre^1;2

¹Hospital Center University Dijon Bourgogne, FHU TRANSLAD et équipe GAD INSERM UMR 1231, Dijon, France; ²Hospital Center University Dijon Bourgogne, Centre de Référence Anomalies du Développement et Syndromes Malformatifs Est, Dijon, France; ³Hospices civils de Lyon, Service de génétique clinique, Centre de Référence Anomalies du Développement et Syndromes Malformatifs Sud Est, Bron, France; ⁴CHU Strasbourg, Laboratoire de Biologie Moléculaire, Strasbourg, France; ⁵University of Burgundy Franche-Comté, Laboratory of Sociology and Anthropology (LaSA, EA3189), Besançon, France; ⁶Hospital Center University Dijon Bourgogne, FHU TRANSLAD, Dijon, France; ⁷University of Burgundy Franche-Comté, Laboratory of economy (LEDI), EA7467, Dijon, France; ⁸Hospital Center University Dijon Bourgogne, INSERM, CIC1432, module épidémiologie Clinique, Dijon, France; ⁹University of Paris, Sorbonne Paris City, Clinical Psychology Laboratory, Psychopathology, Psychoanalysis (EA4056, ED 261), Paris, France; ¹⁰GH APHP, Genetics Department, Reference Center for Hereditary Cardiac Disorders, Paris, France; ¹¹APHP, Centre de Référence Anomalies du Développement et Syndromes Malformatifs, Paris, France; ¹²Univ Lyon, Univ Lyon 1, CNRS, INSERM, Physiopathologie et Génétique du Neurone et du Muscle, UMR5261, U1315, Institut NeuroMyoGène, Lyon, France; ¹³Université de Rennes, ERN Ithaca, Institut de Génétique et Développement de Rennes, CNRS, UMR 6290, Rennes, France; ¹⁴CHU de Rennes, Service de Génétique Clinique, Centre labélisé pour les Anomalies du Développement ouest, Rennes, France; ¹⁵University Hospital of Angers, Biochemistry and Genetics Department, Angers, France; ¹⁶CHU de Bordeaux, Reference Center AD SOOR, AnDDI-RARE, INSERM U 1211, Medical Genetics Department, Bordeaux, France; ¹⁷CHU de Lille, Clinique de Génétique, Lille, France; ¹⁸University of Claude Bernard Lyon 1, INSERM U1028, CNRS UMR5292, CRNL, GENDEV Team, Bron, France; ¹⁹Hôpital Timone Enfant, Département de Génétique Médicale, Marseille, France; ²⁰CHU Montpellier, Reference Center for Rare Disease, Medical Genetic Department for Rare Disease and Personalize Medicine, Montpellier, France; ²¹CHU Nantes, Service de Génétique Médicale, Nantes, France; ²²Université de Nantes, CNRS, INSERM, L’Institut du Thorax, Nantes, France; ²³Université de Rouen Normandie, Inserm U1245, Rouen, France; ²⁴CHU Rouen, Department of Genetics, Rouen, France; ²⁵CHU de Grenoble, Service de génétique médicale, Grenoble, France; ²⁶Université Paris Cité, Laboratoire de Psychologie Clinique, Psychopathologie, Psychanalyse, Boulogne-Billancourt, France; ²⁷Pitié-Salpêtrière University Hospital, Institute of Myologie, Paris, France

Consortium: FIND Consortium and DEFIDIAG-DS Consortium

Background/objectives: In the context of global debates on reporting additional findings (AF) such as incidental or secondary findings, particularly in Europe and France, our team has conducted extensive studies. The FIND study, involving 340 patients across three centers and using exome sequencing, examined patient preferences and the medical, organizational, and psychological challenges associated with reporting three categories of SF: “actionable” diseases, genetic counseling, and pharmacogenetics. After this first study, the DEFIDIAG-DS study, involving 2150 participants across 13 centers and using genome sequencing, delved into the impact of medical discourse and the decision-making process of parents concerning their health in the context of one category of “actionable” SF.

Methods: Mixed methodology combining quantitative and qualitative approaches over a longitudinal period of up to 1 to 3 years post-results disclosure, including interviews and specific questionnaires.

Result: Both studies revealed high patient satisfaction and a lack of regret, despite potential psychological impacts, especially in situations of vulnerability. Patients emphasized the importance of separate information and reporting sessions for primary diagnosis (PD) and AF within the same day. Coordination by geneticists in their care pathway with specialists was also underscored.

Conclusion: the results of these studies provide valuable insights into the perception and impact of AF: importance of information, reporting organization and care pathways coordination. Emphasizing the importance of multidisciplinary support for “opportunistic presymptomatic diagnosis,” these data underline the challenging nature of anticipating effects on patients, given that their primary motivation remains the search for an etiological diagnosis for PD and potential actionability.

Grant: FIND: PREPS (DGOS); DEFIDIAG-DS: DEFIDIAG (PFMG)

Conflict of Interest: None declared

C08.5 Opportunistic genomic screening in acute care settings: pilot and process evaluation of a two-step model

Melissa Martyn ^1;2;3, Ling Lee^1;2, Alli Jan^2;4, Chole Mighton^5;6, Sophie Bouffler⁷, Marc Clausen^5;6, Yvonne Bombard^5;6, Clara Gaff^1;2;3, Zornitza Stark^3;7;8

¹Melbourne Genomics Health Alliance, Melbourne, Australia; ²Murdoch Children’s Research Institute, Parkville, Australia; ³The University of Melbourne, Department of Paediatrics, Faculty of Medicine, Dentistry and Health Sciences, Melbourne, Australia; ⁴NSW Health, Central Coast Local Health District, Gosford, Australia; ⁵St. Michael’s Hospital, Unity Health Toronto, Genomics Health Services Research Program, Toronto, Canada; ⁶University of Toronto, Institute of Health Policy, Management and Evaluation, Toronto, Canada; ⁷Australian Genomics, Parkville, Australia; ⁸Victorian Clinical Genetics Services, Parkville, Australia

Background/Objectives: Opportunistic genomic screening (OGS) is the analysis of clinical genomic data for additional health information. OGS has primarily been offered concurrent with diagnostic testing. An alternative approach is needed in acute care, where parents of children undergoing ultra-rapid sequencing report ‘feeling overwhelmed’. The Acute Care Genomics study piloted and evaluated an alternative approach.

Methods: Families (n = 232) were offered OGS after return of ultra-rapid diagnostic sequencing results: adult-onset for parents; childhood onset for (living) children; couple carrier screening. Parents had access to an online patient platform ‘Genetics Adviser’, modified for this study, followed by genetic counselling. Surveys were administered at 3 timepoints. Genetic counselling sessions were recorded and transcribed. Descriptive statistic and qualitative content analyses of Genetics Adviser utilisation, survey and transcript data focused on uptake and decision-making, understanding and service delivery preferences.

Results: Uptake of OGS was 40-42% across categories. Couple screening was an easier decision than paediatric or adult OGS (p = 0.007). Genetics Adviser users had higher knowledge scores than non-users, before and after counselling (p < 0.02). Differences in counselling sessions were noted, with less information seeking/giving and value clarification for Genetics Adviser users. After results return, 54% supported delaying OGS; 18% preferred a concurrent offer.

Conclusion: Two-step OGS in acute care is acceptable to parents. Couple screening is an easier decision than paediatric or adult OGS; concurrent couple screening merits further investigation. Genetics Advisor improves knowledge and impacts counselling; further research into the potential role of online patient platforms in OGS is warranted.

Grants: Medical Research Future Fund: GHFM76747. Royal Children’s Hospital Foundation: 2020-1259

Conflict of Interest: Melissa Martyn: None declared, Ling Lee: None declared, Alli Jan: None declared, Chole Mighton: None declared, Sophie Bouffler: None declared, Marc Clausen Genetics Adviser, Inc (part owner), Genetics Adviser Creative Director (part-time), Yvonne Bombard Genetics Adviser, Inc, Canadian Institutes of Health Research, Canadian cancer society, Ontario institute of Cancer research, Terry fox research institute, McLaughlin Centre of university of Toronto, Canada Research Chairs Program, Takeda, ASCO, ACMG, Clara Gaff: None declared, Zornitza Stark: None declared

C08.6 Challenges and legislation influencing genetic counselling practice in EU Member States

J. Matt McCrary ¹, Denis Horgan², Els Van Valckenborgh³, Marc Van Den Bulcke³, Anke Katharina Bergmann¹

¹Hannover Medical School, Department of Human Genetics, Hannover, Germany; ²European Alliance of Personalized Medicine, Brussels, Belgium; ³Sciensano, Cancer Center, Department of Epidemiology and Health, Bussels, Belgium

Consortium: CAN.HEAL Consortium

Background/Objectives: Integration of genetic testing into cancer care requires expert genetic counselling, a human interface point between patients and complex genetic technologies. Genetic counselling facilitates informed decision-making by patients and their families. However, its delivery is complicated by growing healthcare workforce shortages coinciding with an increasing inclusion of genetic testing in cancer prevention, diagnosis, and care. This project aims to provide an overview of genetic counselling practice, challenges, and legislation across EU Member States and a foundation for consensus European recommendations for delivering sustainable and equitable genetic counselling.

Methods: National legislative databases were searched in all 27 Member States. Interviews with relevant experts (medical geneticists, genetic counsellors, oncologists) and an online survey of cancer patient organizations from each EU Member State provide detailed insights into genetic counselling practice in each country.

Results: Legislation mandating genetic counselling exists in 22 of 27 Member States, albeit with significant variation in language addressing the ‘who, what, when, and how’ of counselling. Practice is similarly varied. Key barriers to counselling delivery noted by health professionals were remarkably consistent: genetic literacy of patients and non-geneticist health professionals (all 27 Member States); workforce capacity (25 of 27 Member States). Collection of data regarding key barriers from the patient perspective is ongoing, and will be presented at the conference.

Conclusion: This project highlights substantial variability in genetic counselling legislation and practice across the EU, but common challenges. Recommendations will focus on opportunities for collective European action to address these common challenges.

Grants: European Union Grant: 101080009

Conflict of Interest: None declared

C09.1 Clinical pharmacokinetic study of novel CYP2C19 and CYP2D6 alleles identified by genome and exome sequencing in the Estonian Biobank

Jana Lass^1;2, Laura Birgit Luitva¹, Anette Caroline Kõre³, Raul Kokassaar⁴, Kristi Krebs¹, Elisabet Størset⁵, Kadri Maal¹, Magnus Ingelman-Sundberg⁶, Espen Molden⁵, Alar Irs⁷, Kersti Oselin⁴, Lili Milani ^1;8

¹Estonian Genome Centre, Institute of Genomics, University of Tartu, Tartu, Estonia; ²Department of Pharmacy, Tartu University Hospital, Tartu, Estonia; ³Institute of Clinical Medicine, University of Tartu, Tartu, Estonia; ⁴Clinic of Oncology and Haematology, North Estonia Medical Center, Tallinn, Estonia; ⁵Center for Psychopharmacology, Diakonhjemmet Hospital, Oslo, Norway; ⁶Department of Physiology and Pharmacology, Section of Pharmacogenetics, Karolinska Institutet, Stockholm, Sweden; ⁷Heart Clinic, Tartu University Hospital, Tartu, Estonia; ⁸Genetics and Personalized Medicine Clinic, Tartu University Hospital, Tartu, Estonia

Background/Objectives: CYP2C19 and CYP2D6 metabolize over 30% of drugs. While the impact of common genetic variants on drug metabolism is well-established, understanding the effects of rare genetic variants remains challenging due to their low prevalence and absence of intermediate phenotypes for robust testing. In this context, we initiated a clinical study within the Estonian Biobank (EstBB) to investigate the effect of novel CYP2C19 and CYP2D6 alleles identified by genome and exome sequencing (GS and ES).

Methods: Novel rare genetic variants were identified in GS (n = 3,000) and ES (n = 2,500) data from EstBB participants. Putative loss-of-function and missense variants were annotated using LOFTEE and ANNOVAR. We used phased genotype data to call star alleles and invited individuals with different combinations of star alleles and novel variants to participate in the clinical study. The functional effect of the alleles was assessed by intake of omeprazole and metoprolol as probe drugs for CYP2C19 and CYP2D6 activity, respectively, among 114 eligible study participants.

Results: Probe drug and metabolite concentrations from ten time points per participant were quantified by mass-spectrometry. Based on differences in drug/metabolite area-under-the-curve ratios we confirmed the predicted effect of novel missense variants in CYP2C19 and CYP2D6, shifting normal metabolizers into intermediate metabolizers. All deletions in CYP2C19 had clear loss-of-function effects.

Conclusion: This study emphasizes the importance of rare genetic variants in explaining the ‘missing heritability’ in drug response, and how biobanks with the possibility to recall participants can contribute to discoveries in pharmacogenomics.

Grants: Estonian Research Council (PRG184), Swedish Research Council (2021-02732).

Conflict of Interest: None declared

C09.2 A method to estimate the effect of medications using genome-wide data

Ricky Lali ^1;2, Wei Deng^3;4, Shihong Mao², Matteo Di Scipio^2;5, Guillame Paré^1;2;6

¹McMaster University, Department of Health Research Methods, Evidence, and Impact, Hamilton, Canada; ²Population Health Research Institute, David Braley Cardiac, Vascular and Stroke Research Institute, Hamilton, Canada; ³McMaster University, Psychiatry and Behavioural Neurosciences, Hamilton, Canada; ⁴Peter Boris Centre for Addictions Research, Hamilton, Canada; ⁵McMaster University, Medicine, Hamilton, Canada; ⁶McMaster University, Pathology and Molecular Medicine, Hamilton, Canada

Background/Objectives: Understanding the effect of medications on clinically relevant biomarkers is crucial to assess their safety and efficacy. However, the only reliable way to estimate these effects is through comparison of pre and post dose levels, ideally in randomized populations. Though large-scale biobanks house rich data on baseline medication use, effect estimation is subject to indication bias.

Methods: We propose GENEPIE (GENotype-based Effect Prediction for Interventions and Exposures), an unbiased method to estimate medication effects using ultra-fast genome-wide outcome heritability optimization using a restricted maximum likelihood model. We tested the effect of 100 medications on lipids and waist-to-hip ratio in 408,069 British UK Biobank participants at baseline.

Results: Consistent with clinical trial data, GENEPIE determined that rosuvastatin conferred the strongest low-density lipoprotein cholesterol (LDL-C) reduction (-46.5%; 95% CI, -47.0 to -45.9%), atorvastatin the second strongest reduction (-45.4%; 95% CI, -45.6 to -45.1%) and that ezetimibe demonstrated the weakest (-25.1%; 95% CI, -25.9 to -24.3%). Standard epidemiological approaches resulted in indication bias, with marked attenuation of expected effects on LDL-C (e.g., -5.20%; 95% CI, -6.06 to -4.34% for ezetimibe). GENEPIE also revealed that metformin and orlistat decrease waist-to-hip ratio, as expected, and also showed a potential role of fibrates in weight loss.

Conclusion: GENEPIE leverages cross-sectional biobank-scale data to produce drug effect estimates robust to indication bias. Our results recapitulate known medication effects and show a potential effect of fibrates on weight, an observation that currently has basis only in model systems.

Grants: Not applicable

Conflict of Interest: Ricky Lali: None declared, Wei Deng: None declared, Shihong Mao: None declared, Matteo Di Scipio: None declared, Guillame Paré Received consulting fees from Bayer, Sanofi, Amgen, Novartis, and Illumina.

C09.3 Gene-environment interplay of socioeconomic indices and complex diseases

Fiona Hagenbeek ¹, Max Tamlander¹, Zhiyu Yang¹, Tuomo Hartonen¹, Kira Detrois¹, Pekka Martikainen^2;3, Nina Mars^1;4, Andrea Ganna^1;5, Samuli Ripatti^1;5;6

¹Institute for Molecular Medicine Finland, Helsinki, Finland; ²Helsinki Institute for Demography and Population Health, Faculty of Social Sciences, University of Helsinki, Helsinki, Finland; ³Max Planck – University of Helsinki Center for Social Inequalities in Population Health, Helsinki, Finland; ⁴Stanley Center for Psychiatric Research, Broad Institute of MIT and Harvard, Cambridge, United States; ⁵Massachusetts General Hospital and Broad Institute of MIT and Harvard, Cambridge, United States; ⁶University of Helsinki, Department of Public Health, Helsinki, Finland

Consortium: FinnGen

Background/Objectives: Low socioeconomic status (SES) is a risk factor for several diseases. Here, we aim to explore whether differences in SES (educational attainment (EA) and occupation) modify polygenic risk for 19 complex diseases.

Methods: We used genomics, SES, and health data from approximately 280,000 Finnish individuals in the FinnGen study aged 35-80 at cohort entry. We considered 19 common diseases for which we calculated polygenic scores (PGS). EA was classified according to the 1997 international standard classification of education (ISCED) and categorized into low (ISCED ≤ 4), or high (ISCED ≥ 5) EA. Occupational information was categorized as either manual labor, self-employment, lower-level or high-level administrative occupations.

Results: Using Cox proportional hazard models, we observed higher PGS hazard ratios (per SD change in PGS) for the low EA group in 5 diseases (type 2 diabetes (T2D), rheumatoid arthritis, lung cancer, depression, and alcohol use disorder) and for the high EA group in breast, prostate, and all cancers. For example, in T2D the PGS hazard ratios per SD were 1.71 [95% CI 1.70,1.73] in low EA and 1.65 [1.63,1.67] in high EA groups and in breast cancer 1.53 [1.50,1.56] and 1.71 [1.67,1.74], respectively. Occupation results mostly mirrored the EA results. We are currently performing replication across European biobanks from the INTERVENE consortium.

Conclusion: We observed heterogeneity in PGS effects across SES groups which may improve equitable use of genetic information in healthcare.

Grants: This project is funded by the European Union’s Horizon 2020 research and innovation programme under grant agreement No 101016775.

Conflict of Interest: None declared

C09.4 From ATP7B variant spectrum to underdiagnosed cases: Estonian biobank supports preventing the preventable

Miriam Nurm ¹, Anu Reigo¹, Tarmo Annilo¹, Toomas Toomsoo^2;3, Margit Nõukas¹, Tiit Nikopensius¹, Vasili Pankratov¹, Tuuli Reisberg⁴, Georgi Hudjašov¹, Toomas Haller¹, Neeme Tonisson^1;5

¹Institute of Genomics, University of Tartu, Tartu, Estonia; ²School of Natural Sciences and Health, Tallinn University, Tallinn, Estonia; ³Confido Medical Center, Tallinn, Estonia; ⁴Core Facility of Genomics, University of Tartu, Tartu, Estonia; ⁵Tartu University Hospital, Genetics and Personalized Medicine clinic, Tartu, Estonia

Background/Objectives: Wilson’s disease (WD) is a rare copper metabolism disorder with highly variable phenotypic presentation, caused by deleterious variants in the ATP7B gene. WD often remains underdiagnosed and undertreated, but could lead to debilitating health outcomes without treatment. The Estonian Biobank (EstBB) holds genotype and phenotype data of over 213,000 adult participants, thereby providing an excellent platform for studying the ATP7B variation spectrum in the Estonian population.

Methods: We identified and classified according to ACMG/AMP guidelines 49 pathogenic or likely pathogenic ATP7B variants in the cohort. We performed a comprehensive analysis of all available electronic health records (lab measurement results, ICD-10 codes, drug prescription data, epicrises) for EstBB participants diagnosed with copper metabolism disorders (ICD-10 code E83.0) (n = 59) or at high risk for developing WD (n = 17). Our study culminated in a recall project for biobank participants at high risk of WD for precise phenotyping (including family history, clinical biochemistry, neurological examination, etc.).

Results: The predominant pathogenic ATP7B variant in Europe, p.His1069Gln, has significantly higher prevalence in the Estonian population (AF 0.006) compared to other non-isolated populations (gnomAD global AF 0.001). Based on electronic health records, thirteen EstBB participants were ascertained as probably undiagnosed WD cases. Regardless of age, genotype, and prior WD diagnosis, all recall participants showed serum copper and ceruloplasmin values below reference, and 86.7% had signs of early to late neurodegeneration.

Conclusion: Population-based biobanks are excellent resources for combining continuous multilevel data and providing personalized feedback on disease risks, enabling better disease diagnostics in the future.

Grants: Estonian Research Council (#PRG555)

Conflict of Interest: None declared

C09.5 The multi-omics landscape of systemic inflammation in people living with HIV

Nienke van Unen ^1;2, Javier Botey Bataller^1;2;3, Jessica Dos Santos³, Maartje Cleophas³, Marc Blaauw³, Willem Vos³, Louise van Eekeren³, Albert Groenendijk^3;4, Xun Jiang^1;2, manoj Kumar Gupta^1;2, Nhan Nguyen^1;2, Cheng-Jian Xu^1;2, Leo A.B. Joosten³, Mihai Gheorghe Netea³, Andre van der Ven³, Yang Li^1;2;3

¹Centre for Individualised Infection Medicine (CiiM), Hannover, Germany; ²Twincore, Zentrum für Experimentelle und Klinische Infektionsforschung GmbH, Hannover, Germany; ³Radboud University Medical Center, Nijmegen, Netherlands; ⁴Erasmus MC, Rotterdam, Netherlands

Consortium: 2000HIV

Background: People living with HIV (PLHIV) with viral suppression because of HIV-treatment have a higher susceptibility to non-AIDS co-morbidities, such as cardiovascular diseases (CVD), due to systemic inflammation. Systemic inflammation is potentially caused by immune activation by viral glycoproteins, but its underlying mechanisms are largely unknown.

Methods: We analyzed 1342 virally suppressed PLHIV participating in the 2000HIV project. By integrating five different omics layers together with immune function profiling, we deciphered mechanisms underlying persistent inflammation.

Results: Integrating epigenomics, transcriptomics, proteomics, metabolomics, and cytokine responses resulted in a set of co-morbidity related factors capturing inter-individual variation across layers. Specifically, we identified latent factors associated with CVD and accelerated aging. Furthermore, we used genomics to understand the sources of inter-individual variation in the different omics layers and identified 4,766 expression-QTLs, 1,155 protein-QTLs, 22 metabolite-QTLs and 2 cytokine-QTLs at study-wide significance level. The identified QTLs were largely concordant with existing QTL studies performed in healthy cohorts. Mendelian randomization analysis revealed baseline determinants of immune responses, such as gene TNFRSF8 and protein IL17D as regulators of immune responses to HIV. Lastly, by linking genetic variation to the latent factors we identified, we discovered that the NLRP12 locus regulates the inflammasome complex at different layers. This suggests a causal factor in determining systemic inflammation in PLHIV.

Conclusion: All in all, we discovered intricate molecular mechanisms underlying inter-individual variation in systemic inflammation in 1342 virally suppressed PLHIV, identified molecular QTLs in PLHIV, and pinpointed baseline regulators orchestrating the immune response to HIV stimulation.

Conflict of Interest: None declared

C09.6 Discovery of core genes for type 1 diabetes via genome-wide aggregation of trans-effects using UK Biobank proteomics study.

Xuan Zhou¹, Andrii Iakovliev ², Stuart McGurnaghan², Athina Spiliopoulou^1;2, Helen M Colhoun², Paul McKeigue^1;2

¹Usher institute, Collage of Medicine and Veterinary Medicine, Edinburgh, United Kingdom; ²The Institute of Genetics and Cancer, College of Medicine and Veterinary Medicine, Edinburgh, United Kingdom

Background/Objectives: Type 1 diabetes (T1D) is a complex autoimmune disease involving destruction of insulin-producing beta cells. Our objective is to identify putative “core” effector genes for T1D using genome-wide aggregated trans-effects (GATE) analysis to inform drug discovery.

Methods: We used pQTL summary statistics from UKB-PPP and deCODE to compute GATE scores in a target dataset (4964/7497 cases/controls) from the SDRNT1BIO and the Generation Scotland respectively. We tested for association between T1D and GATE scores, one protein at a time, and identified core genes based on the strength of the association, the number of independent trans-loci contributing to each score and retained core genes with uncorrelated GATE scores.

Results: We identified 16 putative core genes at P values < 10^-6, 13 were novel compared with our previous analysis. The strongest finding is PDCD1 (LogOR = 0.19, P = 10^-24) followed by CXCL10, LGALS9, PRSS27, FASLG, IDO1, SDK2, AGR2, FGF19, SELL, TCN1, and KLRB1. Of these, 7 are in immune pathways, 6 are expressed on immune cells, 2 are expressed on NK-cells. Expression of PD-1, encoded by PDCD1, measured among 253 T1D cases in UKB-PPP is strongly associated with T1D (LogOR = 0.47, P = 2x10^-15).

Conclusion: We highlight the role of adaptive (PDCD1, LGALS9, IDO1) as well as innate (FASLG, KLRB1) immune pathways in T1D, and suggest new targets for drug development including PDCD1 -- immune checkpoint in a clinical trial for rheumatoid arthritis. Our investigation substantiates the “omnigenic” model and sheds light on the pathogenesis of T1D.

Grants: Diabetes UK Project Grant [15/0005301]; Cross Disciplinary Fellowship (XDF) Programme [MC_FE_00035]

Conflict of Interest: None declared

C10.1 The Australian national genomic autopsy study: a summary of results, outcomes and instructive families from 406 trios/quads.

Thuong Ha^1;2, Alicia Byrne^1;3;4, Peer Arts^1;5, Karin Kassahn^5;6, Lynn Pais^4;7, Anne O’Donnell-Luria^4;7, Milena Babic⁸, Mahalia Frank¹, Jinghua Feng^3;9, Paul Wang², David Lawrence^2;6, Leila Eshraghi¹, Luis Arriola^1;2, John Toubia², Hung Nguyen⁶, Alison Gardiner^1;5, Jarrad Dearman¹⁰, Hannah Kovilpillai¹⁰, Genomic Autopsy Study Research Network¹¹, George McGillivray¹², Jason Pinner¹³, Fiona McKenzie^14;15, Rebecca Morrow^1;16, Jill Lipsett¹⁶, Nick Manton¹⁶, Yee Khong^5;16, Theresa Power¹⁶, Lynette Moore^5;16, Jan Liebelt^5;10, Suzanne Sallevelt¹⁰, Lesley McGregor¹⁰, Shannon LeBlanc¹⁰, Arjan Bouman¹⁰, Andreas Schreiber^2;3;17, Sarah King-Smith^1;11, Tristan Hardy^5;6, Matilda Jackson^1;11, Hamish S. Scott^1;2;3;5;11, Christopher P. Barnett ^5;10;11

¹Centre for Cancer Biology, Genetics and Molecular Pathology, Adelaide, Australia; ²Centre for Cancer Biology, ACRF Genomics Facility, Adelaide, Australia; ³University of South Australia, UniSA Clinical and Health Sciences, Adelaide, Australia; ⁴Broad Institute of MIT and Harvard, Program in Medical and Population Genetics, Cambridge, United States; ⁵University of Adelaide, Adelaide Medical School, Adelaide, Australia; ⁶SA Pathology, Genetics and Molecular Pathology, Adelaide, Australia; ⁷Boston Children’s Hospital, Division of Genetics, Boston, United States; ⁸Centre for Cancer biology, Genetics and molecular pathology, Adelaide, Australia; ⁹Centre for Cancer biology, ACRF Genomics Facility, Adelaide, Australia; ¹⁰Women’s and Children’s Hospital, Paediatric and Reproductive Genetics Unit, North Adelaide, Australia; ¹¹Australian Genomics, Melbourne, Australia; ¹²Murdoch Children’s Research Institute, Victorian Clinical Genetics Services and Royal Women’s Hospital, Melbourne, Australia; ¹³Sydney Childre’s Hospital Network, Centre for Clinical Genetics, Randwick Sydney, Australia; ¹⁴Genetics Services of Western Australia, Perth, Australia; ¹⁵University of Western Australia, School of Paediatrics and Child Health, Perth, Australia; ¹⁶SA Pathology at Women’s and Children’s Hospital, Anatomical Pathology, North Adelaide, Australia; ¹⁷University of Adelaide, School of Biological Sciences, Adelaide, Australia

Consortium: Genomic Autopsy Study Research Network

Background/Objectives: The cause of pregnancy loss and perinatal death remains unexplained in ~ 25% of cases, despite a high perinatal autopsy rate in Australia. The genomic autopsy study is a national collaborative study using WES and WGS to identify genetic causes of fetal/newborn abnormalities that result in termination of pregnancy, death in utero or in the newborn period, in view to providing families with answers regarding cause and likelihood of recurrence.

Methods: WES and/or WGS was performed on prospective families referred to the Genetics unit (parent-fetus trios/quads). Priority cases were consanguineous families, fetuses with multiple malformations, and unexplained fetal/newborn death. Statistical, bioinformatic and experimental laboratory techniques were used to confirm variant causality.

Results: 406 prospective trios or quads have been recruited and sequenced from around Australia. 170/406 (42%) are either solved or have a strong candidate gene identified. Of the 170 solved families, 87 (51.2%) have been clearly solved by identification of a pathogenic mutation in a known gene. An additional 33/170 (19.4%) have a novel VUS in a known OMIM disease gene. Numerous instructive cases involving numerous genes (e.g., KIF14, SIK3, LAMC3, ARSL) will be presented.

Conclusion: Extensive genomic investigation of pregnancy loss and perinatal death should be offered as standard care, particularly when congenital abnormalities are present. Many families have benefitted from a clear diagnosis which could not be made on clinical grounds. Fetal genetic causes of late stillbirth are rare.

Grants: MRFF-GHFM Project Grant GHFM76777; NHMRC Project Grant ID APP1123341 CI Barnett and Scott are co-senior authors

Conflict of Interest: None declared

C10.2 Scalable automated reanalysis in rare disease: achieving high performance while limiting curation burden in diverse clinical and research cohorts

Matthew Welland^1;2, Kaileigh Ahlquist^3;4, Paul de Fazio⁵, Christina Austin-Tse^3;6, Lynn Pais⁴, Laura Wedd^1;2, Samantha Bryen^1;2, Rocio Rius^1;2, Michael Franklin^2;7, Giles Hall⁶, Laura Gauthier³, Andrew Mallett⁸, Amali Mallawaarachchi⁷, Paul Lockhart⁹, Richard Leventer^9;10, Ingrid Scheffer^9;10, Katherine Howell^9;10, Karin Kassahn¹¹, Hamish S. Scott¹¹, Christopher Semsarian¹², Julie McGaughran¹³, Bryony Thompson¹⁴, Chirag Patel¹³, Greg Smith¹⁵, Anne O’Donnell-Luria^3;4, Simon Sadedin⁵, Heidi Rehm^3;4, Sebastian Lunke⁵, Jeremiah Wander¹⁵, Kaitlin Samocha^3;4, Cas Simons^1;2, Daniel MacArthur^1;2, Zornitza Stark ⁵

¹Garvan Institute of Medical Research, Centre for Population Genomics, Sydney, Australia; ²Murdoch Children’s Research Institute, Center for Population Genomics, Melbourne, Australia; ³Massachusetts General Hospital, Center for Genomic Medicine, Boston, United States; ⁴Broad Institute of MIT and Harvard, Program in Medical and Population Genetics, Cambridge, United States; ⁵Murdoch Children’s Research Institute, Victorian Clinical Genetics Services, Melbourne, Australia; ⁶Broad Institute of MIT and Harvard, Data Sciences Platform, Cambridge, United States; ⁷Garvan Institute of Medical Research, Sydney, Australia; ⁸James Cook University, College of Medicine and Dentistry, Townsville, Australia; ⁹Murdoch Children’s Research Institute, Melbourne, Australia; ¹⁰Royal Children’s Hospital, Melbourne, Australia; ¹¹SA Pathology, Department of Genetics and Molecular Pathology, Adelaide, Australia; ¹²Royal Prince Alfred Hospital, Department of Cardiology, Sydney, Australia; ¹³Royal Brisbane and Women’s Hospital, Genetic Health Queensland, Brisbane, Australia; ¹⁴Royal Melbourne Hospital, Melbourne, Australia; ¹⁵Microsoft Research, Redmond, United States

Background: Reanalysis of existing genomic data in rare disease is highly effective in increasing diagnostic yield but remains limited by current manual approaches.

Methods: We developed an automated reanalysis framework, leveraging new, publicly available data on gene-disease relationships (PanelApp Australia) and variant pathogenicity (ClinVar). The framework was validated in a phenotypically diverse cohort of 400 rare disease patients from the USA and Australia. It was then applied prospectively in 2,002 unselected patients, who remained undiagnosed following clinical testing between 2018-2020 (1761 singletons, 241 trios; 573 genomes, 1429 exomes).

Results: In the validation cohort, our framework identified 159/196 (81%) of known diagnoses with data analysed as trios, and 145/196 (74%) if analysed as singletons. On average, it returned one candidate variant per trio, and two variants per singleton. Iterative reanalysis of unsolved patients (N = 204) identified five additional diagnoses in 12 months (2.5%). In the prospective cohort, the framework has identified 80 new diagnoses to date (4.0% yield): 42% due to new gene-disease relationships, 13% due to additional variant-level evidence, 40% due to process issues in the original analysis, and 5% due to bioinformatic improvements. All but one of the variants returned to diagnostic laboratories for evaluation have been classified as pathogenic or likely pathogenic.

Conclusion: We demonstrate the feasibility of an automated reanalysis model, optimised for high specificity and scalable to thousands of unsolved rare disease patients across multiple clinical and research environments. The model operates iteratively, raising the near-term prospect of delivering continuous and systematic automated reanalysis at scale.

Grants: GHFM APP2008820

Conflict of Interest: Matthew Welland: None declared, Kaileigh Ahlquist: None declared, Paul de Fazio: None declared, Christina Austin-Tse: None declared, Lynn Pais: None declared, Laura Wedd: None declared, Samantha Bryen: None declared, Rocio Rius: None declared, Michael Franklin: None declared, Giles Hall: None declared, Laura Gauthier: None declared, Andrew Mallett: None declared, Amali Mallawaarachchi: None declared, Paul Lockhart: None declared, Richard Leventer: None declared, Ingrid Scheffer: None declared, Katherine Howell: None declared, Karin Kassahn: None declared, Hamish S. Scott: None declared, Christopher Semsarian: None declared, Julie McGaughran: None declared, Bryony Thompson: None declared, Chirag Patel: None declared, Greg Smith Microsoft, Anne O’Donnell-Luria: None declared, Simon Sadedin: None declared, Heidi Rehm: None declared, Sebastian Lunke: None declared, Jeremiah Wander Microsoft, Kaitlin Samocha: None declared, Cas Simons: None declared, Daniel MacArthur GlaxoSmithKline, Insitro, Foresite Labs, Zornitza Stark: None declared

C10.3 Genome sequencing of a Pan-European cohort of 883 rare disease cases leads to new diagnoses in the Solve-RD project

Steven Laurie ^1;2, Leslie Matalonga^1;2, Ida Paramonov^1;2, german demidov³, Galuh Astuti⁴, Wouter Steyaert⁴, Kornelia Ellwanger³, Stephan Ossowski³, Christian Gilissen⁴, Katja Lohmann⁵, Holm Graessner³, Richarda de Voer⁴, Ana Töpf⁶, Lisenka Vissers⁴, Alexander Hoischen⁴, Sergi Beltran^1;7

¹Centro Nacional de Análisis Genómico - CNAG, Barcelona, Spain; ²Universitat de Barcelona, Barcelona, Spain; ³Universitätsklinikum Tübingen – Institut für Medizinische Genetik und angewandte Genomik, IMGAG, Tübingen, Germany; ⁴Radboud University Medical Center, Nijmegen, Netherlands; ⁵University of Lübeck, Institute of Neurogenetics, Lübeck, Germany; ⁶John Walton Muscular Dystrophy Research Centre (JWMDRC), Translational and Clinical Research Institute, Newcastle, United Kingdom; ⁷Universitat de Barcelona, Departament de Genètica, Microbiologia I Estadística, Facultat, de Biologia,, Barcelona, Spain

Consortium: Solve-RD consortium

Background/Objectives: A key objective of the H2020 Solve-RD project is to provide a diagnosis for thousands of families affected by rare diseases (RD) across Europe, the majority of which have previously undergone exhaustive analysis of exome sequencing (ES) data, without resolution. To this end, 883 families underwent short-read genome sequencing (GS), with the goal that the broader and more even coverage provided by GS will allow the identification of previously missed causative variants.

Methods: DNA from affected probands (n = 883), and relatives (970) was sequenced to generate 30X coverage on a BGI-DNBseq platform. In parallel, deep phenotypic descriptions of probands were collated using Human Phenotype Ontology terms. Alignment and short variant calling were performed using BWA-MEM and GATK, and a suite of tools was applied to identify rare de novo, mtDNA, copy-number, structural, and short tandem repeat variants. Variants were prioritised for clinical interpretation by Solve-RD partnering European Reference Network (ERN) experts if they impacted upon candidate genes in curated lists provided by the submitting ERNs

Results: While clinical interpretation of identified variants remains ongoing, to date >10% of families for which prioritised variants have been interpreted have had their diagnostic odysseys brought to a conclusion as a result of this pan-European effort. We report here on some interesting cases highlighting why they were not resolved through analyses of the original genetic data.

Conclusion: Systematic analysis of short-read GS data, accompanied by detailed patient descriptions, can elevate the diagnostic rate in RD cases markedly, even where prior comprehensive analysis of genetic data proved inconclusive.

Grants: EU H2020-779257

Conflict of Interest: None declared

C10.4 Rare disease gene association discovery from burden analysis of the 100,000 Genomes Project data

Valentina Cipriani ^1;2;3, Letizia Vestito¹, Damian Smedley¹

¹William Harvey Research Institute, Queen Mary University of London; ²UCL Institute of Ophthalmology, University College London; ³UCL Genetics Institute (UGI), University College London

Consortium: the Genomics England Clinical Interpretation Partnership (GeCIP) leads and collaborators

Background/Objectives: High-throughput sequencing has accelerated gene discovery and molecular diagnosis in rare diseases. However, the genetic cause for 50% of Mendelian diseases remains unknown. We set out to address the “missing heritability” in rare diseases.

Methods: We have developed an Exomiser-based gene burden analytical framework to identify enrichments of rare, protein-coding, predicted pathogenic and segregating variants in probands of families with a specific phenotype relative to control families and applied it to 35,008 rare disease families recruited to the 100,000 Genomes Project (100KGP).

Results: Following in silico and clinical expert triaging, we identified 88 novel associations (38 with existing experimental evidence; 5 interim published confirmations via collaboration/independent study). Here, we describe 7 compelling associations: hypertrophic cardiomyopathy with DYSF and SLC4A3 (high/specific heart expression; existing associations to skeletal dystrophies/short QT syndrome); monogenic diabetes with UNC13A (known role in regulation of β cells; mouse model with impaired glucose tolerance); epilepsy with KCNQ1 (mouse model shows seizures; existing long QT syndrome association may be linked); early onset Parkinson’s disease with RYR1 (existing links to tremor pathophysiology; mouse model with neurological phenotypes); anterior segment ocular abnormalities associated with POMK (expression in corneal cells; zebrafish model with developmental ocular abnormalities); and cystic kidney disease with COL4A3 (high renal expression; prior evidence for a digenic/modifying role in renal disease).

Conclusion: Confirmation of our findings would lead to potential diagnoses in 456 molecularly undiagnosed cases within the 100KGP and additional patients worldwide, highlighting the clinical impact of a large-scale statistical approach to rare disease gene discovery.

Grants: 1R01HD103805-01;GACR 24-10324S;ASAP-000478;ASAP-000509.

Conflict of Interest: Valentina Cipriani: None declared, Letizia Vestito: None declared, Damian Smedley 1R01HD103805-01, D.S. was seconded to and received salary from Genomics England, a wholly owned Department of Health and Social Care company, from 2016 to 2018.

C10.5 Structural and non-coding variants increase the diagnostic yield of clinical whole genome sequencing for rare diseases

Jenny Taylor ¹, Alistair Pagnamenta¹, Carme Camps¹, Edoardo Giacopuzzi², John Taylor³, Eduardo Calpena⁴, Usha Kini³, Andrew Wilkie⁴, Niko Popitsch⁵

¹University of Oxford, Centre for Human Genetics, Oxford, United Kingdom; ²Human Technopole, Milan, Italy; ³Oxford Universities Hospital Trust, Oxford, United Kingdom; ⁴University of Oxford, Weatherall Institute of Molecular Medicine, Oxford, United Kingdom; ⁵University of Vienna, Department of Biochemistry and Cell Biology, Vienna, Austria

Consortium: OxClinWGS Consortium

Background/Objectives: Whole genome sequencing (WGS) has revolutionised the diagnosis of patients with rare diseases. However, the diagnostic yields of large-scale clinical programmes are still modest at ~25-30%, in part because the full genome is not interrogated.

Methods: We undertook WGS on a cohort of 122 unrelated rare disease patients and their relatives (300 genomes) who were referred with a broad spectrum of rare diseases and had been pre-screened by arrays, panel and/or exome sequencing. We incorporated algorithms SVRare, ALTSPLICE and GREEN-DB in the bioinformatics pipeline to enhance the annotation of structural, splicing and regulatory variants, respectively.

Results: We identified pathogenic or likely pathogenic variants in 43/122 cases (35%). Seven structural variants were identified, three of which were in genes novel for disease or phenotype, whilst the remaining four had not been detected by prior clinical testing. Two were complex rearrangements.

Of sixteen splice site variants identified, seven were in non-canonical sites of known genes and two, were in genes, MCM10 and DOCK7, novel at the time of discovery. Our WGS results expanded the clinical spectrum for RMND1, changed the clinical diagnoses for six patients and informed treatment in a further eight patients, which for five patients was considered life-saving. Incidental findings were discovered in FBN1 and KCNQ1 genes and led to diagnoses of Marfan and long QT syndromes, respectively.

Conclusion: Our results emphasize the importance of analyzing the full genome sequence, including introns, to maximize clinical diagnoses for rare disease patients.

Grants: Health Innovation Challenge Fund [R6-388 / WT 100127]. Oxford NIHR Biomedical Research Centre

Conflict of Interest: None declared

C10.6 Genome sequencing: efficiently feeding the health care system.

Pierre Blanc ¹, Boris Keren^1;2, Benjamin Cogne^1;3, Thomas Smol^1;4, Jean-Madeleine de Sainte Agathe^1;2, Jonathan Levy^1;5, Audrey Briand-Suleau¹, Lamisse Mansour-Hendili¹, Paul Gueguen^1;6, Tania Attie-Bitach^1;7, Sophie Rondeau^1;7, Giulia Barcia^1;7, marion lesieur-sebellin^1;7, Sophie Valleix^1;7, François Lecoquierre^1;8, Christele Dubourg^1;9, Cyril Burin Des Roziers^1;7, VALÉRIE MALAN^1;7, Corinne Collet^1;7, Olivier Grunewald^1;4, Jérémy Bertrand¹, Jérôme Champ¹, Sylvain Dugat¹⁰, Amélie Souchet¹, Jérôme Chansard¹, Floriane Carallis¹, Steve Ho Tam Chay¹, Béréthé Diarra¹, Ines Detrait¹, Audrey Adechokan¹, Marielle Letchimy¹, Leslie Delachaux¹, Camille Barette¹, Sami Mahamoudi¹, Nicolas Derive¹, Thomas Rambaud¹, Zinara Lidamahasolo¹, Pierre Marijon¹, Virginie Saillour¹, Anaïs Lharidon¹, Aurélien De Reynies¹, Antonio RAUSELL¹, Sacha Schutz^1;11, Laurent Frobert¹, Mouhamadou Niang¹, Kevin Jousselin¹, Marion Ducrotverdun¹, Lara Vinauger¹, Alban Lermine¹, Michel Vidaud¹

¹Laboratoire SeqOIA; ²APHP.SU; ³CHU de Nantes; ⁴CHRU de Lille; ⁵APHP.Nord; ⁶CHU de Tours; ⁷APHP.Centre; ⁸CHU de Rouen; ⁹CHU de Rennes; ¹⁰Integragen; ¹¹CHU de Brest

Background/Objectives: Genomic medicine is a cutting edge and rapidly expanding domain in the health care landscape. However, if genome sequencing data production is no longer a challenge, the clinical pathway back to the patient still represents a burden. In 2018, the French government launched a genomic medicine plan (PFMG 2025) and set up two national laboratories, accepting the idea that they could adopt distinct tools and strategies. We present the core output and choices of SeqOIA, the lab that operates in the northern half of France.

Methods: Between 2020 and 2023, short-read genome sequencing was performed on ~31,000 rare disease participants including ~12,000 affected individuals in 60 medical conditions.

Results: The overall diagnostic rate is 31%. Intellectual disability and congenital abnormalities are the most represented conditions, respectively with 3,142 and 2,674 prescriptions and a raw diagnostic yield of 35% and 31%. Over the first 3 years, the total pending results fraction is 10% and by the end of 2023, 82% of the prescriptions of the first semester of the year led to a report with a median turnaround time of 3.4 months. Interestingly, among our strategies some may be considered as orthogonal to worldwide current trends. Using these, the mean time from genome variant analysis to report delivery felt down to 50 minutes.

Conclusion: We propose an alternative operational lab framework that enables genome sequencing results to efficiently return to the clinicians.

Grants: The SeqOIA Laboratory is funded by the French Ministry for Solidarity and Health.

Conflict of Interest: None declared

C11.1 KICS2 dysfunction impairs KICSTOR complex-mediated mTORC1 activation causing intellectual disability and epilepsy

Rebecca Buchert ¹, Martin Burkhalter², Linda Sofan¹, Timo Roser³, Kirsten Cremer⁴, Javeria Raza Alvi⁵, Stephanie Efthymiou⁶, Tawfiq Froukh⁷, Sughra Guliyeva⁸, Ulviyya Guliyeva⁸, Moath Hamdallah⁹, Muriel Holder¹⁰, Rauan Kaiyrzhanov¹¹, Doreen Klingler¹, Mahmoud Koko¹², Reza Maroofian¹³, Lars Matthies⁴, Joohyun Park¹, Ana Velic¹⁴, Stephanie Spranger¹⁵, Tipu Sultan¹⁶, Hartmut Engels⁴, Holger Lerche¹², Henry Houlden¹³, Alistair Pagnamenta¹⁷, Ingo Borggräfe¹⁸, Yvonne Weber¹², Penelope Bonnen¹⁹, Olaf Riess¹, Melanie Philipp², Tobias Haack¹

¹University of Tübingen, Institute of Medical Genetics and Applied Genomics, Tübingen, Germany; ²University of Tübingen, Department of Experimental and Clinical Pharmacology and Pharmacogenomics, Division of Pharmacogenomics, Tübingen, Germany; ³Ludwig-Maximilians-University, Division of Pediatric Neurology, Developmental Medicine and Social Pediatrics, Department of Pediatrics, Munich, Germany; ⁴University of Bonn, Institute of Human Genetics, School of Medicine & University Hospital Bonn, Bonn, Germany; ⁵Children’s Hospital Lahore, Department of Pediatric Neurology, Institute of Child Health, Lahore, Pakistan; ⁶UCL Queen Square Institute of Neurology, Department of Neuromuscular Disorders, London, United Kingdom; ⁷Philadelphia University, Department of Biotechnology and Genetic Engineering, Amman, Jordan; ⁸MediClub Hospital, Baku, Azerbaijan; ⁹Farah Hospital, Amman, Jordan; ¹⁰Guy’s & St Thomas’ NHS Foundation Trust, Clinical Genetics Department, Guy’s Hospital, London, United Kingdom; ¹¹UCL Institute of Neurology, Department of Neuromuscular diseases, London, United Kingdom; ¹²University of Tübingen, Department of Neurology and Epileptology, Hertie-Institute for Clinical Brain Research, Tübingen, Germany; ¹³University College London, Department of Neuromuscular Diseases, London, United Kingdom; ¹⁴University of Tübingen, Proteome Center Tübingen, Tübingen, Germany; ¹⁵Practice of Human Genetics, Bremen, Germany; ¹⁶Children’s Hospital & the Institute of Child Health, Lahore, Pakistan; ¹⁷University of Oxford, NIHR Oxford Biomedical Research Centre, Centre for Human Genetics, Oxford, United Kingdom; ¹⁸Ludwig-Maximilians-University, Dr. von Hauner Children̛s Hospital, Department of Pediatric Neurology and Developmental Medicine, Munich, Germany; ¹⁹Baylor College of Medicine, Molecular and Human Genetics, Houston, United States

Background/Objectives: Nutrient-dependent mTORC1 regulation upon amino acid deprivation is mediated by the KICSTOR complex, comprising SZT2, KPTN, ITFG2, and KICS2, which recruits GATOR1 to lysosomes. Previously, pathogenic SZT2 and KPTN variants have been associated with autosomal recessive intellectual disability (ID) and epileptic encephalopathy. Here we report on eight families with ID and biallelic loss of function variants in KICS2 (previously C12orf66).

Methods: Exome or genome sequencing identified five different variants as well as a 1.1 Mb deletion involving KICS2. We generated HEK293T cells carrying the respective variants using CRISPR/Cas9 gene-editing and performed mTORC1 activity assays and dimethyl-labeled phosphoproteomics. Furthermore, we depleted Kics2 in zebrafish using morpholinos and attempted rescue using either wildtype or mutant RNA.

Results: We identified biallelic KICS2 variants in eight index cases presenting with ID and epilepsy. In HEK293T cells, the identified variants demonstrated a deleterious impact on nutrient-dependent mTORC1 regulation, as confirmed by phosphoproteome analysis. Depletion of Kics2 in zebrafish resulted in ciliary dysfunction consistent with a role of mTORC1 in cilia biology which could not be rescued with mutant Kics2. Cilia analysis in zebrafish and human fibroblasts confirmed the pathogenicity of the identified variants.

Conclusion: Our genetic and experimental data provide evidence that KICS2 is a novel factor involved in ID presumably due to its dysfunction impacting mTORC1 regulation and cilia biology.

Grants: German Academic Exchange Service (DAAD, Project-ID 57166498), Fundació La Marató de TV3 (grant number 202019-32) and the Deutsche Forschungsgemeinschaft (German Research Foundation, grant numbers PH144/4-3, 6-1, 418081722, 433158657 and INST 37/1049-1).

Conflict of Interest: None declared

C11.2 De novo variants in the SF3B1 gene are responsible of a new spliceosomopathy with neurodevelopmental disorders.

Kevin Uguen^1;2, Tiffany Bergot¹, Marie Pier Scott-Boyer³, Solene Chapalain¹, Severine Commet¹, Claudia Gonzaga-Jauregui⁴, Susan Hiatt⁵, Devon Haynes⁶, Michael Kruer^7;8, Patrick Sulem⁹, Cynthia Curry¹⁰, Trine Prescott¹¹, Kohji Kato¹², Merry Ferre¹³, Changlian Zhu¹⁴, Theresa Brunet¹⁵, Martine Doco-Fenzy¹⁶, Thomas Courtin¹⁷, Poirsier Céline¹⁸, Trine Bjørg Hammer¹⁹, Tony Roscioli²⁰, Melissa MacPherson²¹, Jacqueline Leonard²², Dong Li²², Ian Glass²³, Scott Ward²⁴, Philippe M. Campeau²⁵, Eric Lippert¹, Laurent Corcos¹, Claude Férec¹, Arnaud Droit³, Sebastien Küry¹⁶, Delphine Bernard ¹

¹Université de Bretagne Occidentale, Inserm U1078, Brest, France; ²CHRU de Brest, Service de Génétique Médicale, Brest, France; ³Université de Laval, Research Center of Quebec CHU, Quebec, Canada; ⁴Universidad Nacional Autónoma de México, Laboratorio Internacional de Investigación sobre el Genoma Humano, Juriquilla, Mexico; ⁵HudsonAlpha Institute for Biotechnology, Huntsville, United States; ⁶Arnold Palmer Hospital for Children-Orlando Health, Division of Genetics, Orlando, United States; ⁷Phoenix Children’s Hospital, Phoenix, Pediatric Movement Disorders Program, Division of Pediatric Neurology, Barrow Neurological Institute, Phoenix, United States; ⁸University of Arizona College of Medicine, Departments of Child Health, Neurology, Cellular & Molecular Medicine and Program in Genetics, Phoenix, United States; ⁹deCODE Genetics/Amgen, Inc., Reykjavik, Iceland; ¹⁰University of California, San Francisco, Genetic Medicine, Fresno, United States; ¹¹Telemark Hospital Trust, Department of Medical Genetics, Skien, Norway; ¹²Nagoya University, Department of Genetics, Research Institute of Environmental Medicine (RIeM), Nagoya; ¹³Laboratory Corporation of America Holdings (LabCorp), Durham, United States; ¹⁴University of Gothenburg, enter for Brain Repair and Rehabilitation, Institute of Neuroscience and Physiology, Gothenburg, Sweden; ¹⁵Technical University of Munich, Institute of Human Genetics, Munich, Germany; ¹⁶Nantes Université, CHU de Nantes, Service de Génétique médicale, Nantes, France; ¹⁷Institut du Cerveau - Paris Brain Institute, Paris, France; ¹⁸CHU Reims, Département de génétique médicale, Reims, France; ¹⁹Filadelfia Danish Epilepsy Centre, Department of Epilepsy Genetics, Dianalund, Denmark; ²⁰Prince of Wales Hospital, New South Wales Health Pathology Randwick Genomics, Syndey, Australia; ²¹Faculty of Medicine and Dentistry, University of Alberta, Department of Medical Genetics, Edmonton, Canada; ²²Children’s Hospital of Philadelphia, Divisions of Human Genetics, Philadelphia, United States; ²³University of Washington, Department of Pediatrics, Seattle, United States; ²⁴Vanderbilt University Medical Center, Department of Pediatrics, Division of Medical Genetics and Genomic Medicine, Nashville, United States; ²⁵CHU Sainte-Justine, Centre Hospitalier Universitaire Saint-Justine Research Center, Montreal, Canada

Background/Objectives: SF3B1 (splicing factor 3b subunit 1) is an essential splicing factor that plays a role in recognizing the branch point within the snRNP U2 complex. Recurrent missense mutations in the SF3B1 gene are frequently found in cancers. To date, no constitutional mutations have been described. However, mutations in other splicing machinery genes, such as SF3B2, SF3B4, or PUF60, have been associated with various developmental syndromes.

Methods: We describe an international cohort of 23 patients with a syndromic neurodevelopmental disorder and in whom a de novo missense (n = 15) or loss-of-function variation (n = 8) in the SF3B1 gene was identified. We assessed the effect of missense variants on SF3B1 function by targeted splicing analyses using K562 and HEK293 cells. We also performed genome-wide analyses (RNA-Seq) on K562 cells with inducible expression of three different variants.

Results: Targeted analyses show that each missense variant rescues the splicing defect induced by SF3B1 silencing, suggesting that none can inactivate SF3B1 function. Only one variant has an effect on splicing similar to cancer-associated missense variations. Genome-wide analyses suggest that all three variants studied lead to splicing remodeling, primarily involving exon skipping. These data, combined with available information on the effects of SF3B1 loss of function, support the involvement of these variants in a developmental phenotype. We highlight a genotype-phenotype correlation, with a more severe phenotype in patients with missense variants.

Conclusion: In conclusion, this work describes a new syndrome within the spectrum of spliceosomopathies. Further transcriptomic analyses are needed to elucidate the mechanisms associated with these SF3B1 variants.

Conflict of Interest: None declared

C11.3 ANO4 is a novel gene causing developmental and epileptic encephalopathy or GEFS+ and temporal lobe epilepsy

Fang Yang¹, Anais Begemann ², Nadine Reichhart¹, Akvile Haeckel³, Katharina Steindl², Eyk Schellenberger³, Ronja Fini Sturm¹, Magalie Barth⁴, Sissy Bassani², Paranchai Boonsawat², Thomas Courtin^5;6, Bruno Delobel⁷, Boudewijn Gunning⁸, Katia Hardies⁹, Jennesson Mélanie¹⁰, Tarja Linnankivi¹¹, Clément Prouteau⁴, Marta Spodenkiewicz¹², Sandra Toelle¹³, Koen L. I. van Gassen¹⁴, Wim Van Paesschen¹⁵, Nienke Verbeek¹⁴, Alban Ziegler⁴, Markus Zweier², Anselm Horn^2;16, Heinrich Sticht¹⁶, Holger Lerche¹⁷, Sarah Weckhuysen^9;18, Olaf Strauß¹, Anita Rauch²

¹Charité – Universitätsmedizin Berlin, Department of Ophthalmology, Berlin, Germany; ²University of Zurich, Institute of Medical Genetics, Schlieren-Zurich, Switzerland; ³Charité –Universitätsmedizin Berlin, Institute for Radiology and Children’s Radiology, Berlin, Germany; ⁴University Hospital of Angers, Department of Genetics, Angers, France; ⁵Sorbonne Université, INSERM, CNRS, Institut du Cerveau, Paris, France; ⁶Hôpital Pitié-Salpêtrière, DMU BioGe’M, AP-HP, Paris, France; ⁷Hopital Saint Vincent de Paul, GH de l’Institut Catholique de Lille, Lille, France; ⁸Stichting Epilepsie Instellingen Nederland, Zwolle, Netherlands; ⁹University of Antwerp, Applied & Translational Neurogenomics Group, Antwerp, Belgium; ¹⁰CHU Reims, Department of Pediatrics, Reims, France; ¹¹University of Helsinki and Helsinki University Hospital, Epilepsia Helsinki, Helsinki, Finland; ¹²La Réunion University Hospital, Department of Genetics, Saint-Pierre, France; ¹³Children’s University Hospital Zurich, Department of Pediatric Neurology, Zurich, Switzerland; ¹⁴University Medical Center Utrecht, Department of Genetics, Utrecht, Netherlands; ¹⁵KU Leuven, Laboratory for Epilepsy Research, Leuven, Belgium; ¹⁶Friedrich-Alexander-Universität Erlangen-Nürnberg, Division of Bioinformatics, Institute of Biochemistry, Erlangen, Germany; ¹⁷University of Tübingen, Department of Neurology and Epileptology, Tübingen, Germany; ¹⁸Antwerp University Hospital, Department of Neurology, Antwerp, Belgium

Background/Objectives: Anoctamins are a family of Ca²⁺-activated proteins that assemble as homodimers and may act as ion channels and/or phospholipid scramblases with limited understanding of function and disease association. To date, no human disease has been associated with variants in ANO4 (alias TMEM16D).

Methods: We assembled a cohort of 7 patients with ANO4 variants and assessed them by variant mapping and structural modeling. We evaluated functional effects of all variants in a heterologous expression system by patch-clamp recordings, immunocytochemistry, and surface expression of annexin A5 as a measure of phosphatidylserine scramblase activity.

Results: ANO4 missense variants occurred de novo in 5 patients with fever sensitive developmental and epileptic encephalopathy and were inherited with reduced penetrance in 2 families with generalized epilepsy with febrile seizures plus or temporal lobe epilepsy. Variant mapping and structural analysis suggested a preliminary genotype-phenotype correlation. Experimentally, all ANO4 variants showed severe loss of ion channel function and mild loss of surface expression due to impaired plasma membrane trafficking. Increased levels of Ca²⁺-independent annexin A5 at the cell surface suggested an increased apoptosis rate in most mutant ANO4 expressing cells, but no changes in Ca²⁺-dependent scramblase activity were observed. Co-transfection with ANO4 wildtype suggested a dominant negative effect.

Conclusion: In summary, we present a novel genetic base for both, encephalopathic sporadic and inherited fever sensitive epilepsies, and a first hereditary disease caused by variants in ANO4.

Grants: SNSF 320020_179547 and University of Zurich ITINERARE to A.R., DFG STR480/14-1 to O.S., and FWO 1861424N and G056122N to SW.

Conflict of Interest: None declared

C11.4 Heterozygous variants in BICRA and BICRAL, sub-units of ncBAF complex, cause neurodevelopmental disorders in human

Yuwei Shi ¹, Ruizhi Deng¹, Sourav Ghosh², Anita Nikoncuk¹, Scott Barish³, Daryl Scott³, Hugo J Bellen³, Axel Schmidt⁴, Sarah Dyack^5;6, Andrea Rideout^5;6, Isabelle Maystadt⁷, Colombine Meunier⁷, YLINE CAPRI⁸, Viviana Lupo⁸, Jonathan Levy⁹, Alain Verloes⁸, Nathalie Couque⁹, Andrea Holubová¹⁰, Miroslava Hančárová¹⁰, Didier LACOMBE^11;12, Vincent MIchaud^11;12, Miquel Raspall¹³, Clément Prouteau¹⁴, Vincent Milon¹⁴, Katharina Steindl¹⁵, Francesco Brancati¹⁶, Antonio Novelli¹⁷, Dario Cocciadiferro¹⁷, Maria Cristina Digilio¹⁸, Kai Muru^19;20, Hilde Van Esch²¹, Anita Gorrie²², Jessica Planner²², Hülya Kayserili²³, Gael Nicolas²⁴, Anne-Marie Guerrot²⁴, Mona Grimmel²⁵, Tobias Haack²⁵, Haley McConkey², Jessica Rzasa², Jennifer Kerkhof², Bekim Sadikovic^2;26, Kristina Lanko¹, Stefan Barakat¹

¹Erasmus University Medical Center, Department of Clinical Genetics, Rotterdam, Netherlands; ²London Health Sciences Centre, Verspeeten Clinical Genome Centre, London, Canada; ³Baylor College of Medicine, Department of Molecular and Human Genetics, Houston, United States; ⁴University Hospital Bonn, Institute of Human Genetics, Bonn, Germany; ⁵Dalhousie University, Department of Pediatrics, Halifax, Canada; ⁶IWK Health Centre, Maritime Medical Genetics Service, Halifax, Canada; ⁷Institute of Pathology and Genetics, Gosselies, Center for Human Genetics, Gosselies, Belgium; ⁸Hospital Robert Debré Ap-Hp, Clinical genetics unit, Genetic Department, Paris, France; ⁹Hospital Robert Debré Ap-Hp, Cytogenetics unit, Genetic Department, Paris, France; ¹⁰Charles University Second Faculty of Medicine and University Hospital Motol, Prague, Czech Republic; ¹¹Hospital Center University De Bordeaux, Service de Génétique Médicale, Centre de Référence Anomalies du Développement et Syndrome Malformatifs, Bordeaux, France; ¹²Université de Bordeaux, Laboratoire Maladies Rares: Génétique et Métabolisme, INSERM U1211,, Bordeaux, France; ¹³Vall d’Hebron University Hospital, Barcelona, Spain; ¹⁴Angers University Hospital Center, Department of Medical Genetics and Mitovasc INSERM 1083, CNRS 6015, Angers, France; ¹⁵University of Zurich, Instiute of Medical Genetics, Zürich, Switzerland; ¹⁶University of L’Aquila, Department of Life, Health and Environmental Sciences, L’Aquila, Italy; ¹⁷Bambino Gesù Children’s Hospital, Translational Cytogenomics Research Unit, Roma, Italy; ¹⁸Bambino Gesù Children’s Hospital, Medical Genetics Unit, Medical Genetics Laboratory, Pediatric Cardiology Department, Scientific Rectorate, Roma, Italy; ¹⁹Tartu University Hospital, Genetics and Personalized Medicine Clinic, Tartu, Estonia; ²⁰Tartu University, Institute of Clinical Medicine, Tartu, Estonia; ²¹UZ Leuven, Center for Human Genetics, Leuven, Belgium; ²²Monash Medical Centre, Melbourne, Australia; ²³Koç University Hospital, Istanbul, Türkyie; ²⁴Normandie Université, UNIROUEN, Inserm U1245, CHU Rouen, Department of Genetics and reference center for developmental disorders, FHU G4 Génomique, Rouen, France; ²⁵University of Tuebingen, Institute of Medical Genetics and Applied Genomics, Tübingen, Germany; ²⁶Western University, Department of Pathology and Laboratory Medicine, London, Canada

Background/Objectives: SWI/SNF-related intellectual disability disorders (SSRIDDs) are rare neurodevelopmental disorders caused by variants in several members of SWI/SNF chromatin-remodeling complexes. A new ncBAF complex with unique core subunits BICRA (a.k.a. GLTSCR1) and its homolog BICRAL (a.k.a. GLTSCR1L) were recently discovered. Interestingly we previously showed that variants in BICRA cause a unique SSRIDD in human.

Methods: Through international collaborations, we collected clinical data of 33 additional individuals carrying variants in BICRA and 5 individuals with variants in BICRAL. Blood-derived DNA was used to generate an episignature. Human embryonic stem cells based disease models were generated and differentiated towards neural stem cells and excitatory neurons, followed by RNA-sequencing and morphology analysis.

Results: Individuals carrying rare variants in BICRA present with neurodevelopmental phenotypes, including intellectual disability, developmental delay, autism spectrum disorders and behavioral abnormalities as well as dysmorphic features.

Similarly, individuals with variants BICRAL, showed overlapping clinical phenotypes.

A predominantly hypomethylated DNA methylation profile was identified in patients with pathogenic BICRA variants. The RNA-sequencing analysis revealed that either reduced BICRAL or complete absence of BICRA leads to transcriptome change in neural stem cells, altering pathways related to axon guidance. Both BICRA and BICRAL haploinsufficiency result in morphologically increased branching and decreased neurites’ length in excitatory neurons.

Conclusion: Our findings showed that haploinsufficiency of both BICRA and BICRAL are causes of NDDs, that are related to a specific episignature. Furthermore, our functional studies provide insights in the pathomechanisms of these novel NDDs.

Grants: China Scholarship Council (CSC), ZonMw Vidi 09150172110002

Conflict of Interest: None declared

C11.5 A recurrent homozygous single amino acid deletion (p.Glu133del) in MED22 leads to a progressive neurodevelopmental disorder with microcephaly, dystonia, seizures and neurodegeneration

Elisa Cali ¹, Lea Lescouzeres², Mikhail Savitskiy³, MariaSavina Severino⁴, Sara Fortuna⁵, Miguel Soler⁶, Vladimir Katanaev³, Dylan Taatjes⁷, Henry Houlden¹, Shunmoogum A. Patten², Reza Maroofian¹

¹University College London; ²Université de Montréal; ³University of Geneva, Geneva, Switzerland; ⁴Giannina Gaslini Institute, Genova, Italy; ⁵IIT - Center for Robotics and Intelligent Systems, Genova, Italy; ⁶Università degli Studi di Udine, Udine, Italy; ⁷University of Colorado Boulder, Boulder, United States

Background/Objectives: Proper neurodevelopment relies on gene expression programs influenced by signalling molecules, transcription factors (TFs), and epigenetic processes. In this study, we investigate an ultra-rare homozygous single amino acid deletion (p.Glu133del) in MED22, a component of Mediator, a key regulator of transcription, gene expression and posttranslational epigenetic modifications. Disruption of Mediator has been implicated in various neurodevelopmental and neurodegenerative disorders.

Methods: We analysed the variant’s impact into the Mediator structure and function through cellular studies (Western blotting and IP-MS) and computational simulations. We investigated MED22 protein function generating knock-out zebrafish model and knock-in Drosophila model. We performed scRNA-seq on zebrafish mutants.

Results: We identified eight affected individuals from six unrelated families with a severe progressive neurodevelopmental disorder, characterized by global developmental delay, microcephaly, dystonia, seizures, and microcephaly. In vivo models of Drosophila melanogaster and zebrafish demonstrate significant developmental disruption, mirroring major aspects of the human disorder and displaying Parkinson-like features. scRNA-sequencing data obtained from zebrafish mutants implicate MED22 in dopaminergic network dysfunction.

Conclusion: We establish the first association of MED22 with a human disorder, through integration of clinical, genetic and neuroradiological data. We provide insights in the molecular dynamics of Mediator dysfunction and in the neuronal pathways that might be affected by MED22 dysfunction.

Grants: Wellcome Trust (WT093205MA and WT104033AIA).

Conflict of Interest: None declared

C11.6 Elucidating the clinical and genetic spectrum of inositol polyphosphate phosphatase INPP4A-related neurodevelopmental disorder

Lettie Rawlins ^1;2, Reza Maroofian³, Stuart Cannon¹, Muhannad Daana⁴, Mina Zamani^5;6, Shamsul Ghani^1;7, Joseph Leslie¹, Nishanka Ubeyratna¹, Nasar Khan⁸, Hamid Khan⁷, Annarita Scardamaglia³, Robin Cloarec⁹, Shujaat Ali Khan⁷, Muhammad Umair^10;11, Saeid Sadeghian¹², Hamid Galehdari⁵, Almundher Al-Maawali^13;14, Adila Al-Kindi^13;14, Reza Azizimalamiri¹², Gholamreza Shariati^6;15, Faraz Ahmed¹⁶, Amna Mohammed Al-Futaisi¹⁶, Alireza Sedaghat^6;17, Mohammad Hamid¹⁸, Maha Zaki¹⁹, Abdullah Alhashem²⁰, Makaram Obeid²¹, Amjad Khan^22;23, Ahmad Beydoun²¹, Marwan Najjar²¹, Homa Tajsharghi²⁴, Giovanni Zifarelli²⁵, Peter Bauer²⁵, Rose-Mary Boustany²¹, lydie BURGLEN^26;27, Adam Gunning¹, Ehsan Ghayoor Karimiani²⁸, Joseph Gleeson^29;30, Mathieu MILH^9;31, Somaya Salah³², Jahangir Khan⁷, Volker Haucke^33;34, Caroline Wright¹, Lucy McGavin^35;36, Orly Elpeleg³², Muhammad Shabbir⁷, Henry Houlden³, Michael Ebner³³, Emma Baple^1;2, Andrew Crosby¹

¹University of Exeter, Clinical and Biomedical Sciences, Exeter, United Kingdom; ²Royal Devon and Exeter Hospital, Peninsula Clinical Genetics Service, Exeter, United Kingdom; ³University College London, Department of Neuromuscular Diseases, London, United Kingdom; ⁴Clalit Health Care Services, Child Development Center, Jerusalem, Israel; ⁵Shahid Chamran University of Ahvaz, Department of Biology, Ahvaz, Iran; ⁶Narges Medical Genetic & PND Laboratory, Ahvaz, Iran; ⁷International Islamic University Islamabad (IIUI), Department of Biological Sciences, Islamabad, Pakistan; ⁸Aarhus University, Interdisciplinary Nanoscience Center, Aarhus, Denmark; ⁹Timone Infant Hospital, APHM, Marseille, France; ¹⁰University of Management and Technology, Department of Life Sciences, Lahore, Pakistan; ¹¹King Abdullah International Medical Research Center (KAIMRC), Riyadh, Saudi Arabia; ¹²Ahvaz Jondishapur University of Medical Sciences, Department of Pediatric Neurology, Ahvaz, Iran; ¹³Sultan Qaboos University, Department of Genetics, Muscat, Oman; ¹⁴Sultan Qaboos University Hospital, Genetic and Developmental Medicine Clinic, Muscat, Oman; ¹⁵Ahvaz Jondishapur University of Medical Sciences, Department of Medical Genetics, Ahvaz, Iran; ¹⁶Sultan Qaboos University, Deaprtment of Child Health, Muscat, Oman; ¹⁷Ahvaz Jondishapur University of Medical Sciences, Diabetes Research Center, Ahvaz, Iran; ¹⁸Pasteur Institute of Iran, Department of Molecular Medicine, Tehran, Iran; ¹⁹National Research Centre (NRC), Department of Clinical Genetics, Cairo, Egypt; ²⁰Prince Sultan Military Medical City, Department of Radiology, Riyadh, Saudi Arabia; ²¹American University of Beirut, Beirut, Lebanon; ²²University of Lakki Marwat, Department of Zoology, Lakki Marwat, Pakistan; ²³University of Tübingen, Institute for Medical Genetics and Applied Genomics, Tübingen, Germany; ²⁴University of Skövde, School of Health and Education, Skovde, Sweden; ²⁵CENTOGENE GmbH, Rostock, Germany; ²⁶Inserm, Developmental Brain Disorders Laboratory, Paris, France; ²⁷Sorbonne University, Department of Genetics, Paris, France; ²⁸St George’s University of London, Molecular and Clinical Sciences Institute, London, United Kingdom; ²⁹University of California San Diego, Department of Neurosciences, La Jolla, United States; ³⁰Rady Children’s Institute for Genomic Medicine, San Diego, United States; ³¹Aix-Marseille University, APHM, Marseille, France; ³²Hebrew University Medical Center, Department of Genetics, Jerusalem, Israel; ³³Leibniz-Forschungsinstitut für Molekulare Pharmakologie, Berlin, Germany; ³⁴Free University Berlin, Department of Biology, Berlin, Germany; ³⁵University Hospitals Plymouth NHS Trust, Plymouth, United Kingdom; ³⁶University of Plymouth, Plymouth, United Kingdom

Background/Objectives: Biallelic INPP4A variants have been associated with severe neurodevelopmental disease in single case reports. Our studies expand and elucidate the clinical-genetic spectrum and provide a pathomechanistic explanation for genotype-phenotype correlations.

Methods: Clinical and genomic investigations of 26 individuals with INPP4A-related neurodevelopmental disorder were undertaken alongside molecular and in silico modelling and translation reinitiation studies.

Results: We characterize a clinically variable disorder with cardinal features including global developmental delay, severe-profound intellectual disability, microcephaly, limb weakness, cerebellar signs and short stature. A more severe presentation associated with biallelic INPP4A variants downstream of exon 4 has additional features of (ponto)cerebellar hypoplasia, reduced cerebral volume, peripheral spasticity, contractures, intractable seizures and cortical visual impairment. Our studies identify the likely pathomechanism of this genotype-phenotype correlation entailing translational reinitiation in exon 4 resulting in an N-terminal truncated INPP4A protein retaining partial functionality, associated with less severe disease. We also identified identical reinitiation site conservation in Inpp4a^-/- mouse models displaying similar genotype-phenotype correlation. Additionally, we show fibroblasts from affected individuals exhibit disrupted endocytic trafficking pathways with largely unaffected Akt and mTORC1 signalling pathways, defining the likely biological basis of disease. Lastly, our work defined a new MANE Plus Clinical transcript (NM_001351425.2) as the most relevant to this neurodevelopmental disorder.

Conclusion: Our studies comprehensively characterise INPP4A-related neurodevelopmental disorder and suggest genotype-specific clinical assessment guidelines. We propose the potential mechanistic basis of observed genotype-phenotype correlations entails exon 4 translation reinitiation.

Grants: Wellcome Trust: 209083/Z/17/Z, 221951/Z/20/Z, 226083/Z/22/Z. Medical Research Council: G1002279, MR/S01165X/1, MR/S005021/1. Academy of Medical Sciences: SGL029\1079.

Conflict of Interest: None declared

C12.1 Comprehensive molecular analysis in advanced thymic epithelial tumors uncovers germline alterations in one fifth of patients impacting therapy recommendations

Lino Möhrmann^1;2, Lysann Rostock¹, Maximilian Werner^1;2, Małgorzata Oleś³, Jonas Arnold ^4;5;6, Nagarajan Paramasivam³, Korinna Jöhrens⁷, Luise Rupp⁸, Marc Schmitz⁸, Daniela Richter^1;2;9, Sebastian Uhrig³, Martina Fröhlich³, Barbara Hutter³, Jennifer Hüllein³, Arne Jahn^4;5;6, Marie Arlt^4;5;6, Elena Möhrmann^1;2, Dorothea Hanf^1;2, Laura Gieldon¹⁰, Simon Kreutzfeldt^11;12, Christoph Heilig^11;12, Maria-Veronica Teleanu^11;12, Daniel Lipka¹³, Katja Beck^11;12;14, Annika Baude-Müller^11;12;14, Andreas Mock¹⁵, Ivan Jelas¹⁶, Damian Rietke¹⁶, Marcel Wiesweg¹⁷, Christian Brandts^18;19;20, Melanie Börries^9;21, Anna Lena Illert²², Alexander Desuki^23;24;25, Thomas Kindler^23;24;25, Angela Krackhardt²⁶, Christoph Benedikt Westphalen^27;28, Petros Christopoulos^29;30, Leonidas Apostolidis³¹, Albrecht Stenzinger³², Michael Allgäuer³², Olaf Neumann³², Irina Kerle^1;2;33, Peter Horak^11;12, Christoph Heining^1;2, Heidrun Grosch^30;34, Evelin Schröck^4;5;6, Daniel Hübschmann³, Stefan Fröhling^11;35;36, Hanno Glimm^1;2;37

¹TMO, NCT/UCC, a partnership between DKFZ, Faculty of Medicine and University Hospital Dresden and HZDR; ²TMO, Faculty of Medicine and University Hospital Dresden, Germany; DKTK, Dresden, Germany; ³Computational Oncology Group, Molecular Precision Oncology Program, NCT Heidelberg and DKFZ, Heidelberg, Germany; ⁴Institute for Clinical Genetics, University Hospital Dresden, Dresden, Germany; TMO, (NCT/UCC), a partnership between DKFZ, Faculty of Medicine and University Hospital Dresden and HZDR; ⁵ERN GENTURIS, Hereditary Cancer Syndrome Center Dresden, Germany; TMO, Faculty of Medicine and University Hospital Dresden, Germany; ⁶DKTK, Dresden, Germany; Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany; ⁷Department of Pathology, University Hospital Carl Gustav Carus, Dresden, Germany; ⁸Institute of Immunology, Faculty of Medicine Carl Gustav Carus, TU Dresden, Dresden, Germany; ⁹German Cancer Research Center (DKFZ) Heidelberg, Germany; ¹⁰Institute of Medical Genetics, Carl von Ossietzky University, Oldenburg, Germany; ¹¹Division of Translational Medical Oncology, German Cancer Research Center (DKFZ), Heidelberg, Germany; ¹²National Center for Tumor Diseases (NCT), NCT Heidelberg, a partnership between DKFZ and Heidelberg University Hospital, Heidelberg, Germany; ¹³Section of Translational Cancer Epigenomics, Division of Translational Medical Oncology, German Cancer Research Center (DKFZ) and NCT Heidelberg, Heidelberg, Germany; ¹⁴Molecular Precision Oncology Program, NCT Heidelberg and German Cancer Research Center (DKFZ), Heidelberg, Germany; ¹⁵Institute of Pathology, Ludwig-Maximilians-Universität München (LMU), Munich, Germany; ¹⁶Charité Comprehensive Cancer Center, Universitätsmedizin Berlin, Freie Universität Berlin and Humboldt-Universität zu Berlin, Berlin, Germany; ¹⁷Department of Medical Oncology, West German Cancer Center, University Hospital Essen, University Duisburg-Essen, Germany; ¹⁸Department of Medicine, Hematology/Oncology, University Hospital, Goethe University, Frankfurt, Germany; ¹⁹University Cancer Center Frankfurt (UCT), University Hospital, Goethe University, Frankfurt, Germany; ²⁰German Cancer Consortium (DKTK), Heidelberg, Germany; ²¹Institute of Medical Bioinformatics and Systems Medicine, Medical Center—University of Freiburg, Faculty of Medicine, University of Freiburg and DKTK, partner site Freiburg, a partnership between DKFZ and Medical Center - University of Freiburg; ²²Department of Internal Medicine I, Medical Center—University of Freiburg, Faculty of Medicine, University of Freiburg, Freiburg im Breisgau, Germany; ²³University Cancer Center and Department of Internal Medicine III, University Medical Center Mainz, Germany; ²⁴TRON-Translational Oncology, University Medical Center, Johannes Gutenberg University Mainz, Germany; ²⁵German Cancer Consortium (DKTK), partner site Frankfurt/Mainz, a partnership between DKFZ and University Medical Center Mainz, Germany; ²⁶Klinik und Poliklinik für Innere Medizin III, Klinikum rechts der Isar, School of Medicine, Technical University of Munich (TUM), Munich, Germany; ²⁷Department of Internal Medicine III, University Hospital, LMU Munich and Comprehensive Cancer Center, Munich, Germany; ²⁸German Cancer Consortium (DKTK), partner site Munich, a partnership between DKFZ and University of Munich, Germany; ²⁹Translational Lung Research Center, member of the German Center for Lung Research (DZL), Heidelberg, Germany; ³⁰Department of Thoracic Oncology, Thoraxklinik, Heidelberg University Hospital and National Center for Tumor Diseases (NCT), NCT Heidelberg, a partnership between DKFZ and Heidelberg University Hospital, Germany; ³¹Department of Medical Oncology, NCT Heidelberg and Heidelberg University Hospital, Heidelberg, Germany; ³²Institute of Pathology, University Hospital Heidelberg, Heidelberg, Germany; ³³DKTK, Dresden, Germany; ³⁴Translational Lung Research Center, member of the German Center for Lung Research (DZL), Heidelberg, Germany; ³⁵National Center for Tumor Diseases (NCT), NCT Heidelberg, a partnership between DKFZ and Heidelberg University Hospital, Heidelberg, Germany; DKTK, Heidelberg, Germany; ³⁶Institute of Human Genetics, Heidelberg University, Heidelberg, Germany; ³⁷Translational Functional Cancer Genomics, German Cancer Research Center (DKFZ) Heidelberg, Germany

Background/Objectives: The molecular landscape and genetic risk factors of malignant thymic epithelial tumors (TETs), a clinically challenging entity, are largely unknown. We characterized a cohort of patients with advanced TETs clinically and molecularly using a broad multiomics approach.

Methods: 46 thymic carcinomas (TC), 8 thymic neuroendocrine neoplasms (TNEN) and 27 thymomas from the prospective observational German multicenter precision oncology study DKFZ/NCT/DKTK MASTER were included. Parallel tumor and control genome or exome sequencing, RNA analysis and DNA methylation profiling were performed. A molecular tumor board assessed clinical actionability of somatic and germline alterations.

Results: We observed mutational heterogeneity between entities with targetable alterations being more frequent in TC. Assessment of germline variants within 142 cancer predisposition genes (CPGs) among 76 patients with available germline data revealed 14 pathogenic germline variants (PGVs) in 14 patients (18%) with WHO type B2 thymoma or higher. PGVs in eight autosomal dominant CPGs occurred in nine patients. PGVs in BRCA1 and MUTYH occurred more than once and there was one biallelic pathogenic MUTYH variant. Loss of heterozygosity was observed in five tumors with PGVs in autosomal dominant CPGs. PGVs impacted therapy recommendations in seven patients, all of them targeting impaired DNA damage response. We detected BRCAness-signature SBS3 in 24/67 (36%) samples. Therapy recommendations were applied in 28/68 (41%) cases.

Conclusion: Comprehensive molecular analysis provides clinically relevant information. Based on the higher PGV yield and potential clinical impact on patients and their relatives, we propose germline testing parallel to molecular tumor characterization for all patients with TC, TNEN and B2/B3 thymoma.

Grants: DKFZ-HIPO-H021

Conflict of Interest: None declared

C12.2 Looking beyond coding mutations: the role of regulatory noncoding sequences in hereditary diffuse gastric cancer

Celina São José ¹, Marta Ferreira^1;2, Lilian Cordova³, Ana Maria Pedro¹, Raquel Santos¹, Ricardo Amorim¹, ana andré¹, J. Garcia-Pelaez¹, janine senz³, Pardeep Kaurah³, Fiona Puntieri⁴, Juliane Glaser⁴, david huntsman^3;5, Stefan Mundlos^4;6;7, Kasmintan Schrader³, Carla Oliveira^1;8

¹Instituto de Investigação e Inovação em Saúde, Porto, Portugal; ²Faculty of Science of the University of Porto, Computer Science, Porto, Portugal; ³University of British Columbia, Pathology and Laboratory Medicine, Vancouver, British Columbia, Canada; ⁴Max Planck Institute for Molecular Genetics, Berlin, Germany; ⁵Genetic Pathology Evaluation Centre, University of British Columbia and Vancouver General, Vancouver, British Columbia, Canada; ⁶Berlin-Brandenburg Center for Regenerative Therapies (BCRT), Charité Universitätsmedizin Berlin, Berlin, Germany; ⁷Institute of Medical and Human Genetics, Charité Universitätsmedizin Berlin, Berlin, Germany; ⁸Faculty of Medicine of the University of Porto, Porto, Portugal

Background/Objectives: CDH1 or CTNNA1 germline inactivation explains 10-40% of families with hereditary diffuse gastric cancer (HDGC). Missing heritability prevents proper management and disease prevention in HDGC-suspected families. We hypothesized that defects in CDH1-regulatory elements contribute to HDGC missing heritability.

Methods: We called single-nucleotide/copy-number variants (SNV/CNV) from 19 HDGC probands whole-genome sequencing data, performed gene-ontology analysis, 4C-seq and ATAC-seq in normal stomach epithelia, CRISPR-Cas9, RT-PCR and flow cytometry in cell lines, immunohistochemistry and Microsatellite instability (MSI) analysis in tumours and enhancer assays in mouse embryos.

Results: We found a MLH1 2.7Kb germline deletion overlapping a stomach-specific regulatory element in 1/19 HDGC-suspected families. Gastric tumour displayed MLH1 protein loss and MSI. We crossed ATAC-seq (46.249 peaks) and 4C-seq (370 CDH1 interactor-regions) data from normal stomach with WGS data (1.882 rare CNVs) from 19 HDGC-suspected families. A CDH3 20kb-deletion and a 39bp-intergenic deletion downstream of CDH1 were mimicked with CRISPR-Cas9 (homozygous) triggering 100% or 50% CDH1 mRNA downregulation, respectively, similarly to a CDH1 coding deletion. Two sequences within the CDH3 CNV displayed enhancer activity in mouse embryos. Combined analysis of deleted accessible chromatin regions revealed a pattern of impaired immune-related pathways in 9 families.

Conclusion: We identified regulatory regions within a MLH1 stomach-specific accessible sequence, and within the CDH1 TAD that harbour CNVs potentially explaining the missing heritability in three HDGC-suspected families. Germline CNVs in stomach-specific regulatory regions favouring an immune-deficient phenotype may predispose 9 additional families to HDGC.

Grants: FCT: PTDC/BTM-TEC/30164/2017, PTDC/BTM-TEC/6706/2020, SFRH/BD/140796/2018; EU: SOLVE-RD_Ref 22184, HORIZON_HORIZON-HLTH-2022-CARE-08.

Conflict of Interest: None declared

C12.3 Identification and functional characterization of MSH2 variants with partial effects at RNA splicing and/or protein level: implications for Lynch syndrome diagnosis

Manon Quilan ¹, Nawel Malouche², Abhijit Rath³, Marine Cauchois-Lemière¹, Laetitia Meulemans¹, Aurelie Drouet¹, Julie Leclerc⁴, Marie-Pierre Buisine⁴, Florence Coulet², Stéphanie Baert-Desurmont⁵, Pascaline Gaildrat¹, Christopher D. Heinen³, Martine Muleris², Alexandra Martins¹

¹Inserm U1245, Inserm, Univ Rouen Normandie, Rouen, France; ²Département de Génétique Médicale et Inserm UMR-S 958, Hôpital Pitié-Salpêtrière, AP-HP, Sorbonne Université, Paris, France; ³Center for Molecular Oncology, UConn Health, Farmington, CT, United States; ⁴UMR9020-U1277 CANTHER, Univ. Lille, CNRS, Inserm, CHU Lille, and Molecular Oncogenetics Unit, Department of Biochemistry and Molecular Biology, Lille University Hospital, Lille, France; ⁵Inserm U1245 et Département de Génétique Moléculaire, CHU Rouen, Univ Rouen Normandie, Rouen, France

Background/Objectives: MSH2 is a major gene involved in Lynch syndrome (LS), an inherited predisposition to cancer, especially to colorectal/endometrial tumours. Identification of a causal variant is essential for LS diagnosis. However, ≥30% of the variants detected in MSH2 remain classified as of uncertain significance (VUS). It is therefore critical to develop strategies to improve their interpretation, particularly of those with partial/dual effects at RNA and/or protein level.

Methods: We identified and characterized such VUS using three complementary approaches: (i) cell-based minigene splicing assays (ii) protein functional tests based on cDNA complementation of MSH2-deficient cells and methylation tolerance assessment, and (iii) integrated RNA/protein analyses using CRISPR/Cas9-modified human embryonic stem cells (hESCs-tests). We chose MSH2 exon 5 (MSH2ex5) as a model-system because this exon is essential for MSH2 function.

Results: Our minigene assays uncovered multiple non-coding (intronic/synonymous) and missense MSH2ex5 variants causing RNA splicing defects of different severities (weak to total exon skipping). Selected missense variants were subsequently analysed for their impact on protein function, which revealed a large spectrum of outcomes (none, intermediate, and drastic defects). Preliminary results from the hESCs-tests on the non-coding spliceogenic variants indicate that hESCs can tolerate a substantial decrease in full-length MSH2 transcript produced by the mutant MSH2 alleles without significantly impacting DNA damage response nor DNA mismatch repair functions. Three missense variants with partial impacts on splicing, combined or not with intermediate protein effects, were also selected for hESCs-tests.

Conclusion: This work contributes to the interpretation of MSH2 VUS, notably those with partial/dual effects at RNA/protein levels.

Grants: RIN2021-100, INCA_16094.

Conflict of Interest: None declared

C12.4 BPTF Mutation Unravels Viral Oncogenicity in Heritable Kaposi Sarcoma

Yuval Yogev ^1;2, Moshe Schaffer^3;4, Mark Shlapobersky^4;5, Matan M. Jean¹, Ohad Wormser¹, Max Drabkin¹, Daniel Halperin¹, Riad Kassem^6;7, Alexandra A. Tsitrina⁸, Noam Asna^3;4, Alejandro Livoff^4;5, Ohad Shmuel Birk^1;9

¹Ben-Gurion University of the Negev Faculty of Health Sciences, The Morris Kahn Laboratory of Human Genetics at the National Institute of BiotechnologyThe Morris Kahn Laboratory of Human Genetics at the National Institute of Biotechnology, Beer-Sheva, Israel; ²The Edmond and Lily Safra Children’s Hospital, Sheba Medical Center, Ramat Gan, Israel; ³Barzilai University Medical Center, Department of Oncology, Ashkelon, Israel; ⁴Ben-Gurion University of the Negev, Faculty of Health Sciences, Beer-Sheva, Israel; ⁵Barzilai University Medical Center, Department of Pathology, Ashkelon, Israel; ⁶Sheba Medical Center, Department of Dermatology, Ramat Gan, Israel; ⁷Tel Aviv University, Sackler School of Medicine, Tel Aviv, Israel; ⁸Ben-Gurion University of the Negev, Ilse Katz Institute for Nanoscale Science and Technology, Beer-Sheva, Israel; ⁹Soroka University Medical Center, Genetics Institute, Beer-Sheva, Israel

Background/Objectives: Kaposi sarcoma (KS), a rare neoplasm caused by Kaposi’s sarcoma-associated herpesvirus (KSHV), exhibits distinct sub-forms with varied clinical presentations. While classical KS (cKS) is predominantly sporadic, rare familial cases suggest genetic predisposition.

Methods: We utilize WES for variant characterization, and functional assays using CRISPR-KI alongside with reporter viruses, confocal0microscopy, flow-cytometry and transcriptomics.

Results: We report two Moroccan Jewish families with seven individuals across two generations affected by cKS, segregating with a heterozygous missense mutation (p.I2012T) in Bromodomain and PHD finger-containing transcription factor (BPTF), known to interact with the KSHV protein LANA-1. Functional studies using a rKSHV.219-infected CRISPR-KI model reveal that the BPTF^I2012T variant promotes a higher latent-to-lytic infection ratio, with increased LANA expression and decreased virion production. Transcriptomic analysis highlights dysregulated pathways related to tumorigenesis, immune response, blood vessel development pathways and viral genome replication, consistent with cKS. Notably, BPTF^I2012T enriched pathways of viral genome regulation and replication, immune response, and chemotaxis, including downregulation of IFI16, SHFL HLAs, TGFB1 and HSPA5, all previously associated with KSHV infection and tumorigenesis, and influences NF-κB pathway, suggesting its role in KSHV oncogenesis.

Conclusion: Our findings suggest that BPTF plays a crucial role in KSHV-induced tumorigenesis by modulating viral life cycle and immune response. Understanding the genetic basis of heritable KS sheds light on the molecular mechanisms underlying viral oncogenicity and may lead to targeted therapies for this rare malignancy

Grants: Israel Science Foundation (grant no. 2034/18) the Foulkes foundation, ISCORT grant, the National Knowledge Center for Rare/Orphan Diseases of the Israel Ministry of Science, Technology and Space.

Conflict of Interest: None declared

C12.5 A FGFR1 susceptibility allele modulates the oncogenic potential of driver mutations in low-grade brain tumors

Jacopo Boni ¹, Míriam Fernández-González¹, Hyerim Han¹, Cassandra Wong², Jonathan Boulais³, Islam Elkholi³, Carla Roca¹, Cristina Rioja¹, Clara Nogué¹, Cristina Jou⁴, Anne-Claude Gingras², Jean-François Côté³, Barbara Rivera^1;5;6

¹IDIBELL Institut d’Investigació Biomèdica de Bellvitge, L’Hospitalet de Llobregat, Spain; ²Lunenfeld-Tanenbaum Research Institute, Sinai Health System, Toronto, Canada; ³Institut de recherches cliniques de Montréal (IRCM), Montréal, Canada; ⁴Sant Joan de Déu Barcelona Hospital, Pathology department, Esplugues de Llobregat, Spain; ⁵Lady Davis Institute at the Jewish General Hospital, Montréal, Canada; ⁶McGill University, Gerald Bronfman Department of Oncology, Montréal, Canada

Background/Objectives: Alterations in the FGFR1 receptor are associated with a wide range of human diseases, including different types of brain tumors. We identified a germline FGFR1 variant (p.R661P) in a family with DNETs (low-grade glioneuronal tumors). Secondary oncogenic somatic hits (p.N546K or p.K656E) in-cis with the p.R661P were identified in the tumors of these individuals. This pattern of multiple FGFR1 mutants was validated in therapy-naïve sporadic cases, suggesting intrinsic mechanisms of selective pressure that promote FGFR1 mutational events in the context of a quiet genome.

Methods: To gain insights into the molecular mechanisms triggered by different FGFR1 mutations, we mapped the proximal interactome of FGFR1 mutants through a BioID-MS approach. We generated FGFR1-mutant human oligodendroglioma cell lines by CRISPR-Cas9 editing to assess cell phenotypic changes.

Results: The study revealed similar profiles for the germline p.R661P and the WT condition, whereas all oncogenic mutants show pronounced differences in their proximal interactome. We identified a modulating effect played by the R661P variant, hampering the oncogenic potential of N546K and K656E mutants by reducing affinity for downstream effector PLC-y and enhancing receptor degradation, respectively. The p.R661P mutation alone abolished the self-renewing capacity of these cells, while in double mutants (N546K-R661P, K656E-R661P) weaking the proliferative potential of the oncogenic N546K and K656E clones.

Conclusion: These results reveal a severe impairment phenotype associated to the susceptibility variant, likely compensated in patients by the presence of the wt copy, and, suggest oncogenic modulation potential exerted by multiple FGFR1 alterations during low-grade brain tumorigenesis.

Grants: ALSF YIA award & la Caixa (LCF/BQ/PI19/11690009) award.

Conflict of Interest: None declared

C12.6 Genomic mechanisms of aggressive and metastatic phenotypes in non-melanoma skin cancers

Andrey Yurchenko ¹, Andrea Forschner², Fatemeh Rajabi¹, Lukas Flatz², Sergey Nikolaev¹

¹Gustave Roussy Cancer Campus, Université Paris Saclay, France, INSERM U981, Villejuif, France; ²University Hospital Tübingen, University of Tübingen, Department of Dermatology, Tübingen, Germany

Background/Objectives: Basal Cell Carcinoma (BCC) and cutaneous Squamous Cell Carcinoma (cSCC) are the most common human malignancies of keratinocyte origin with increasing incidence. Usually, these cancers can be effectively treated, but rarely, they can be aggressive or metastatic with very poor prognosis and still unknown genomic mechanisms.

Methods: We have assembled the largest to date cohorts of BCC (n = 389) and cSCC (n = 223) representing major histological subtypes, lymph node metastasis and distant metastasis. The samples were profiled with WES/WGS/RNA-seq and clinically annotated.

Results: The mutational profiles of BCC and cSCC were determined by UV-light exposure with high TMB (80 and 100 mt/mb respectively) and a median of six cancer driver alterations per sample. We found that loss of TP53 is typically the first driver event in keratinocyte carcinomas and is associated with a dramatic increase in successive driver mutation acquisition. Aggressive morpheaphorm BCC samples did not demonstrate specific genetic alterations but had fibrotic TME and increased activity of HIPPO-YAP pathway according to RNA-seq analysis and immunohistochemistry. Metastatic BCC were characterised by high genomic instability and aneuploidy not typical for this type of cancer. Metastatic cSCC tumours had an increased frequency of MAPK pathway alterations and high levels of aneuploidy and whole genome duplication events. These findings were confirmed by an independent cohort of the GENIE project. Lymph node metastasis of cSCC samples demonstrated lymphocyte-enriched TME favourable for immunotherapeutic treatment.

Conclusion: Genomic instability and transcriptional activation of HIPPO-YAP and MAPK pathways define the high-risk and metastatic potential of keratinocyte carcinomas.

Grants:

Conflict of Interest: None declared

C13.1 Uncovering cell-type specific gene-regulation using single-cell multiomics in the sc-eQTLGen consortium

Maryna Korshevniuk ¹, Dan Kaptijn¹, Roy Oelen¹, Martijn van der Werff¹, Jelmer Niewold¹, Harm-Jan Westra¹, Lude Franke¹, Monique van der Wijst¹, Marc Jan Bonder¹

¹University Medical Center Groningen, Genetics Department, Groningen, Netherlands

Consortium: sc-eQTLGen

Background/Objectives: One approach to interpreting disease-related genetic variation is through the identification of expression quantitative trait loci (eQTLs). Linking genetic variation to single-cell expression data (scRNA-seq) enables the identification of cell type-specific eQTLs (ct-eQTLs), thus providing insight into the relevant context in which these variants act. Overlaying ct-eQTLs with cell type-specific DNA accessibility data (scATACseq) allows us to gain insight into the mechanisms underlying ct-eQTLs and suggest potential disease mechanisms.

Methods: Within the sc-eQTLGen consortium we conducted the largest ct-eQTL mapping to date in PBMC (n-donors 2,000). Furthermore, we generated a large single-cell multiomics (scMO) dataset on PBMCs derived from 271 healthy individuals to gain mechanistic insights into the ct-eQTL SNPs.

Results: Here we present the first results of the sc-eQTLGen consortium where we identified in total 14,749 eQTLs (FDR 10%) for the six major PBMC cell types corresponding to 7,310 eGenes and 2,810 unique ct-eQTLs. In the preliminary analysis of the scMO data we identified several overlapping ct-specific ATAC peaks and ct-eQTLs. Among them is a monocytes-specific eQTL on the TREM1 which overlaps a KLF family motive differentially accessible in monocytes. Leveraging our complete sc-MO dataset, we aim to identify ct-specific regulators of gene expression, thus providing deeper insights into the regulatory mechanisms of the ct-eQTLs.

Conclusion: Integrating scMO data enhances our understanding of cell-type-specific gene regulation mediated by eQTLs. This study underscores the importance of such integrative approaches in elucidating the genetic mechanisms underlying complex traits and diseases, with potential implications for precision medicine and disease research.

Conflict of Interest: None declared

C13.2 Spatial dissection of colorectal cancer initiation using whole transcriptome imaging

helena crowell¹, hu zedong², Michael Rhodes³, Joseph Beechem³, sabine tejpar², holger heyn¹, Anna Pascual Reguant ⁴

¹Centre Nacional d’Anàlisi Genòmica, Single Cell Genomics, Barcelona, Spain; ²KU Leuven, Laboratory of Molecular Digestive Oncology, Leuven, Belgium; ³NanoString® Technologies, Seattle, United States; ⁴Centre Nacional d’Anàlisi Genòmica, Spatial Genomics, Barcelona, Spain

Background/Objectives: Revolutionizing clinical pathology requires new technologies that excel currently applied approaches that multiplex selected biomarkers to guide clinical decisions. While higher-plex imaging techniques recently became available, none is scalable to appreciate the full spectrum of molecular biomarkers holistically. We argue that biomarker discovery requires substantially more features to cope with the complexity of diseases. Here, thousands of genes or proteins, imaged at single-cell resolution, will be required to comprehensively identify diagnostic and therapy-prognostic biomarkers.

Methods: Spatial Molecular Imaging (SMI) allows for transcriptome-wide analyses at molecule resolution, enabling spatial single-cell profiling at a definition that was only achieved through scRNA-seq until now. Through digitalization of full transcriptomes in tissue sections and improving the current state-of-the-art methods in pathology by orders of magnitude, next-generation pathology tools are in sight.

Results: We applied SMI’s whole transcriptomics to spatially resolve tumor evolution within colon samples, progressing from premalignant lesions (tubullovillous adenoma) to colorectal cancer (CRC). We generated spatially resolved single-cell landscapes of these tissues and detected vast spatial heterogeneity that matched H&E pathologist annotations, with intertwined cell state niches resembling stem cells or differentiated colonocytes, phenotypes previously linked to the progression into different CRC subtypes. We associated gene programs of stemness, fetal development and immunogenicity, the latter spatially correlated with immune-cell infiltration patterns.

Conclusion: Importantly, the enrichment of clinically relevant pre-malignant cell states at discrete niches could only be resolved using whole transcriptome spatial data and not through scRNA-seq, providing evidence of the method’s superiority in enabling next-generation pathology strategies.

Grants: RYC2022-035848-I

Conflict of Interest: helena crowell: None declared, hu zedong: None declared, Michael Rhodes NanoString Technologies, NanoString Technologies, Joseph Beechem NanoString Technologies, NanoString Technologies, sabine tejpar: None declared, holger heyn co-founder of Omniscope, SAB member of Nanostring and MiRXES, and consultant to Moderna and Singularity, Anna Pascual Reguant: None declared

C13.3 Convergence of neurodevelopmental signatures in Prader-Willi syndrome from an allelic series of CRISPR-engineered neuronal models

Celine De Esch ¹, Derek Tai¹, Xander Nuttle¹, Rachita Yadav¹, Yating Liu¹, Jaewon Shin¹, Nicholas Burt¹, Serkan Erdin¹, John Lemanski¹, Bimal Jana¹, Benjamin Currall¹, Yulia Mostovoy¹, Dadi Gao¹, Lisa Burnett¹, James Gusella¹, Michael Talkowski¹

¹Massachusetts General Hospital/Harvard Medical School, Center for Genomic Medicine, Boston, United States

Background/Objectives: Prader-Willi Syndrome (PWS) is a neurodevelopmental disorder caused by a loss of expression of paternally inherited genes from the imprinted chromosome 15q11-q13 region. Recurrent deletions spanning 5-6 Mb (Type I/II deletions) are the most common cause of PWS and the affected genomic region includes five paternally expressed protein-coding genes, six small nucleolar RNAs (SNORDs), and IPW, a long non-coding RNA. To investigate the potential drivers of neurodevelopmental changes associated with PWS, we created an allelic series of human induced pluripotent stem cell (hiPSC) models harboring CRISPR-engineered paternal deletions across the PWS locus and characterized derivative neuronal models.

Methods and Results: We differentiated 221 isogenic hiPSC models with PWS-related mutations and matched controls for each edit, including Type I/II deletions, the critical region deletion (CRD, including SNORD116, IPW, and SNORD109A), and individual gene deletions in male and female iPSC backgrounds into Ngn2-induced neurons (iNs) followed by systematic transcriptomic analyses (small RNA, mRNA), along with functional microelectrode array (MEA). Meta-analyses of differentially expressed and co-expressed genes across Type I/II edits were strongly enriched for nervous system and synaptic functions (false discovery rate < 0.1), including retrograde endocannabinoid signaling, GABAergic synapse, serotonergic synapse, and taste transduction pathways. The Type I/II edits displayed significant (p < 0.05) reductions in synaptic activity, synchrony (strength of synaptic connections), and oscillation (network activity).

Conclusion: Collectively, our findings demonstrate the functional impact of PWS-associated gene deletions in human-derived neural models and highlight alterations to neuronal and endocrine pathways associated with PWS pathogenesis.

Conflict of Interest: Celine De Esch PWS grant, MGH, Derek Tai: None declared, Xander Nuttle: None declared, Rachita Yadav: None declared, Yating Liu: None declared, Jaewon Shin: None declared, Nicholas Burt: None declared, Serkan Erdin: None declared, John Lemanski: None declared, Bimal Jana: None declared, Benjamin Currall: None declared, Yulia Mostovoy: None declared, Dadi Gao: None declared, Lisa Burnett: None declared, James Gusella: None declared, Michael Talkowski: None declared

C13.4 Colocalization between human aortic smooth muscle cells molecular QTLs and vascular disease GWAS identified loci with pleiotropic effect

Ilakya Selvarajan¹, Nadine Odenthal ^2;3;4, Luis Eichelmann^2;3;4, Tobias Reinberger^2;3;4, Zouhair Aherrahrou^2;3;4, Minna Kaikkonen-Määttä¹, Redouane Aherrahrou^1;2;3;4;5

¹A.I. Virtanen Institute for Molecular Sciences, University of Eastern Finland, Kuopio, Finland; ²Institute for Cardiogenetics, University of Lübeck, Lübeck, Germany; ³DZHK (German Centre for Cardiovascular Research); ⁴University Heart Centre Lübeck, Lübeck, Germany; ⁵Center for Public Health Genomics, University of Virginia, Charlottesville, United States

Background/Objectives: Smooth muscle cells (SMCs) play a crucial role in cardiovascular diseases (CVD), undergoing phenotypic changes under pathological conditions, contributing to vascular remodeling and CVD pathogenesis. This study aims to identify genetic variants regulating SMC gene expression and phenotypes associated with vascular disease-related loci from genome-wide association studies (GWAS).

Methods: SMCs from 151 heart transplant donors of various genetic ancestries were cultured in proliferative and quiescent conditions and the expression and splicing quantitative trait loci (eQTL and sQTL) were conducted. Colocalization analysis was performed between these SMCs molecular QTL and GWAS loci associated with vascular disorders to identify effector transcripts for CVD pathogenesis. Potential risk genes were prioritized and validated for their effects on the vascular phenotype using in vivo and in vitro experiments

Results: We found 336 eQTLs and 580 sQTLs co-localizing with CVD loci, showing genetic regulation of mRNA expression and splicing in CVD genetic risk. Filtering these loci for accessibility to chromatin regions and high expression in SMCs revealed key genes, such as MAP3K7CL, which colocalizes with CAD, heart attack, and blood pressure GWAS loci. MAP3K7CL modulates SMC migration, proliferation, and calcification and its knockdown in zebrafish altered blood flow, heart rhythm patterns, and angiogenesis, highlighting its role in CVD pathogenesis.

Conclusion: Our study reveals that SMCs influence vascular disorders, and their genetic regulation is linked to CVD pathogenesis. These findings may guide targeted therapeutic interventions for vascular diseases.

Grants: This work was supported by Finnish Foundation for Cardiovascular Research, DZHK and the CORONA Foundation.

Conflict of Interest: None declared

C13.5 Cell type signatures in cell free DNA fragmentation profiles reveal disease biology

Kate Stanley ¹, Joris Vermeesch², dhanya sudhakaran³, Stefania Tuveri³, Tatjana Jatsenko³, Bernard Thienpont³, Kristel Van Calsteren³, Lore Lannoo³, Kris Van Den Bogaert³, Kevin Punie³, Marie de Borre³, sabine tejpar³, Liesbeth Lenaerts³, Peter Vandenberghe³, Rodrigo Almeida³, Ilse Parijs³, Leen Vancoillie³, Laura Rengifo³

¹KU Leuven, Centre for Human Genetics, Leuven; ²KU Leuven, Centre for Human Genetics, Leu, Belgium; ³KU Leuven, Centre for Human Genetics, L, Belgium

Background/Objectives: Circulating cell free DNA (cfDNA) fragments have characteristics that are specific to the cell types that release them. Current methods for cfDNA deconvolution typically use disease tailored marker selection in a limited number of bulk tissues or cell lines. We utilize single cell transcriptome data as a comprehensive cellular reference set for disease-agnostic cfDNA cell-of-origin analysis.

Methods: We performed whole-genome paired-end sequencing of plasma cfDNA from healthy individuals and patients with cancer and pregnancy complications. Sequencing data was aligned to the human reference genome and the position of fragments was used to infer nucleosome spacing. We then correlated cfDNA-inferred nucleosome spacing with gene expression data from the Tabula Sapiens database to rank the relative contribution of over 450 cell types to plasma cfDNA.

Results: In 744 healthy individuals and patients, we uncover cell type signatures in support of emerging disease paradigms in oncology and prenatal care. We trained predictive models that differentiated patients with colorectal cancer (84.7%), early-stage breast cancer (90.1%), multiple myeloma (AUC 95.0%), and preeclampsia (88.3%) from matched controls using cell type ranks. Importantly, our approach performed well in ultra-low coverage cfDNA datasets.

Conclusion: CfDNA fragmentation patterns can be used to estimate the relative abundance of fragments from different cell types. These relative abdundances are altered under certain conditions and can be used to predict disease in diverse clinical settings for the potential expansion of liquid biopsy.

Grants: EU Horizon 2020 (824110, 813707), Kom Op Tegen Kanker (KOTK/2018/11468), FWO-SBO (S003422N), and Agentschap Innoveren en Ondernemen (HBC.2018.2108).

Conflict of Interest: None declared

C13.6 Interpreting GWAS Signals and the Genetic Architecture of Ventricular Hypertrophy in African Americans trough the Context of Cellular Expression Networks in Human iPSC-Cardiomyocytes.

Jakob Blamer¹, Andrea Matter¹, Elizabeth Virlee¹, Amy Turner¹, Ulrich Broeckel ¹

¹Medical College of Wisconsin, Pediatrics, Milwaukee, United States

Background: GWAS studies provide many susceptibility loci for common diseases, yet the underlying genetic architecture is not well understood. Evidence indicates that GWAS-identified genes interact systematically through networks. We explore this concept using results from a GWAS study of Left Ventricular Hypertrophy (LVH) in an African American cohort combined with relevant expression data.

Methods: We generated hiPSC lines from 110 participants in the NHLBI HyperGEN GWAS study and differentiated those into cardiomyocytes (hiPSC-CMs) for expression analysis. We identified a set association signals in the full HyperGEN cohort and mapped variants to respective candidate genes. Subsequently, we paired GWAS results with hiPSC-CM cardiac phenotypes and expression data to prioritize genes using network-based analyses. We identified expression modules and functionally relate those to LVH which are enriched for GWAS genes. Finally, we performed CRISPR gene knockout studies of candidate genes to explore changes in expression and network properties.

Results: We identified two distinct expression modules to be functionally related to LVH and enriched in candidate GWAS genes. Additionally, GWAS genes were local hubs in one of these two modules. We identified two novel (ETV6, KLF2) and one known (BNC2) transcription factor GWAS genes differentially co-expressed between hiPSC-CMs from normal and affected donors. Knock-down of these genes through CRISPR/Cas9 in hiPSC-CM resulted in significant hypertrophy marker expression changes and strong enrichment of cardiac pathways.

Conclusion: Network-based analyses using disease relevant cells allow for the functional annotation of GWAS genes. Human iPSCs combined with CRISPR editing provide a platform for the functional annotation of GWAS results.

Grants: NHLBI R56HL167896

Conflict of Interest: Jakob Blamer: None declared, Andrea Matter: None declared, Elizabeth Virlee: None declared, Amy Turner: None declared, Ulrich Broeckel NHLBI

C14.1 A multi-omics approach increases the diagnostic yield ‘beyond the exome’ in unsolved patients with inherited eye diseases

Miriam Bauwens ^1;2, Marta Del Pozo-Valero^1;2, Marieke De Bruyne^1;2, Filip Van den Broeck^3;4, Quinten Mahieu^1;2, Mattias Van Heetvelde^1;2, Stijn Van de Sompele^1;2, Alfredo Dueñas Rey^1;2, Audrey Meunier⁵, Thomy de Ravel⁶, Joke Ruys³, Irina Balikova⁷, Sascha Vermeer⁸, Julie De Zaeytijd^3;4, Bart Leroy^3;4, Elfride De Baere^1;2

¹Center for Medical Genetics, Ghent University Hospital, Gent, Belgium; ²Department of Biomolecular Medicine, Ghent University, Universiteit Gent, Gent, Belgium; ³Department of Ophthalmology, Ghent University Hospital, Gent, Belgium; ⁴Department of Head & Skin, Universiteit Gent, Gent, Belgium; ⁵Department of Ophthalmology, University Hospital Saint-Pierre, Université Libre de Bruxelles (ULB), Brussels, Belgium; ⁶Center for Medical Genetics, University Hospital Brussels, Brussels, Belgium; ⁷Department of Ophthalmology, UZ Leuven, Leuven, Vlaams-Brabant, Belgium; ⁸Center for Human Genetics, UZ Leuven, Leuven, Vlaams-Brabant, Belgium

Background/Objectives: Exome sequencing (ES)-based testing in patients with inherited eye diseases (IED) has an overall diagnostic yield of ~60%. To overcome the missing heritability in molecularly undiagnosed ES-tested IED patients, we adopted a multi-omics approach, combining genomics, transcriptomics and functional assays.

Methods: Short-read genome sequencing (GS) of 111 cases (103 probands) was analyzed using Franklin (Geenox) and Seqplorer (in-house). Coding single nucleotide variants (SNVs) in known IED genes were prioritized, followed by non-coding variants assessed for an effect on splicing or cis-regulation, and by structural variants (SVs). Candidate genes were selected based on single-cell expression profiles and function. GS-data directed whole-blood RNA-seq and/or functional assays.

Results: A likely genetic diagnosis was obtained for 44.5% of probands. Our multi-omics approach pinpointed likely pathogenic non-coding variants in 18.5%, predominantly affecting splicing (16.5%). These include OPA1 variant c.1608+622A > G, leading to pseudo-exon inclusion (PEI) and a novel OPA1 variant, leading to out-of-frame PEI. A first deep-intronic ALMS1 variant and a novel deep-intronic RPGRIP1 variant were also shown to induce aberrant splicing. Moreover, we demonstrated a reduced (40%) regulatory activity of a novel RPGRIP1 promotor variant. Variants in novel candidate genes, including SAMD7 and GPATCH11, were identified in 14.5%, while coding SNVs (8.5%) and SVs (3%) impacting known IED genes were found in 11.5%.

Conclusion: Our multi-omics approach solved missing heritability in 44.5% of molecularly undiagnosed IED cases, mainly attributed to non-coding deep-intronic splicing variants and variants in novel genes. Overall, its added diagnostic value urges its implementation in standard-of-care genetic testing of IED.

Grants: EJPRD19-234, BeSolveRD, BOF20/GOA/023

Conflict of Interest: None declared

C14.2 Mutations in COQ8B are a novel cause of recessive non-syndromic retinitis pigmentosa.

Karolina Kaminska ^1;2, Ana Belén Iglesias-Romero^1;2, Cristina Santos^3;4, Siying Lin^5;6, Mathieu Quinodoz^1;2, Marc Folcher^1;2, Joaquim Calado⁷, Gavin Arno^5;6, Andrew R Webster^5;6, Ana Berta Sousa^8;9, Luisa Coutinho Santos³, Carlo Rivolta^1;2;10

¹Institute of Molecular and Clinical Ophthalmology Basel (IOB), Basel, Switzerland; ²University of Basel, Department of Ophthalmology, Basel, Switzerland; ³Instituto de Oftalmologia Dr. Gama Pinto, Lisboa, Portugal; ⁴NOVA Medical School, iNOVA4Health, Lisboa, Portugal; ⁵Moorfields Eye Hospital NHS Foundation Trust, National Institute of Health Research Biomedical Research Centre, United Kingdom; ⁶University College London, UCL Institute of Ophthalmology, United Kingdom; ⁷NOVA Medical School, ToxOmics, Lisboa, Portugal; ⁸Hospital de Santa Maria, Department of Medical Genetics, Lisboa, Portugal; ⁹University of Lisbon, Laboratory of Basic Immunology, Faculty of Medicine, Lisboa, Portugal; ¹⁰University of Leicester, Department of Genetics and Genome Biology, United Kingdom

Background/Objectives: Retinitis pigmentosa (RP) is a highly heterogeneous genetic disease characterized by progressive retinal blindness. Mutations in more than 90 genes, with very diverse functions, have been so far identified to be associated with this condition.

The objective of this study was to determine the molecular cause of RP in 5 patients from 4 unrelated families, who were all negative for mutations in previously known disease genes.

Methods: All patients underwent a detailed ophthalmic examination. Genotyping was performed by Next Generation or Sanger sequencing.

Mutant or wild-type versions of COQ8B were obtained by transfecting HEK293T cells with synthetic expression plasmids. Then, an assay using Bioluminescence Resonance Energy Transfer Technology (NanoBRET) was used to assess the functional consequences of the variants on COQ8B activity. Western blots were performed on extracts from human retinas.

Results: Analysis of sequencing data identified biallelic assortments of 5 different variants in COQ8B (3 missense and 2 nonsense), a gene involved in the biosynthesis of coenzyme Q, in all patients. Segregation analysis within families confirmed the recessive nature of the disease.

NanoBRET assays showed that COQ8B bearing the combination of variants detected in patients had significantly reduced ligand binding affinity, compared to the wild-type protein. Western blots showed the presence of COQ8B in the human retina.

Conclusion: We identified a new disease gene associated with recessive and non-syndromic RP, COQ8B. Functional studies indicate that the variants detected in patients likely impair COQ8B protein function, leading to defects in coenzyme Q biosynthesis and, ultimately, to retinal disease.

Grants:

Conflict of Interest: None declared

C14.3 High throughput screening: identification of candidate molecules for the treatment of SLC7A8- dependent age-related hearing loss (ARHL)

Paola Tesolin ¹, Paola Bartoccioni^2;3;4, Giulia Pianigiani¹, Maria Caballero², Anna Morgan¹, Cristina Bon¹, Gema Rojas Granado², Israel Ramos², Patrick Aloy², Paolo Gasparini^1;5, Virginia Nunes^6;7, Giorgia Girotto^1;5, Manuel Palacin^2;3;4

¹Institute for Maternal and Child Health – IRCCS, Burlo Garofolo; ²Institute for Research in Biomedicine (IRB Barcelona), The Barcelona Institute of Science and Technology; ³Centro de Investigación Biomédica en Red de Enfermedades Raras (CIBERER), Instituto de Salud Carlos III; ⁴University of Barcelona, Department of Biochemistry and Molecular Biomedicine, Faculty of Biology; ⁵University of Trieste, Department of Medicine, Surgery and Health Sciences; ⁶Genes, Disease and Therapy Program, Molecular Genetics Laboratory - IDIBELL; ⁷University of Barcelona, Genetics Section, Physiological Sciences Department, Health Sciences and Medicine Faculty

Background/Objectives: ARHL is a common multifactorial sensory impairment, but few causative genes have been identified, hampering the development of targeted therapies. Our previous study (PMID:29355479) proposed SLC7A8 as a novel candidate by showing that knock-out (KO) mice develop ARHL and by identifying damaging variants in humans displaying ARHL. These findings suggest that SLC7A8 is an interesting therapeutic target to investigate.

Methods: We performed a high throughput screening (HTS) to detect SLC7A8 transcription regulators. A reporter line was generated by inserting the luciferase gene in SLC7A8 genomic locus. Subsequently, it was incubated for 48 hours with a library of 2240 approved drugs, 960 drug-like compounds and 44 molecules selected as regulators of SLC7A8 transcription according to literature data.

Results: The HTS led to the identification of 56 hit compounds. The most promising ones were re-screened to define the dose-response curve in luciferase and cell viability assays. Furthermore, three molecules, which are already approved drugs, were finally selected. The Compound-1 mechanism of action is linked with increased SLC7A8 expression. Compound-2 was chosen to represent a drug category over-present among the hits. Finally, Compound-3 is already used to treat other forms of hearing loss. An RT-qPCR confirmed that the treatment significantly increased SLC7A8 expression in the analysed cell lines.

Conclusion: These results highlight the possible use of the selected molecules in treating SLC7A8-dependent ARHL. Subsequently, the hits will be tested in HEK293T carrying the variants identified in humans, in inner ear-related cell lines, and in Slc7a8 heterozygous KO mice.

Conflict of Interest: None declared

C14.4 Whole genome sequencing and functional analysis in families with inherited retinal dystrophies (IRDs)

Valerie Wenzel ¹, Tanja Plieger¹, Anna-Zoe Häusler¹, Christina Kiel¹, Heidi Stöhr¹, Georg Spital², Ines Schönbuchner³, Herbert Jägle⁴, Ulrich Kellner⁵, Bernhard Weber^1;6

¹University Regensburg, Institute for Human Genetics, Regensburg, Germany; ²St. Franziskus-Hospital Münster, Eye Center, Münster, Germany; ³Human Genetics Ostbayern, Germany; ⁴University Hospital Regensburg, Clinic and Polyclinic for Ophthalmology, Regensburg, Germany; ⁵Eye Center Siegburg, MVZ Ophthalmologic Diagnostic and Therapy Center Siegburg GmbH, Center for Rare Retinal Diseases, Münster, Germany; ⁶University Hospital Regensburg, Institute of Clinical Human Genetics, Regensburg, Germany

Consortium: Bavarian Genomes Network for Rare Disorders, Munich, Deutschland

Background: Inherited retinal dystrophies (IRDs) are a group of rare diseases characterized by phenotypic and genetic heterogeneity. Despite major advances in molecular genetic diagnosis, approximately 30-40% of IRD cases remain unsolved. Whole Genome Sequencing (WGS) represents a promising approach in routine diagnostics by capturing genetic variants throughout the human genome and addressing structural variation as disease causation. The study aimed to evaluate the efficacy of WGS in routine diagnostics.

Methods: As part of the multicenter study “Bavarian Genomes Network for Rare Disorders”, WGS was performed in 65 IRD families comprising 143 individuals with no clear result after exhaustive analysis with traditional DNA diagnostics tools. WGS data evaluation focused on variant type, allele frequency, computational predictions, and homozygosity analysis. Furthermore, a luciferase reporter assay was established to study the effects of non-coding variants on promoter activity.

Results: WGS analysis determined disease causation in 13 of 65 (20%) families. Moreover, WGS analysis identified at least one disease-causing gene candidate in 10 families (15%). The latter findings include a promising homozygous non-coding variant located in the Rhodopsin (RHO) promoter, found in a patient of consanguineous parents. A luciferase assay studying the functional consequences of the identified variant showed strongly reduced RHO promoter activity.

Conclusion: This study underscores the advantages of WGS for the detection of genetic variants in DNA diagnostics. It should be noted, however, that elaborate functional analyses are generally indispensable to determine pathogenicity of candidate variants.

Grants: The “1,000 clinical genomes” project was funded by “Masterplan Digital II” of the Bavarian State Ministry of Science and Arts.

Conflict of Interest: None declared

C14.5 Biallelic loss-of-function variants in NDUFA7 cause autosomal recessive Leber’s hereditary optic neuropathy by impairing complex I repair

Christiane Michaela Neuhofer ^1;2;3;4, Theresa Brunet^1;4;5, Robert Kopajtich^1;2, Riccardo Berutti^1;2;4;6, Thomas Meitinger^1;4, Hans A. Mayr^3;7, Ilka Wittig⁸, Thomas Klopstock⁹, Holger Prokisch^1;2;4

¹Institute of Human Genetics, Technical University of Munich, Munich, Germany; ²Institute of Neurogenomics, Helmholtz Zentrum München, Neuherberg, Germany; ³Institute of Human Genetics, Paracelsus Medical University, University Hospital Salzburg, Salzburg, Austria; ⁴Bavarian Genomes Network for Rare Disorders; ⁵Dr. v. Hauner Children’s Hospital, Department of Pediatric Neurology and Developmental Medicine, LMU- University of Munich, Munich, Germany; ⁶NEXT - NGS core facility, CRS4 - Center for Advanced Studies, Research and Development in Sardinia, Pula, Italy; ⁷University Children’s Hospital, Paracelsus Medical University, University Hospital Salzburg, Salzburg, Austria; ⁸Institute for Cardiovascular Physiology, Goethe-University Frankfurt am Main, Frankfurt am Main, Germany; ⁹Friedrich-Baur-Institute, Department of Neurology, LMU- University of Munich, Munich, Germany

Background/Objectives: Leber’s hereditary optic neuropathy (LHON) is the most common mitochondrial disorder. It is caused in >90% of cases by one of three mitochondrial point mutations in mitochondrial complex I (CI) subunits. Recently DNAJC30 has been identified as the first nuclear gene to cause LHON and implicated CI repair as underlying pathomechanism. We present genetic and functional findings in a patient clinically diagnosed with LHON to introduce NDUFA7 as a novel nuclear gene causing LHON. The patient showed a classic LHON phenotype with subacute, painless central vision loss first of the right eye, within one week bilaterally, at age 17.

Methods: We performed WGS after targeted LHON diagnostics was unable to reveal the patient`s disease cause. Functional work-up included transcriptome, proteome and complexome turnover analysis.

Results: WGS revealed a homozygous 1bp duplication at a splice donor site of NDUFA7, which encodes a CI N-module subunit that directly interacts with DNAJC30 to enable rapid N-module turnover. All four unaffected siblings of the patient were not homozygous for this NDUFA7 variant. Transcriptome analysis confirmed a 1bp frameshift at the transcript level and proteomics showed a significant reduction of NDUFA7 protein levels in fibroblasts. Dynamic complexome analysis of CI showed a reduction of three N-module subunits and DNAJC30 interactors, NDUFA7, NDUFS6 and NDUFA12, which physiologically show a high turnover rate.

Conclusion: We report NDUFA7 as a novel disease gene, extend the genetic landscape of complex I defects leading to LHON and emphasize the importance of the N-module repair for LHON disease.

Grants: GENOMIT, Bavarian Genomes

Conflict of Interest: Christiane Michaela Neuhofer: None declared, Theresa Brunet: None declared, Robert Kopajtich: None declared, Riccardo Berutti: None declared, Thomas Meitinger PI of Bavarian Genomes that in part funded this project, Hans A. Mayr: None declared, Ilka Wittig: None declared, Thomas Klopstock PI in GENOMIT which partly funded this project, Holger Prokisch PI and coordinator of GENOMIT, which partly funded this project

C14.6 Biallelic variants in GPN2, an essential RNA polymerase assembly factor, result in pleiotropic Perrault syndrome phenotypes

Thomas Smith ^1;2, Leigh Demain^1;2, Rabia Faridi³, Huw Thomas^1;2, Alessandro Rea^1;2, Arshia Maqbool⁴, Isabelle Schrauwen⁵, Khurram Liaqat^5;6, Zubair Ahmed^7;8, Sondhya Ghedia⁹, Andrew Green^10;11, Ruth Sheffer¹², Hagar Mor-Shaked¹², Mathilda Wilding⁹, Robin Hay⁹, Sayaka Inagaki³, Saima Riazuddin^7;8, Langping He^13;14, Glenda Beaman^1;2, Wasim Ahmad¹⁵, Suzanne Leal^5;16, Sheikh Riazuddin⁴, Robert Taylor¹⁴, Thomas Friedman³, Raymond O’Keefe^1;2, William Newman^1;2

¹University of Manchester, School of Biological Sciences, Manchester, United Kingdom; ²University of Manchester, Manchester Centre for Genomic Medicine; ³National Institutes of Health, Laboratory of Molecular Genetics, National Institute on Deafness and Other Communication Disorders; ⁴University of Health Sciences, Jinnah Burn and Reconstructive Surgery Centre, Allama Iqbal Medical Research, Lahore, Pakistan; ⁵Columbia University Medical Center, Center for Statistical Genetics, Gertrude H. Sergievsky Center, and the Department of Neurology; ⁶Quaid-i-Azam University, Department of Biotechnology, Faculty of Biological Sciences; ⁷University of Maryland School of Medicine, Department of Otorhinolaryngology Head and Neck Surgery; ⁸University of Maryland, Department of Molecular Biology and Biochemistry, School of Medicine; ⁹Royal North Shore Hospital, Department of Clinical Genetics, St Leonards, Australia; ¹⁰Children’s Health Ireland at Crumlin, Department of Clinical Genetics; ¹¹University College Dublin, Ireland School of Medicine and Medical Science; ¹²Hadassah-Hebrew University Medical Center, Department of Genetics; ¹³Newcastle upon Tyne Hospitals NHS Foundation Trust, NHS Highly Specialised Rare Mitochondrial Disorders Service; ¹⁴Newcastle University, Wellcome Centre for Mitochondrial Research; ¹⁵Quaid-i-Azam University, Department of Biochemistry; ¹⁶Columbia University Medical Center, Taub Institute for Alzheimer’s Disease and the Aging Brain, New York, United States

Background/Objectives: Perrault syndrome is a rare, autosomal recessive disorder characterised by sensorineural hearing loss and primary ovarian insufficiency. To date, biallelic variants in nine genes affecting mitochondrial and peroxisomal function have been associated with this condition, with presentations encompassing leukodystrophy, intellectual disability and developmental delay. However, in many individuals with Perrault syndrome the genetic cause remains unknown.

Methods: Exome sequencing was undertaken on affected individuals with Perrault syndrome alongside unaffected family members to identify disease-associated variants. Biallelic variants in GPN2, a GTPase involved in RNA polymerase biogenesis and assembly, were identified in four unrelated families. Individuals from two families presented with Perrault syndrome and cerebellar ataxia with homozygosity for p.Asn222Asp (RefSeq 54707). A 662kb common haplotype at the GPN2 locus was defined. A further family of ten affected individuals with Perrault syndrome homozygous for p.His121Gln was identified. Through reverse phenotyping, a child homozygous for p.Ser276Cys, presented with progressive ataxia, intellectual disability and seizures, but no hearing loss.

Results: Previously known to localize to the nucleus, we demonstrate that GPN2 also localizes to the mitochondria using immunofluorescence in patient fibroblasts. Respiratory chain complex activities were not altered in p.Asn222Asp fibroblasts, and GPN2 mRNA and protein levels were unchanged via qPCR and Western blotting, suggesting a subtle molecular phenotype. However, disease-associated variants may lead to altered GTPase activity and protein stability.

Conclusion: We present compelling genetic evidence that biallelic variants in GPN2 result in a multisystem phenotype, with preliminary data suggesting that GPN2 also has a previously unrecognized role in mitochondrial function.

Grants: MRC(MR/W019027/1);RNID&MCF(S60_Newman);AMR(GN2494); NIHR_MBRC(NIHR203308);WCMR_RWT(203105/Z/16/Z);NIDCD_NIH_TBF(DC000039).

Conflict of Interest: None declared

C15.1 Insights from 1.2M structural variants across global populations in the Genome Aggregation Database (gnomAD)

Xuefang Zhao ^1;2;3, Ryan Collins⁴, Mark Walker^2;3, Alex Yenkin¹, Nehir Kurtas^1;2;3, Arthur Lee^1;2;3, Lily Wang¹, Daniel Ben-Isvy¹, Jack Fu^1;2;3, Cal Liao^3;5, Philip Darnowsky³, Katherine Chao³, Stacey Gabriel^3;5, Eric Banks³, Anne O’Donnell-Luria^2;3, Daniel MacArthur³, Heidi Rehm^2;3, Benjamin Neale^3;5, Mark Daly^3;5, Kaitlin Samocha^2;3, Konrad Karczewski^3;5, Dadi Gao^1;6, Harrison Brand^1;2;3, Michael Talkowski^1;2;3

¹Harvard Medical School, Neurology, Boston, United States; ²Massachusetts General Hospital, Center for Genomic Medicine, Boston, United States; ³Broad Institute, Medical and Population Genetics, Boston, United States; ⁴Dana-Farber Cancer Institute, Division of Medical Oncolog, Boston, United States; ⁵Massachusetts General Hospital, The Analytic and Translational Genetics Unit, Boston, United States; ⁶Massachusetts General Hospital, Neurology, Boston, United States

Consortium: gnomAD consortium

Background/Objectives: Structural variantion (SVs) contribute to the diversity of human genomes and myriad human trait associations.

Methods: As part of the Genome Aggregation Database (gnomAD), we present the completion of an SV atlas in gnomADv4 that encompasses 1,199,117 unique SVs genotyped across 63,046 unrelated individuals. Importantly, over half of these genomes were inferred non-European, enabling unique insights into genetic diversity across global ancestries.

Results: The average genome harbored 11,844 SVs discovered by short-read sequencing, with long-read sequences from 15 matched genomes supporting ~97% of SVs discovered outside of highly repetitive genomic regions. Beyond canonical SVs, we resolved ~45 complex SVs in each genome (most involving inversions) and found a highly complex yet polymorphic variant involving 49 breakpoints across 7 chromosomes in 3 individuals. From these data we estimate that SVs accessible to short-read sequencing cause, on average, loss-of-function (LoF) of 188 genes and copy gain (CG) of 34 genes per genome. The per-gene frequency of LoF and CG in SVs are significantly correlated with metrics of constraint against LoF for short variants, including the surprising finding of selection against SVs that only duplicate transcription start sites or UTRs despite the absence of direct disruption to the genic open reading frames. Further, we find that metrics measuring an excess of singleton SVs over neutral selection show positive correlation with expression levels (r = 0.73) of genes they interrupt.

Conclusion: These preliminary findings underscore the substantial impact of SVs on genome structure and function and the value of publicly accessible reference resources such as gnomAD (https://gnomad.broadinstitute.org/)

Grants: R01MH115957

Conflict of Interest: None declared

C15.2 Whole genome sequencing and analysis of 5750 Indians: Phase-1 insights from the Genome India project. (virtual)

Analabha Basu ¹, Bratati Kahali², Mohd Faruq³, Divya Tej Sowpati⁴, Kumarasamy Thangaraj⁴, Yadati Narahari⁵

¹iBRIC-National Institute of BioMedical Genomics, Statistical and Computational Genomics, Kalyani, India; ²Centre for Brain Research - Indian Institute of Science, Bangalore, India; ³CSIR-Institute of Genomics and Integrative Biology, New Delhi, India; ⁴CSIR-Centre for Cellular and Molecular Biology, Hyderabad, India; ⁵Centre for Brain Research - Indian Institute of Science Campus, Bengaluru, India

Consortium: GenomeIndia Project

Background/Objectives: India, the most populous country in the world, (1.43billion) is grossly under-represented in global genomic databases and biobanks. India harbours genetic diversity next only to Africa with more than 5000 prominent ethnolinguistic groups. The GenomeIndia Project is initiated by the Department of Biotechnology, Government of India, to catalogue the genomic diversity of India and build a roadmap for genomic-medicine, pharmacogenetics and public health benefits for the subcontinent and the world at large.

Methods: WGS of 10,000 individuals from 99 populations spanning geography, ethno-linguistic identity and social sub-structuring. In phase-1, a preliminary analysis of 5750 individuals from 65 populations has been completed.

Results:

1. 1)

   Identification of 135.48million SNVs (117.8million biallelic SNVs, 10.6million biallelic INDELs, 6.97million multiallelic).

2. 2)

   More than 4.7million common variant candidates for association study design.

3. 3)

   27 million variants, which are overall rare (<1%overall) but common in atleast one population(>5%). About 7million of 27million are novel. This potentially can reclassify the definition of 213 InterVar variants which were identified by ACMG as deleterious.

4. 4)

   Identified 118 Level1A and 1B variants, classified by PharmGKB as adversely impacting drug response, of which 38 are in high frequency in different populations.

5. 5)

   More than two-third of the populations analyzed in phase-1, have homozygosity and shared identical-by-descent segments significantly more numerous and extensive compared to the Finnish.

Conclusion:

1. 1)

   An Unprecedented diversity, unraveling new understanding of population history

2. 2)

   Possible reclassification of hundreds of variants identified as deleterious.

3. 3)

   Enormous improvement in imputing power of genotypes using the GenomeIndia Data compared to what was previously available.

Grants: BT/GenomeIndia/2018

Conflict of Interest: None declared

C15.3 A PCA and F-statistics based framework to infer human population history

Divyaratan Popli ¹, Benjamin Peter²

¹Max Planck Institute for Evolutionary Anthropology, Evolutionary Genetics, Leipzig, Germany; ²Max Planck Institute for Evolutionary Anthropology, Leipzig, Germany

Background/Objectives: F-statistics and Principal Component Analysis (PCA) are among the most widely used methods to study human population history. PCA is a dimensionality reduction method to visualize the genetic distance of individuals, and is generally not used for statistical testing. In contrast, F-statistics measure the genetic drift between two, three, and four populations, and can be used to test hypotheses about admixture and structure. Unlike PCA, F-statistics require assignment of individuals to discrete populations.

Methods: In this study we provide a statistical framework to estimate F-statistics from PCA. Our approach is based on geometrical interpretation of F-statistics (Oteo-García 2021 and Peter, 2022). The accuracy of estimated F-statistic depends on the PCA method used. Therefore, using simulations as well as Neandertal genomes, we evaluate estimation of F-statistics from different PCA methods, and compare them to a popular software ADMIXTOOLS 2.

Results: We find that PCA methods like probabilistic PCA (PPCA) and Latent Subspace Estimation (LSE), which model sampling bias, are more accurate compared to classical PCA. An advantage of our framework is that we can reliably estimate F-statistics with low-coverage genomes with missing genotypes, which are common in ancient DNA studies. Additionally, calculating genetic distances on PCA allows us to use the entire dataset without the need to group individuals.

Conclusion: We provide a statistical framework to combine PCA and F-statistics, and show that this framework addresses limitations of both PCA and F-statistics, when applied independently.

Grants: European Research Council (Grant No.: 694707) to Svante Pääbo, European Research Council (Grant No.: 101042421) to Benjamin M. Peter

Conflict of Interest: None declared

C15.4 Recombination enables higher numbers of recessive genes, contributing to the emergence of sexual mating in complex organisms

Luis Aniello La Rocca ¹, Peter Krawitz², Anton Bovier¹, Konrad Gerischer², Hannah Klinkhammer², Leonie Fohler²

¹Institute for Applied Mathematics, University of Bonn, Bonn, Germany; ²Institute for Genomic Statistics and Bioinformatics, University of Bonn, Bonn, Germany

Background/Objectives: The drift-barrier hypothesis states that random genetic drift constraints the refinement of a phenotype under natural selection. The influence of effective population size and the genome-wide deleterious mutation rate were studied theoretically and an inverse relationship between mutation rate and genome size has been observed for many species. However, the effect of the recessive gene count, an important feature of the genomic architecture, is unknown.

Methods: In a Wright-Fisher model we studied the spread of N completely recessive, lethal diseases. Diploid individuals are represented with a 2xN matrix denoting wild type (0) and mutated alleles (1). Analytic results for specific cases were complemented by simulations across a broad parameter regime for number of genes, mutation- and recombination rates.

Results: Simulations revealed a generational transition to higher mutation burden and prevalence, linked to the extinction of least-loaded haplotype classes. Metastability persists over 100,000 generations. Recombination proves pivotal, preventing population collapse and maintaining mutation burden below 10. The drift-barrier-hypothesis is confirmed by a high mutation burden resulting population collapse.

Conclusion: This study advances our understanding of gene pool stability, and particularly the role of number of recessive disorders. Simulations showed the emergence of mutually exclusive haplotypes with lethal equivalents above 0.02 per generation. Recombination’s essential role in curbing mutation burden and stabilizing the gene pool is demonstrated. Insights into Muller’s ratchet dynamics are provided, stressing the significance of recessive gene count in genomic architecture for population stability.

Grants: DFG - German Excellence Strategy GZ 2047/1, Project-ID 390685813; GZ 2151, Project-ID 390873048

Conflict of Interest: None declared

C15.5 Estimating ancestry profiles by nearest neighbor regression in principal component space

Juha Riikonen ¹, Sini Kerminen¹, Aki Havulinna^1;2, Matti Pirinen^1;3;4

¹Institute for Molecular Medicine Finland, Helsinki, Finland; ²Finnish Institute for Health and Welfare, Helsinki, Finland; ³University of Helsinki, Department of Public Health, Helsinki, Finland; ⁴University of Helsinki, Department of Mathematics and Statistics, Helsinki, Finland

Background: Haplotype-based tools are ideal for ancestry inference but computationally impractical for biobank-scale analyses. We describe an efficient alternative using principal components (PCs) for individual-level ancestry estimation. Contrary to existing tools like Rye¹ that use reference samples with discrete population assignment, our approach enables using reference samples with ancestry proportions to multiple populations. We had data on 18,125 samples from the biobank of the Finnish Institute for Health and Welfare (THL) with PC-coordinates and ancestry profiles estimated using the haplotype-based tool SOURCEFIND², based on 1,124 reference samples allocated to 10 Finnish source populations².

Methods: For a test set of 1,000 non-reference samples we computed their pairwise eigenvalue-weighted Euclidean distances to the reference samples in the 10-dimensional PC space. Using k-nearest neighbor (kNN) regression over the reference sample profiles, we estimated the test samples’ ancestry proportions to each source population.

Results: Estimating over k = 25 nearest reference samples, the mean total variation distance (TVD) between the test samples’ SOURCEFIND and kNN-profiles was 0.173. With the same reference samples, Rye showed a slightly larger mean TVD of 0.224. Using kNN-regression on all samples instead of the reference samples alone, reduced the TVD to 0.113.

Conclusion: As measured by TVD, our kNN-approach outperformed Rye. The method relies only on PCs and enables inference based on reference samples with continuous ancestry profiles, making it a promising tool for ancestry estimation in large-scale genomic studies.

Grants: Juselius foundation

1. Kerminen, PLoS Genetics 2021

2. Chacón-Duque, Nature Communications 2018

3. Conley, Nucleic Acids Research 2023

Conflict of Interest: Juha Riikonen: None declared, Sini Kerminen Nightingale Health Plc, stock options, Nightingale Health Plc, Full-time, Aki Havulinna: None declared, Matti Pirinen: None declared

C15.6 Heterozygote selection reduces recessive intellectual disability

Hila Fridman ^1;2;3, Gelana Khazeeva¹, Ephrat Levy-Lahad^2;3, Christian Gilissen¹, Han Brunner^4;5

¹Radboud University Medical Center, Department of Human Genetics, Radboud Institute for Molecular Life Sciences, Nijmegen, Netherlands; ²Shaare Zedek Medical Center, Medical Genetics Institute, Jerusalem, Israel; ³The Hebrew University of Jerusalem, Faculty of Medicine, Jerusalem, Israel; ⁴Radboud University Medical Centre, 4Department of Human Genetics and Donders Center for Neuroscience, Nijmegen, Netherlands; ⁵Maastricht University Medical Center, 5Department of Clinical Genetics, GROW-School for Oncology and Developmental Biology and MHENS School for Mental Health and Neuroscience, Maastricht, Netherlands

Background/Objectives: For recessive conditions, there is strong selection against biallelic pathogenic variants (PVs) resulting in severe disease. Here we investigated whether heterozygosity for PVs in recessive genes has phenotypic consequences that may also be subject to selection.

Methods: We used exome data from 406,194 unrelated European individuals of the UK Biobank, and identified 55,746 different PVs in a curated set of 1929 recessive disease genes. We used three different constraint metrics to perform association analysis of the impact of heterozygosity for PVs in these genes on reproductive success and cognition.

Results: We find that heterozygous carriers in recessive constrained genes are more likely to be childless (OR:1.89, p:1.8E-3 Bonferroni corrected), especially heterozygotes for PVs in genes underlying intellectual disability (ID; OR:2.18, p:4.4E-2). Heterozygous variants in recessive ID genes also affect cognition (OR:0.29, p:1.0E-4). This effect is not seen for genes associated with other disorders. This negative selection is reflected in a threefold reduction in the frequency of heterozygosity for recessive ID in the population relative to other recessive disorders.

Conclusion: We find evidence for fitness effects in heterozygous carriers for recessive genes, especially for variants in constrained genes that underlie intellectual disability (ID), and show that such variants affect cognition. This negative selection is reflected in a drastically altered genetic landscape.

Grants: VIDI grant from the Dutch Research Council (917-17-353 to C.G.) and an AI for Health PhD grant from Radboudumc.

Conflict of Interest: None declared

C16.1 Investigating germline instability in triplet repeat disorders in 11,415 family trios

Anupriya Dalmia ^1;2, Valentina Galassi Deforie³, Kristina Ibañez², Chris Clarkson², Davina Hensman Moss⁴, Peter Holmans⁵, Arianna Tucci²

¹UCL, UK DRI, London, United Kingdom; ²QMUL, Clinical Pharmacology, London; ³UCL, Neuromuscular Diseases; ⁴St George’s University of London; ⁵Cardiff University, Division of Psychological Medicine and Clinical Neurosciences

Consortium: 100,000 Genomes Project

Background/Objectives: Polyglutamine diseases stem from abnormal expansions of repetitive CAG trinucleotide repeats, notably in Huntington disease patients where one HTT gene copy exceeds 36 repeats. Intermediate-sized repeats (27-35) occur in ~6% of the population, often remaining asymptomatic but prone to expansion. We investigate intergenerational instability in normal and intermediate allele carriers across various polyglutamine disease genes.

Methods: In this study, we employed Expansion Hunter v3 to genotype the CAG tract in ATN1, ATXN1, ATXN2, ATXN3, ATXN7, CACNA1A, and HTT, utilizing whole-genome sequencing data from 11,415 family trios in the 100,000 Genomes Project. We, then, ascertain parental origin through haplotyping. We used regression-based analyses to identify key factors influencing germline instability.

Results: Consistent with existing literature, 5.9% of our parent population carries an intermediate allele for HTT. The instability percentage ranges from 0.2% to 2.5% across the seven analyzed genes. Although the prevalence of intermediate allele carriers and the clinically defined intermediate range vary among the genes, our analysis indicates that allele size is the primary predictor of instability. We observed a consistent pattern of increasing germline instability across six different genes (ATN1, ATXN1, ATXN2, ATXN3, ATXN7, and HTT), with instability becoming exponential at a common cut-off point.

Conclusion: Our research breaks new ground by exploring germline repeat instability in a largely neurologically healthy population. The consistent rise in instability at a shared threshold across multiple genes carries profound implications in the genetic counselling setting.

Grants: Barts Charity

Conflict of Interest: None declared

C16.2 Biallelic variants in the ribosomal protein S6 kinase C1 (RPS6KC1) cause a complex neurodevelopmental disorder

Laura Planas-Serra ^1;2, Mar Rodríguez^1;2, Agustí Rodriguez-Palmero^1;2;3, Valentina Vélez-Santamaria^2;4;5, Agatha Schlüter^1;2, Edgard Verdura^1;2, Gorka Gereñu^6;7;8;9, Andrés Jiménez-Zúñiga⁶, Adolfo López de Munaín^6;8;10, Alejandro Iñañez^11;12, Josefina Casas^11;12, Joan Josep Bech¹³, Carolina De La Torre¹³, Maria Iascone¹⁴, Romano Tenconi¹⁵, Reza Maroofian¹⁶, Ehsan Ghayoor Karimiani¹⁷, Suzanne Leal^18;19, Isabelle Schrauwen¹⁸, Susanna Öhman²⁰, Irma Järvelä²¹, Johanna Granvik²², Karit Reinson^23;24, Elvira Kurvinen²³, Katrin Ounap^23;24, Tuğba Kalaycı²⁵, Shahrashoub Sharifi²⁵, Carlos Casasnovas^2;4;26, Eric Anderson²⁷, Udai Pandey²⁷, Aurora Pujol^2;4;28

¹Bellvitge Biomedical Research Institute (IDIBELL), Neurometabolic Diseases Laboratory, L’Hospitalet de Llobregat, Spain; ²Instituto de Salud Carlos III, Centre for Biomedical Research on Rare Diseases (CIBERER), Madrid, Spain; ³Hospital Universitari Germans Trias i Pujol, Pediatric Neurology Unit, Barcelona, Spain; ⁴Bellvitge Biomedical Research Institute (IDIBELL), Neurometabolic Diseases Laboratory, L’Hospitalet de Llobregat, Spain; ⁵Hospital Universitari de Bellvitge, Neuromuscular Unit, L’Hospitalet de Llobregat, Spain; ⁶Biodonostia Health Research Institute, Neuromuscular Diseases Group, Donostia, Spain; ⁷Instituto de Salud Carlos III, Center on Biomedical Research Networking Neurodegenerative Diseases (CIBERNED), Madrid, Spain; ⁸Basque Foundation for Science, Ikerbasque, Bilbao, Spain; ⁹University of the Basque Country UPV-EHU, Department of Physiology, Leioa, Spain; ¹⁰University of the Basque Country UPV-EHU, Department of Neurosciences, Leioa, Spain; ¹¹Institute of Advanced Chemical of Catalonia of CSIC (IQAC-CSIC), RUBAM, Department of Biological Chemistry, Madrid, Spain; ¹²Instituto de Salud Carlos III, Liver and Digestive Diseases Networking Biomedical Research Centre (CIBEREHD), Madrid, Spain; ¹³Josep Carreras Leukaemia Research Institute Barcelona, Proteomics Unit, Barcelona, Spain; ¹⁴ASST Papa Giovanni XXIII, Laboratory of Genetic Medicine, Bergamo, Italy; ¹⁵Università di Padova, Department of Pediatrics, Genetica Clinica, Padova, Italy; ¹⁶University College London, Department of Neuromuscular Diseases UCL Queen Square Institute of Neurology, London, United Kingdom; ¹⁷University of London, Molecular and Clinical Sciences Institute, St. George’s, London, United Kingdom; ¹⁸Columbia University Medical Center, Center for Statistical Genetics, Sergievsky Center, Department of Neurology, New York, United States; ¹⁹Columbia University Medical Center, Taub Institute for Alzheimer’s Disease and the Aging Brain, New York, United States; ²⁰Kårkulla Joint Authority, Kirjala, Finland; ²¹University of Helsinki, Department of Medical Genetics, Helsinki, Finland; ²²Wellbeing services county of Central Ostrobothnia, Kokkola, Finland; ²³Tartu University Hospital, Department of Clinical Genetics, Genetic and Personalized Medicine Clinic, Tartu, Estonia; ²⁴University of Tartu, Institute of Clinical Medicine, Tartu, Estonia; ²⁵Istanbul University, Department of Medical Genetics, Istanbul Medical Faculty, Istanbul, Türkyie; ²⁶Hospital Universitari de Bellvitge, Neuromuscular Unit, L’H; ²⁷University of Pittsburgh Medical Center, Department of Pediatrics, Children’s Hospital of Pittsburgh, Pittsburgh, United States; ²⁸Catalan Institution of Research and Advanced Studies (ICREA), Barcelona, Spain

Background/Objectives: The ribosomal protein S6 kinase family members play essential biological functions in disease, from cancer to intellectual disability. Little is known about RPS6KC1, aside from its lack of phosphorylation capacity and its roles in sphingosine-1-phosphate signaling and peroxiredoxin-3 transport to mitochondria.

Methods: We used whole-exome and -genome sequencing of undiagnosed patients with spastic paraplegia in a Spanish cohort and additional patients through GeneMatcher. Functional studies included proteomics, phosphoproteomics, and lipidomics in patients’ PBMCs and plasma, along with disease modeling in HAP1 cells and Drosophila melanogaster.

Results: We identified biallelic RPS6KC1 loss-of-function variants in 9 individuals from 6 independent families. The broad spectrum of phenotypic manifestations included neurodevelopmental delay, hypotonia, spastic paraplegia, and dysmorphic features, overlapping with those of Coffin-Lowry syndrome caused by RPS6KA3 loss.

Functional studies on patients’ samples indicated decreased expression and phosphorylation of RPS6, with subsequent impact on ribosomal protein synthesis, and a decrease in known interactors PRDX3 and SPHK1, along with a marked repression of mTOR and PI3K pathways. We detected an overall dysregulation of the lipidome, particularly phosphoinositides, and the enzymes controlling lipid synthesis. Functional studies in HAP1 RPS6KC1 knockdown cells suggested that RPS6KC1 may regulate PRDX3 and SPHK1 activities by facilitating their anchor at endosomes. In Drosophila melanogaster, RPS6KC1 knockdown resulted in motor dysfunction, defects at neuromuscular junctions, reduced lifespan, and decreased mTOR activity. Overexpression of mTOR improved motor function and lifespan in this model.

Conclusion: RPS6KC1 exerts crucial roles in neurodevelopment by controlling ribosomal protein synthesis, lipid signaling, and the mTOR pathway.

Conflict of Interest: None declared

C16.3 Deciphering the Genetic Landscape of Unsolved Congenital Myopathies: A Multinational Research Approach

Mridul Johari ¹, Chiara Folland¹, Marion ONNEE^2;3, Machteld Oud⁴, Ana Töpf⁵, Yoshihiko Saito⁶, Anita Cairns⁷, Francesca Clementina Radio⁸, Adele D Amico⁹, Thatjana Gardeitchik¹⁰, Isabella Herman¹¹, Sergei A. Kurbatov¹², Pilar Marti¹³, Alayne Meyer¹⁴, Stefan Nicolau¹⁵, Kiran Polavarapu¹⁶, Thomas Robertson¹⁷, Juan Vilchez¹³, Gordana Kovacevic>¹⁸, Alan Beggs¹⁹, Daniel Calame¹¹, Michael C. Fahey²⁰, Nigel Laing¹, Hanns Lochmuller¹⁶, James Lupski¹¹, Edoardo Malfatti³, Ichizo Nishino⁶, Atchayaram Nalini²¹, Vedrana Milic¹⁸, Marco Tartaglia²², Volker Straub⁵, Bjarne Udd²³, Nicol Voermans²⁴, Gina Ravenscroft¹

¹Harry Perkins Institute of Medical Research, Centre for Medical Research, University of Western Australia, Nedlands, WA, Australia, Perth, Australia; ²Institut Mondor de Recherche Biomédicale, Université Paris Est Créteil, INSERM U955, Créteil, France; ³Institut Mondor de Recherche Biomédicale, Université Paris Est Créteil, INSERM U955, Créteil, France, Paris, France; ⁴Department of Human Genetics, Donders Institute for Brain, Cognition and Behaviour, Radboud University Medical Center, Nijmegen, Netherlands; ⁵John Walton Muscular Dystrophy Research Centre, Translational and Clinical Research Institute, Newcastle University and Newcastle Hospitals NHS Foundation Trust, Newcastle upon Tyne, United Kingdom; ⁶Department of Neuromuscular Research, National Institute of Neuroscience, National Center of Neurology and Psychiatry, Tokyo, Japan; ⁷Neurosciences Department, Queensland Children’s Hospital, Brisbane (Meanjin) Queensland, Brisbane, Australia; ⁸Genetics and Rare Diseases Research Division, Ospedale Pediatrico Bambino Gesù IRCCS, Rome, Italy; ⁹Unit of Muscular and Neurodegenerative Disorders, Genetics and Rare Diseases Research Division, Department of Neurosciences, Bambino Gesù Children’s Hospital, IRCCS, Rome, Italy; ¹⁰Department of Human Genetics, Radboud university medical center, Nijmegen, Netherlands; ¹¹Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, United States; ¹²Research Institute of Experimental Biology and Medicine, Voronezh N. N. Burdenko State Medical University, Voronezh, Russian Federation; ¹³Neuromuscular Unit, Department of Neurology, Hospital Universitari i Politècnic La Fe, Valencia, Spain; ¹⁴Division of Genetic and Genomic Medicine, Nationwide Children’s Hospital, Columbus, United States; ¹⁵Center for Gene Therapy, The Abigail Wexner Research Institute at Nationwide Children’s Hospital, Columbus, United States; ¹⁶Children’s Hospital of Eastern Ontario Research Institute; Division of Neurology, Department of Medicine, The Ottawa Hospital; and Brain and Mind Research Institute, University of Ottawa, Ottawa, Canada; ¹⁷Anatomical Pathology, Queensland Pathology, Brisbane, Queensland, Australia, Brisbane, Australia; ¹⁸Faculty of Medicine, University of Belgrade, Belgrade, Serbia; ¹⁹The Manton Center for Orphan Disease Research, Division of Genetics and Genomics, Boston Children’s Hospital, Harvard Medical School, Boston, United States; ²⁰Department of Paediatrics Monash Children’s Hospital, Melbourne, Australia; ²¹Department of Neurology, National Institute of Mental Health and Neuro Sciences (NIMHANS), Bengaluru, India; ²²Molecular Genetics and Functional Genomics, Ospedale Pediatrico Bambino Gesù, IRCCS, Rome, Italy; ²³Folkhälsan Research Center, Department of Medical and Clinical Genetics, Medicum, University of Helsinki, Helsinki, Finland; ²⁴Department of Neurology, Donders Institute for Brain, Cognition and Behaviour, Radboud University Medical Center, Nijmegen, Netherlands

Consortium: Congenital myopathy consortium

Background/Objectives: Congenital myopathies (CM) are a diverse group of inherited neuromuscular disorders most frequently characterized by muscle weakness and structural abnormalities on muscle biopsy. A significant proportion of CM cases remain genetically unsolved at a single-centre level, underscoring the necessity for comprehensive research approaches that leverage international collaborations.

Methods: We analysed exome sequencing and short-read genome sequencing data from unsolved CM cases across multiple centres in Europe, the USA, Asia, and Australia. This analysis was complemented by dedicated expert curation of clinical and histopathological data, along with curation of de novo variants and Variants of Uncertain Significance (VUS). We further explored genetic pathomechanisms through in vitro studies, and bulk and single nucleus RNA sequencing (snRNAseq).

Results: We identified three novel forms of CM. (1) Loss-of-function variants in JPH1 were identified in three unrelated patients with striking bulbar and ocular involvement. (2) Biallelic variants in ASCC3 were identified in eight patients manifesting with a range of overlapping features including early-onset core myopathy, intellectual disability, and respiratory involvement. snRNAseq analysis highlighted the role of ASC-1 complex in translational homeostasis. (3) Our highly collaborative research revealed 12 unique missense variants in TUBA4A in 23 individuals, indicating a novel class of myopathy with de novo, autosomal dominant, and autosomal recessive inheritance. Previously, TUBA4A variants had been associated with FTD, ALS and ataxia. Thus, our study extends the phenotypic landscape of TUBA4A-related disease.

Conclusion: These findings emphasise the critical role of international collaborations in enhancing genetic interpretation and identifying novel disease genes and genotype-phenotype relationships in neuromuscular disease research.

Grants: AFM-Téléthon(24438)_Australian-NHMRC-Ideas-Grant-(APP2002640)

Conflict of Interest: None declared

C16.4 Allele-specific epigenome therapy in COL6-RD patient-derived primary fibroblasts

Franziska Haarich ^1;2;3, Malte Spielmann^2;4;5, Jeanette Erdmann^1;2;3

¹Institute for Cardiogenetics, University of Lübeck, Lübeck, Germany; ²DZHK (German Research Centre for Cardiovascular Reasearch), partner site Hamburg/Lübeck/Kiel, Lübeck, Germany; ³University Heart Center Lübeck, Lübeck, Germany; ⁴Institute of Human Genetics, Universitätsklinikum Schleswig-Holstein, University of Lübeck and University of Kiel, Lübeck and Kiel, Germany; ⁵Human Molecular Genomics Group, Max Planck Institute for Molecular Genetics, Berlin, Germany

Background/Objectives: Collagen VI-related dystrophies (COL6-RD) are rare disorders caused by mutations in COL6A1, COL6A2 and COL6A3. Common symptoms include muscle weakness, contractures, hyperlaxity, skin abnormalities and respiratory insufficiency. There is no treatment available. Since many disease-causing variants are almost private, a mutation-independent therapy approach would be beneficial. In this project we use allele-specific epigenome editing as a potential therapy to knockdown dominant-negative mutation-carrying alleles to enable the healthy allele to function properly without targeting the mutation itself.

Methods: Variants in the promotor region of COL6A2 were identified and phased by long read sequencing with the disease-causing mutation for two unrelated COL6-RD patients harboring different dominant-negative Glycine substitutions. Allele-specific guideRNAs were designed targeting shared common variants. Cultured primary fibroblasts of these individuals were transfected with synthetic sgRNAs and in vitro transcribed mRNA coding for the epigenome editor CRISPRoff. Transfected cells were analyzed on gene expression and DNA methylation. With immunofluorescence staining the collagen VI secretion and network formation was examined.

Results: The expression of the mutated allele was significantly reduced and the DNA was allele-specifically methylated in the targeted region after the treatment with CRISPRoff and designed gRNAs. Successfully treated cells of both cell lines were enabled to secrete collagen VI into the extracellular matrix leading to a network formation.

Conclusion: Allele-specific and mutation-independent epigenome editing leads to a phenotypic improvement in cultured COL6-RD patient-derived primary fibroblasts. Targeting common variants instead of the mutation enables the treatment of very rare mutations.

Grants: The project is funded by CureCMD, MD Canada and DGM.

Conflict of Interest: None declared

C16.5 Following the clues in the transcripts: differential exon usage provides clinical and prognostic insights in the whole titinopathies spectrum

Maria Francesca Di Feo ^1;2;3, Ali Oghabian⁴, Ella Nippala⁴, Francesca Forzano^5;6, David Gómez-Andrés⁷, Angela F Brady⁸, Maria Iascone⁹, Anna Cereda¹⁰, Lidia Pezzani¹⁰, Daniel Natera-de Benito¹¹, Andrés Nascimento¹¹, Berta Estevez-Arias¹¹, Sergei A. Kurbatov¹², Tania Attie-Bitach¹³, Sheela Nampoothiri¹⁴, Erin Ryan¹⁵, Michelle Morrow¹⁵, Svetlana Gorokhova^16;17, Brigitte Chabrol¹⁸, Johanna Nokelainen¹⁹, Lauri Snellman²⁰, Francina Munell²¹, Peter Hackman⁴, Bjarne Udd^4;22, Marco Savarese^4;23

¹Folkhälsan Research Centre, Biomedicum, Helsinki, Finland; ²Giannina Gaslini Institute, Clinical Genetics and Genomics Unit, Genova, Italy; ³University of Genoa, Department of Neurology, Rehabilitation, Ophthalmology, Genetics, Maternal and Child Health (DINOGMI), Genova, Italy; ⁴Folkhälsan Research Centre, Helsinki, Finland; ⁵Guy’s and St Thomas’ NHS Foundation Trust, Dep. of Clinical Genetics, London, United Kingdom; ⁶King’s College London, London, United Kingdom; ⁷Vall d’Hebron University Hospital, Paediatric Neurology, Barcelona, Spain; ⁸North West Thames Regional Genetics Service, Northwick Park Hospital, Middlesex, United Kingdom; ⁹ASST Papa Giovanni XXIII Hospital, Medical Genetics Lab, Bergamo, Italy; ¹⁰ASST Papa Giovanni XXIII Hospital, Clinical Genetics Unit, Bergamo, Italy; ¹¹Sant Joan de Déu Barcelona Hospital, Neuromuscular Disorders Unit, Pediatric Neurology Department, Esplugues de Llobregat, Spain; ¹²Voronezh N.N.Burdenko State Medical University, Voronez, Russian Federation; ¹³Necker Hospital, Service de Médecine Génomique des Maladies Rares, Paris, France; ¹⁴Amrita Institute of Medical Sciences & Research Centre, Dep. of Pediatric Genetics, Kerala, India; ¹⁵GeneDx, Gaithersburg, United States; ¹⁶Aix-Marseille University, Medical Genetics, Marseille, France; ¹⁷Assistance Publique Hôpitaux de Marseille, Dép. de Génétique Médicale, Marseille, France; ¹⁸Assistance Publique Hôpitaux de Marseille, Centre de Référence des Maladies Neuromusculaires de l’enfant PACARARE, Marseille, France; ¹⁹Helsinki University Hospital, Transplantation Laboratory, Helsinki, Finland; ²⁰University of Helsinki, Coronary Disease Research Group, Helsinki, Finland; ²¹Hospital Materno-Infantil Vall d’Hebron, Unitat de Recerca Biomèdica, Barcelona, Spain; ²²Tampere University Hospital, Dep. of Musculoskeletal Diseases, Tampere, Finland; ²³University of Helsinki, Department of Medical Genetics, Helsinki, Finland

Background/Objectives: Titin (TTN), one of the largest genes with its 364 exons, generates numerous isoforms through potentially millions of splice pathways. Despite being well known as the molecular spring of the sarcomere, our understanding of titin’s differential expression has barely scratched the surface. Truncating variants in TTN (TTNtv) cause a wide range of neuromuscular disorders, posing challenges in determining clinical management and prognosis.

Methods: Short and long-read RNA sequencing was performed on a sample cohort including fetal and adult skeletal and heart muscles (over 60 samples) and utilizing ribosomal RNA-depletion for optimal TTN RNA-sequencing. An exon usage database (to be published for free use) was generated by a newly designed software (IntEREst). Then, molecular and clinical data of 12 patients carrying biallelic TTNtv were accrued from multiple hospitals.

Results: By looking at variants’ exon usage in our database, we were able to subdivide the patients in three categories that effectively matched their clinical trajectories: perinatal death, improvement with age, and aggravation with age. In some prenatal cases, exon usage allowed us to exclude titinopathy as a differential diagnosis. Interestingly, some novel fetal isoforms were discovered.

Conclusion: Interpreting TTN variants is crucial in the genomic era, as 1 in 100 individuals carries a TTNtv. Our study provides data for clinical and prognostic application and serves as a model for genetic disease personalized medicine. Almost 70% of human genes have at least 15 transcript isoforms, highlighting the need for a deeper consideration of this finely tuned evolutionary mechanism.

Grants: Academy of Finland, Samfundet Folkhälsan, Sydantutkimussaatio, Magnus Ehrnrooth Foundation

Conflict of Interest: None declared

C16.6 SENEGENE: insights into neuromuscular and neurogenetic diseases in Senegal, West Africa

Pedro M Rodriguez Cruz ^1;2;3, Rokhaya Diagne^3;4, Ainara Salazar Villacorta⁵, ALASSANE MAMADOU DIOP⁶, Alois Benoît Baba Diouf^7;8, Marie Emilie Yandé Ndong⁸, Henriette Senghor⁴, Seynabou Dieng⁸, Christelle Rose Oko⁴, MARIETOU TRAORE⁹, Maimouna Santos¹⁰, Khalifa Ababacar Mbaye¹¹, Lala Bouna Seck⁴, Mamadou Sy⁴, Maouly FALL⁶, Adjaratou Sow^4;12, Ngor Side DIAGNE⁴, Loreto Martorell Sampol¹³, Daniel Natera-de Benito¹⁴, Andrés Nascimento¹⁴, Raul Tonda^1;2, Steven Laurie^1;2, Leslie Matalonga¹, Henry Houlden⁵, Rokhaya Ndiaye¹⁵, Moustapha Ndiaye^3;4, Sergi Beltran^1;16, Amadou Gallo Diop^4;15

¹Centro Nacional de Análisis Genómico - CNAG, Barcelona, Spain; ²Universitat de Barcelona, Barcelona, Spain; ³Universite Cheikh Anta Diop (UCAD), Dakar, Senegal; ⁴Centre Hospitalier National Universitaire de Fann, Neurology, Dakar, Senegal; ⁵UCL Queen Square Institute of Neurology, London, United Kingdom; ⁶Centre Hospitalier National de Pikine, Neurology, Pikine, Senegal; ⁷Centre Talibou Dabo, Dakar, Senegal; ⁸Centre National Hospitalier d’Enfants Albert Royer, Paediatric Neurology, Dakar, Senegal; ⁹Centre Hospitalier Roi Baudouin de Guediawaye, Neurology, Senegal; ¹⁰Centre Hospitalier Régional de Saint-Louis, Neurology, Saint Louis, Senegal; ¹¹Centre Hospitalier Régional de Ziguinchor, Neurology, Ziguinchor, Senegal; ¹²Centre National Hospitalier des Enfants de Diamniadio, Neurology, Diamniadio, Senegal; ¹³Sant Joan de Déu Barcelona Hospital, Unidad de Genética Molecular, Esplugues de Llobregat, Spain; ¹⁴Sant Joan de Déu Barcelona Hospital, Paediatric Neurology, Esplugues de Llobregat, Spain; ¹⁵Universite Cheikh Anta Diop (UCAD), Human genetics, Dakar, Senegal; ¹⁶Universitat de Barcelona, Genètica, Microbiologia i Estadística,, Barcelona, Spain

Background/objective: Rare genetic diseases remain largely understudied in sub-Saharan Africa. Here, we develop a global partnership to jointly advance the field of neuromuscular and neurogenetic diseases in Senegal, West Africa.

Methods: Adults and children suspected of neuromuscular or neurogenetic diseases, and relatives, are recruited from the Neurology Department of CHNU de Fann (Dakar) and systematically phenotyped using HPO terms. Sequencing (Barcelona & London) follows a progressive approach with MLPA, WES & WGS. The bioinformatic analysis uses state-of-the-art alignment, variant calling and annotation for SNV & InDels, CNVs, STRs, SVs and dedicated mtDNA analysis. Annotated variants are evaluated based on phenotype relevance, segregation, frequency and functional consequences using the RD-Connect Genome-Phenome Analysis Platform (GPAP).

Results: 841 individuals (401F:440M) have been recruited to date from over 350 families: 55% neuromuscular disorders, 19% neurometabolic diseases, 14% complex motor disorders and ataxia, 4% epilepsy and 8% miscellaneous. A definitive genetic diagnosis has been established in ∼65% of the cases (inheritance: 60% recessive, 10% dominant, 30% X-linked) allowing improved medical care and genetic counselling. Unreported variants of known disease genes were found in 29 cases and 3 novel disease candidate genes have been identified.

Conclusion: Our study unveils the first description of neuromuscular and neurogenetic diseases in Senegal,increasing diversity and novel pathogenic variant knowledge in human genetics. Collaborative efforts like SENEGENE contribute to advance rare diseases in underrepresented populations, improving clinical management, and facilitating connexions of local rare disease communities to available resources and existing networks for the benefit of patients and families.

Grants: AFM-Telethon (25078), La Caixa Foundation (LCF/BQ/PI21/11830012).

Conflict of Interest: Pedro M Rodriguez Cruz La Caixa Foundation Junior Leader Fellowship (LCF/BQ/PI21/11830012). AFM Telethon Trampoline grant (25078), Rokhaya Diagne: None declared, Ainara Salazar Villacorta: None declared, ALASSANE MAMADOU DIOP: None declared, Alois Benoît Baba Diouf: None declared, Marie Emilie Yandé Ndong: None declared, Henriette Senghor: None declared, Seynabou Dieng: None declared, Christelle Rose Oko: None declared, MARIETOU TRAORE: None declared, Maimouna Santos: None declared, Khalifa Ababacar Mbaye: None declared, Lala Bouna Seck: None declared, Mamadou Sy: None declared, Maouly FALL: None declared, Adjaratou Sow: None declared, Ngor Side DIAGNE: None declared, Loreto Martorell Sampol: None declared, Daniel Natera-de Benito: None declared, Andrés Nascimento: None declared, Raul Tonda: None declared, Steven Laurie: None declared, Leslie Matalonga: None declared, Henry Houlden: None declared, Rokhaya Ndiaye: None declared, Moustapha Ndiaye: None declared, Sergi Beltran: None declared, Amadou Gallo Diop: None declared

C18.1 Metabolomic profiling of the entire UK Biobank enables interpretation of autosomal dominant disease mutations

Luke Jostins-Dean ¹, Kirsten Schut¹, Sini Kerminen¹, Peter Wurtz¹, Jeffrey Barrett¹

¹Nightingale Health Plc., Helsinki, Finland

Consortium: Nightingale Health Biobank Collaborative Group

Background/Objectives: We have generated metabolomic profiles of the entire UK Biobank. We sought to find new potential manifestations of rare diseases in human metabolism, and to investigate whether metabolic profiles can aid interpretation of candidate pathogenic variants.

Methods: We quantified 249 metabolite parameters from 490,830 UK Biobank participants and combined this with exome sequences for 200,643 participants. Based on high-confidence pathogenic mutations in ClinVar, we identified 44 autosomal dominant diseases with ≥5 carriers. We tested for association between each metabolic parameter and mutation status. For familial hypercholesterolemia, we also fitted a logistic predictive model to identify pathogenic mutations, using LDL levels, statin use and 17 principal components of the metabolic parameters.

Results: We identified ≥1 metabolite association at FDR < 0.05 for 8 diseases. These included expected associations, including dyslipidemia in familial hypercholesterolemia (FH) and hyperglycemia in monogenic diabetes, and previously unreported associations, including elevated serum acetone levels in neurofibromatosis and lowered levels of glycoprotein acetylation in Noonan syndrome. A model including all metabolites had a higher accuracy at identifying known pathogenic FH mutations (36/56 at FDR < 0.05), compared to LDL and statin use alone (30/56). Our model also identified 8 FH variants marked as uncertain significance in ClinVar with significant (FDR < 0.05) evidence of predicted pathogenicity.

Conclusion: Combined exome sequencing and large-scale metabolomic profiling in population biobanks can identify new mechanisms of metabolic dysregulation in genetic disease and improve our models for interpreting pathogenic mutations.

Grants: Nightingale Health.

Conflict of Interest: Luke Jostins-Dean Stock options in Nightingale Health Plc, Employee of Nightingale Health, Kirsten Schut Stock and stock options at Nightingale Health, Employee of Nightingale Health, Sini Kerminen Stock options for Nightingale Health, Employee at Nightingale Health, Peter Wurtz Stock and stock options at Nightingale Health, Employee of Nightingale Health, Jeffrey Barrett Stock and stock options at Nightingale Health, Employee of Nightingale Health

C18.2 Analysis of RNAseq from over 5000 individuals in the 100,000 Genomes Project identifies new potential diagnoses for patients with rare disease.

Diana Baralle ¹, Jenny Lord², Carolina Jaramillo Oquendo², Mark Ross^3;4, Terena James⁴, Lily Hoa⁵, Greg Elgar⁵

¹University of Southampton; ²University of Sheffield; ³; ⁴Illumina; ⁵Genomics England

Consortium: N McGiness, A Ho, C Odhams.

Genomics England

Diagnosis of rare disorders has been revolutionised by whole exome and genome sequencing (WES/WGS), but even with WGS, around half of patients’ disorders remain unsolved. Interpretation of non-coding variation, which can affect splicing and gene regulation, has lagged behind coding variation, and improvement in this area will improve diagnostic yields. Here we present analyses of whole blood-based transcriptome sequencing data from over 5,000 probands with rare disorders that underwent WGS in the 100,000 Genomes Project, but did not receive a molecular diagnosis.

We used multiple tools to identify candidate expression (OUTRIDER via DROP) and splicing (LeafCutterMD, FRASER2 via DROP) outliers. PanelApp gene panels relevant to each proband’s phenotypes were applied and cross-referencing with WGS data was utilised to flag candidate diagnostic events.

We will present the 100,000 Genomes Project RNAseq cohort, including information on disease classes and demographics. OUTRIDER identified on average 3.1 expression outliers per proband genome wide, with 8% of the cohort possessing an outlier event in a disease gene in a relevant gene panel. LeafCutterMD with stringent filtering reduced this to 4.0, consistent with FRASER2’s number of identified candidates. Analysis and review of candidates is ongoing, but we estimate around 21% of the cohort has at least one significant outlier in a disease relevant gene. Interesting examples of diagnostic discoveries will be identified and highlighted.

Although work is ongoing, we estimate at least 20% of the sequenced cohort will have a diagnostic candidate, which is likely to bring a significant uplift to diagnostic rates.

Conflict of Interest: None declared

C18.3 Full-length transcript atlas of the developing human cortex uncovers novel candidate diagnoses in developmental disorders

Rosemary Bamford¹, SziKay Leung¹, Kartik Chundru ^1;1, Caroline Wright¹, Jonathan Mill¹

¹University of Exeter, Department of Clinical and Biomedical Sciences, Exeter, United Kingdom

Background/Objectives: Alternative splicing enables multiple RNA isoforms to be produced from a single mRNA precursor, resulting in high levels transcriptomic and proteomic diversity. Long-read sequencing approaches can be used to generate full-length transcript sequences and fully characterise isoform diversity.

Methods: We used Oxford Nanopore Technologies (ONT) transcriptome sequencing to profile transcript diversity in the human cortex across development (n = 26 prenatal; n = 21 postnatal).

Results: We identify widespread transcript diversity in the developing cortex and observe novel transcript isoforms, with 55% of transcripts in protein-coding genes not previously characterised. Of the most abundant transcript within protein coding genes, 19% were novel isoforms.

We identify developmental changes in alternative splicing, with widespread differential transcript expression between human foetal and adult cortex. This has important implications for variant interpretation, as variant consequences may differ between isoforms depending on the open reading frame. We find regions of transcripts predicted to be protein-coding that were previously thought to be non-coding, including novel exons. These novel coding regions are highly conserved with a mean PhyloP conservation of 0.74, significantly higher than GENCODE UTR and intronic regions (0.32 and 0.12 respectively). Within the novel coding regions of genes associated with dominant developmental disorders, we find 15 de novo variants from undiagnosed neurodevelopmental disorder trios in the Genomics England 100,000 genomes project. Of these, 9/15 are predicted to be non-synonymous in the novel transcripts, including two stop-gained variants in patients with phenotypes consistent with the corresponding single gene disorder.

Conclusion: This highlights the potential clinical utility of long-read transcriptomics.

Grants: Simons Foundation [120682]. Wellcome [226083/Z/22/Z]

Conflict of Interest: None declared

C18.4 CRISPR activation: a new tool for RNA diagnostics

Thorkild Terkelsen ^1;2, Nanna S Mikkelsen², Ebbe N Bak³, Johan Vad-Nielsen², Jenny Blechingberg¹, Simone Weiss³, Simon O Drue³, Henning Andersen⁴, Brage S Andresen⁵, Rasmus O Bak², Uffe B Jensen^1;2

¹Aarhus University Hospital, Department of Clinical Genetics, Aarhus, Denmark; ²Aarhus University, Department of Biomedicine, Aarhus, Denmark; ³Aarhus University Hospital, Department of Molecular Medicine, Aarhus, Denmark; ⁴Aarhus University Hospital, Department of Neurology, Aarhus, Denmark; ⁵University of Southern Denmark, Department of Biochemistry and Molecular Biology, Odense, Denmark

Background/Objectives: Inadequate expression of tissue-specific gene transcripts in easily accessible cells from clinical samples is a major challenge for RNA diagnostics. We have recently evaluated the application of CRISPR activation (CRISPRa) to address this issue [1].

Methods: CRISPRa hijacks the capability of an inactive dCas9 protein to lead transcriptional activators to the gene of interest. The applied CRISPRa method uses an in vitro-transcribed mRNA, which encodes the activator complex, combined with guide RNAs that target the gene promotor. Pulsed delivery of this RNA mix to skin fibroblasts by simple electroporation induces gene upregulation within 24 hours. For RNA diagnostics, we have tested CRISPRa with reverse-transcription-PCR, short-read and long-read RNA sequencing platforms.

Results: We have applied CRISPRa in fibroblasts from individuals to revise the classification of suspected splice-altering variants in the myelin protein zero (MPZ) and spastin (SPAST) genes. With short-read RNA sequencing at routine diagnostic sequencing depths, CRISPRa resulted in excellent coverage of otherwise undetectable splice junctions in MPZ. Additionally, CRISPRa facilitated phasing of distant variants in SPAST by long-read RNA sequencing via increased abundance of the gene transcripts. In this way, CRISPRa supported the quantitative measurement of splicing impact.

Conclusion: CRISPRa is a promising new diagnostic tool to rapidly investigate the impact of splice-altering variants using easily accessible cells from individuals with rare diseases.

References: 1. Terkelsen T, Mikkelsen NS, Bak EN, Vad-Nielsen J, Blechingberg J, Weiss S, et al. CRISPR activation to characterize splice-altering variants in easily accessible cells. Am J Hum Genet. 2024;111(2):309-22.

Grants: DFF 9039-00337B

Conflict of Interest: Thorkild Terkelsen: None declared, Nanna S Mikkelsen: None declared, Ebbe N Bak: None declared, Johan Vad-Nielsen: None declared, Jenny Blechingberg: None declared, Simone Weiss: None declared, Simon O Drue: None declared, Henning Andersen: None declared, Brage S Andresen: None declared, Rasmus O Bak Holds equity in Graphite Bio and UNIKUM Tx., Paid consultant of UNIKUM Therapeutics, Uffe B Jensen: None declared

C18.5 Methylation quantitative trait loci mapping in more than 2,500 placentas provides insight into human complex disease.

Ariadna Cilleros-Portet ¹, Sergi Marí¹, Johanna Lepeule², Johanna Tuhkanen³, Marta Cosín-Tomàs⁴, Marika Groleau⁵, Kelly Bakulski⁶, charles breeze⁷, Todd Everson⁸, Mariana F. Fernandez^9;10;11;12, John Dou⁶, Loreto Santa-Marina^9;13;14, barbara heude¹⁵, jia chen¹⁶, jari lahti³, Marie-France Hivert^17;18;19, Mariona Bustamante⁴, Carmen Marsit⁸, Jose Ramon Bilbao¹, Corina Lesseur¹⁶, Nora Fernandez-Jimenez¹

¹University of the Basque Country (UPV/EHU) and Biobizkaia Health Research Institute, Department of Genetics, Physical Anthropology and Animal Physiology, Leioa, Spain; ²University Grenoble Alpes, Inserm, CNRS, Team of Environmental Epidemiology Applied to Reproduction and Respiratory Health, IAB, Grenoble, France; ³University of Helsinki, Department of Psychology and Logopedics, Helsinki, Finland; ⁴Institute for Global Health (ISGlobal), Barcelona, Spain; ⁵Université de Sherbrooke, Département de Biologie, Sherbrooke, Canada; ⁶University of Michigan, School of Public Health, Department of Epidemiology, Ann Arbor, United States; ⁷National Cancer Institute, Division of Cancer Epidemiology and Genetics, Rockville, United States; ⁸Rollins School of Public Health at Emory University, Gangarosa Department of Environmental Health, Atlanta, United States; ⁹Instituto de Salud Carlos III, Spanish Consortium for Research on Epidemiology and Public Health (CIBERESP), Madrid, Spain; ¹⁰University of Granada, Biomedical Research Center (CIBM), Granada, Spain; ¹¹Instituto de Investigación Biosanitaria de Granada (ibs.GRANADA), Granada, Spain; ¹²University of Granada, Department of Radiology and Physical Medicine, School of Medicine, Granada, Spain; ¹³Biogipuzkoa Health Research Institute, San Sebastian, Spain; ¹⁴Subdirectorate of Public Health of Gipuzkoa, Department of Health of the Basque Government, San Sebastian, Spain; ¹⁵Université de Paris, Centre for Research in Epidemiology and Statistics (CRESS), INSERM, INRAE, Paris, France; ¹⁶Icahn School of Medicine at Mount Sinai, Department of Environmental Medicine and Public Health, New York, United States; ¹⁷Harvard Medical School and Harvard Pilgrim Health Care Institute, Department of Population Medicine, Boston, United States; ¹⁸Centre de Recherche du Centre Hospitalier Universitaire de Sherbrooke (CHUS), Sherbrooke, Canada; ¹⁹Massachusetts General Hospital, Diabetes Unit, Boston, Spain

Consortium: Pregnancy And Childhood Epigenetics (PACE) Consortium

Background/Objectives: The Developmental Origins of Health and Disease hypothesis (DOHaD) posits that the early-life environment can impact later-life health. The placenta plays a key role in DOHaD as it regulates the intrauterine environment. Placental DNA methylation (DNAm) has been suggested as a mechanistic link between intrauterine exposures, genetics, and adverse later-life health outcomes. Herein, we characterized genetic factors contributing to disease susceptibility through placental DNAm.

Methods: We meta-analysed placental methylation Quantitative Trait Loci (mQTLs) in 2,583 samples from nine cohorts of the international Pregnancy And Childhood Epigenetics (PACE) consortium. Each cohort mapped cis-mQTLs (0.5 Mb window) with TensorQTL. The meta-analysis was performed with the inverse variance weighted method of the software GWAMA. Afterwards, Stratified-LD Score Regression (S-LDSC) and Summary-based Mendelian Randomization (SMR) were performed to estimate the contribution of placental DNAm to 36 complex traits and disorders, by using the summary statistics of the largest GWAS available.

Results: We obtained more than 25 million placental cis-mQTLs, with 210,476 and 3,531,259 unique CpGs and SNPs, respectively. Placental mQTLs were enriched in genomic features with intermediate methylation values, and placenta specific accessible chromatin regions. S-LDSC and SMR analyses showed that SNPs linked to immune and anthropometric traits, and specific diseases such as schizophrenia, myocardial infraction, and Hashimoto’s thyroiditis are enriched for placental mQTLs.

Conclusion: The results obtained highlight the impact of genetics through placental DNAm in a range of different complex traits and disorders, reinforcing the role of the placenta in the DOHaD hypothesis.

Grants: GVSAN2018111086, PI18/01142, GVSAN2019111085, PI20/01116, PI21/01491, and PRB2-ISCIII.

Conflict of Interest: None declared

C18.6 Genetic regulators of whole blood gene expression and splicing and their impact on molecular traits and complex diseases

Elodie Persyn ^1;2;3, Alex Tokolyi⁴, Artika Nath^3;5, Katie L Burnham⁴, Jonathan Marten^1;2, Thomas Vanderstichele⁴, Manuel Tardaguila^4;6, David Stacey^1;7;8, Ben Farr⁴, Vivek Iyer⁴, Xilin Jiang^1;2;9, Samuel A. Lambert^1;2;3, Guillaume Noell⁴, Michael Quail⁴, Diana Rajan⁴, Scott C. Ritchie^1;2;3, Benjamin B. Sun¹⁰, Scott A.J. Thurston⁴, Yu Xu^1;2;3, Christopher D. Whelan¹¹, Heiko Runz¹⁰, Slavé Petrovski^12;13, Daniel J. Gaffney^4;14, David Roberts¹⁵, Emanuele Di Angelantonio^1;2, James Peters¹⁶, Nicole Soranzo^4;6, John Danesh^1;2;4, Adam Butterworth^1;2, Michael Inouye^1;2;3, Emma Davenport⁴, Dirk Paul^1;2;12

¹British Heart Foundation Cardiovascular Epidemiology Unit, Department of Public Health and Primary Care, University of Cambridge, Cambridge, United Kingdom; ²Victor Phillip Dahdaleh Heart and Lung Research Institute, University of Cambridge, Cambridge, United Kingdom; ³Cambridge Baker Systems Genomics Initiative, Department of Public Health and Primary Care, University of Cambridge, Cambridge, United Kingdom; ⁴Wellcome Sanger Institute, Wellcome Genome Campus, Hinxton, United Kingdom; ⁵Cambridge Baker Systems Genomics Initiative, Baker Heart and Diabetes Institute, Melbourne, Australia; ⁶Human Technopole, Fondazione Human Technopole, Milan, Italy; ⁷Australian Centre for Precision Health, University of South Australia, Adelaide, Australia; ⁸South Australian Health and Medical Research Institute, Adelaide, Australia; ⁹Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, United States; ¹⁰Translational Sciences, Research & Development, Biogen, Cambridge, United States; ¹¹Neuroscience Data Science, Janssen Research & Development, Cambridge, United States; ¹²Centre for Genomics Research, Discovery Sciences, BioPharmaceuticals R&D, AstraZeneca, Cambridge, United Kingdom; ¹³Department of Medicine, University of Melbourne, Austin Health, Melbourne, Australia; ¹⁴Genomics, BioMarin Pharmaceutical Inc., Novato, United States; ¹⁵Radcliffe Department of Medicine, University of Oxford, Oxford, United Kingdom; ¹⁶Department of Immunology and Inflammation, Imperial College London, London, United Kingdom

Background/Objectives: As many genetic risk factors are localised in non-coding regions of the genome, the integration of multi-omics datasets is often used to better understand downstream molecular mechanisms. In this study, we analysed RNA-sequencing data from the INTERVAL cohort to identify genetic regulators of gene expression and splicing events in whole blood and assess the impact of these genetic loci on other molecular phenotypes (proteins and metabolites) and complex diseases.

Methods: We tested the association between genetic variants and the expression of 19,173 genes and 111,937 splicing events, from the RNA-sequencing data of 4,732 INTERVAL participants. Shared genetic factors with other molecular traits and complex diseases were identified by conducting colocalization analyses. As subsets of INTERVAL participants had multiple -omics data available, we further tested the mediating effect of gene expression and splicing on other molecular traits.

Results: We found cis-quantitative trait loci (cis-QTL) for the expression of 17,233 genes and 29,514 splicing events. Genetic colocalization analyses identified 3,430 proteomic and metabolomic traits and 24 selected diseases with a shared genetic factor with either gene expression or splicing. Focusing on those colocalized signals, we found significant mediated effect on 222 molecular phenotypes.

Conclusion: By assessing the shared genetic aetiology with other molecular traits, our study provides new insights on molecular mechanisms. These results can be integrated in genetic studies to better understand complex diseases pathophysiology and identify new drug targets. The full description of our study is available on medRxiv with available data on https://IntervalRNA.org.uk.

Grants: BHF Programme Grant (RG/18/13/33946)

Conflict of Interest: Elodie Persyn E.P. was funded by the EU/EFPIA Innovative Medicines Initiative Joint Undertaking BigData@Heart grant 116074 and is funded by the NIHR BTRU in Donor Health and Behaviour (NIHR203337), Alex Tokolyi A.T. is supported by the Wellcome Trust (PhD studentship 222548/Z/21/Z), Artika Nath: None declared, Katie L Burnham: None declared, Jonathan Marten: None declared, Thomas Vanderstichele: None declared, Manuel Tardaguila: None declared, David Stacey: None declared, Ben Farr: None declared, Vivek Iyer: None declared, Xilin Jiang: None declared, Samuel A. Lambert: None declared, Guillaume Noell: None declared, Michael Quail M.A.Q. is on the KOL panel for New England Biolabs., Diana Rajan: None declared, Scott C. Ritchie S.C.R was funded by a BHF Programme Grant (RG/18/13/33946) and the NIHR Cambridge BRC (BRC-1215-20014; NIHR203312)., Benjamin B. Sun B.B.S. is a stockholder of Biogen, B.B.S. is an employee of Biogen, Scott A.J. Thurston: None declared, Yu Xu Y.X. is supported by the UK Economic and Social Research Council (ES/T013192/1)., Christopher D. Whelan C.D.W. is a stockholder of Johnson & Johnson, C.D.W. is an employee of Johnson & Johnson, Heiko Runz H.R. is a stockholder of Biogen, H.R. is an employee of Biogen, Slavé Petrovski S.P. is a stockholder of AstraZeneca, S.P. is an employee of AstraZeneca, Daniel J. Gaffney D.J.G. is a stockholder of BioMarin Pharmaceutical, D.J.G is an employee of BioMarin Pharmaceutical, David Roberts D.J.R. is an employee of NHS Blood and Transplant., Emanuele Di Angelantonio: None declared, James Peters J.E.P. has received hospitality and travel expenses to speak at Olink-sponsored academic meetings (none within the past 5 years), J.E.P. is supported by a Medical Research Foundation Fellowship (MRF-057-0003-RG-PETE-C0799), Nicole Soranzo: None declared, John Danesh J.D. holds a British Heart Foundation Professorship and a NIHR Senior Investigator Award., Adam Butterworth A.S.B. has received grants outside of this work from AstraZeneca, Bayer, Biogen, BioMarin and Sanofi, Michael Inouye M.I. is a trustee of the Public Health Genomics (PHG) Foundation. M.I has a research collaboration with AstraZeneca that is unrelated to this study, M.I. is supported by the Munz Chair of Cardiovascular Prediction and Prevention and the NIHR Cambridge Biomedical Research Centre (BRC-1215-20014; NIHR203312). M.I. is also supported by the UK Economic and Social Research Council (ES/T013192/1)., M.I is a member of the Scientific Advisory Board of Open Targets, Emma Davenport: None declared, Dirk Paul D.S.P is a stockholder of AstraZeneca, D.S.P is an employee of AtraZeneca

C19.1 Inhibition of phosphodiesterase 10A by a specific antagonist rescues behavioral deficits and normalizes microglial morphology and synaptic pruning in a mouse model of FOXP1 syndrome

Henning Fröhlich ¹, Jing Wang¹, Ferdinand Althammer¹, Tim Schubert¹, Nina Kluck¹, Valery Grinevich², Stefanie Schmitteckert¹, Christian Schaaf¹, Gudrun Rappold¹

¹Institute of Human Genetics, Heidelberg University Hospital, Heidelberg, Germany; ²Central Institute of Mental Health, Medical Faculty Mannheim, University of Heidelberg, Mannheim, Department of Neuropeptide Research in Psychiatry, Mannheim, Germany

Background/Objectives: FOXP1 syndrome caused by FOXP1 haploinsufficiency is characterized by intellectual disability, speech and language impairment, autistic traits and neuropsychiatric abnormalities such as anxiety and hyperactivity. Despite research into its pathogenesis, effective treatment options are still lacking. Behavioral changes in patients are mirrored in Foxp1^+/- mice. Phosphodiesterases (Pde) have been hypothesized to play a role in the pathophysiology of neurodevelopmental disorders. We investigated in Foxp1^+/- animals whether Pde10a inhibition starting immediately after birth has a positive effect on the striatum and alleviates their behavioral deficits.

Methods: Pde10a mRNA and protein levels were determined by quantitative real-time PCR and western blot at P1, P12 and adulthood. Bulk expression of striatal tissue was analyzed with a NanoString nCounter Neuroinflammation Panel. One of the most advanced PDE10A antagonists was administered intraperitoneally and we studied its effect on behavior focusing on ultrasonic vocalization, anxiety and hyperactivity. We assessed astrocyte and microglia morphology on brain sections using 3D reconstruction by IMARIS.

Results: Foxp1^+/- striata show significantly decreased Pde10a mRNA and protein levels from birth to adulthood. Moreover, they exhibit an upregulation of the neuroinflammation markers Cd74, Ifi30, Fcgr2b and Rhoa as well as increased microglial cell volume, decreased filament length and branching and decreased synaptic pruning. Pde10a inhibition in Foxp1^+/- mice not only corrected the observed behavioral deficits, but also normalized microglial morphology and synaptic pruning without detectable side effects.

Conclusion: Reduced PDE10A expression most likely also contributes to the cognitive and behavioral deficits in FOXP1 syndrome, suggesting that PDE10A antagonists also alleviate neurological dysfunction in these individuals.

Grants: FR 3990/3-1

Conflict of Interest: None declared

C19.2 Nanoparticles-mediated siRNA delivery as a noninvasive personalized therapeutic strategy to hamper premature suture ossification in Crouzon syndrome

FEDERICA TIBERIO ^1;2, Martina Salvati², Elisabetta Falvo³, Luca Polito², Noah Giacon⁴, Lorena Di Pietro^1;2, Gianpiero Tamburrini⁵, Giada Tisci⁶, Pierpaolo Ceci³, Luca Massimi⁵, Ornella Parolini^1;2, Alessandro Arcovito⁴, Wanda Lattanzi^1;2

¹Fondazione Policlinico Universatario Agostino Gemelli IRCSS, Rome, Italy; ²Università Cattolica del Sacro Cuore, Scienze della Vita e Sanità Pubblica, Roma; ³Consiglio Nazionale delle Ricerche, Istituto di Biologia Molecolare e Patologia, Rome, Italy; ⁴Università Cattolica del Sacro Cuore, Dipartimento di Scienze Biotecnologiche di Base, Cliniche Intensivologiche, e Perioperatorie, Roma, Italy; ⁵Fondazione Policlinico Universatario Agostino Gemelli IRCSS, Dipartimento di Neuroscienze, Rome, Italy; ⁶Sapienza Università di Roma, Dipartimento di Scienze Biochimiche, Rome, Italy

Background/Objectives: Crouzon syndrome (CS) is a rare genetic condition caused by heterozygous gain-of-function mutations in the Fibroblast growth factor receptor 2 (FGFR2) gene, which lead to the premature ossification of skull sutures. CS treatment is based on craniectomy, burdened by complications. This study aims at developing a noninvasive personalized FGFR2 knockdown strategy using allele-specific siRNAs, delivered by highly biocompatible polymeric nanoparticles (NPs).

Methods: Calvarial-derived mesenchymal stromal cells (CMSC) were isolated from surgical waste tissues of FGFR2 mutation-positive CS patients. siRNAs targeting mutant FGFR2 alleles have been designed and tested in CMSC. Their silencing and allele-discrimination efficiency were analyzed through gene (qPCR) and protein expression analyses. The intracellular delivery and trafficking of selected siRNAs vehicled through nanoparticles based on human recombinant ferritin and of Poly(d,l-lactide-co-glycolide) (PLGA) were evaluated by Incucyte Live-cells analysis system and confocal microscopy. siRNA-ferritin NPs and siRNA-PLGA complexes were produced and tested in CMSC by qPCR and Western Blot.

Results: For each enrolled patient, we developed specific siRNAs that reduced the expression of FGFR2 mutant allele by at least 50% without affecting the wild-type. The intracellular trafficking analysis showed an efficient internalization of both NPs, widely distributing within the CMSC cytoplasm. Our data demonstrated an efficient NP-mediated delivery and release of siRNA in cells, reducing FGFR2 expression levels by 40%.

Conclusion: Polymeric NPs-mediated siRNAs‘ allele-specific FGFR2 knockdown represents a suitable strategy to dampen the constitutive expression of FGFR2 in CS patients’ cells.

Grants: Funded by AFM-Telethon (Research Grant #23607) and by POR FESR Lazio 2014-2020 (Protocol: A0375-2020- 36576)

Conflict of Interest: FEDERICA TIBERIO Post-doc researcher full time, Martina Salvati PhD student full time, Elisabetta Falvo patent N. WO2016051340A1 entitled “A fusion protein, a nanoparticle composed by a plurality of monomers of said fusion protein and uses thereof”. The patent was sold to the company Thena Biotech S.r.l. in 2017. A Research Activity Agreement between the Institution of team#2 (C.N.R.) and Thena Biotech S.r.l. is currently operative for the projects involving the human ferritins as delivery systems., Researcher full time, Luca Polito Post-doc researcher full time, Noah Giacon PhD student full time, Lorena Di Pietro Researcher full time, Gianpiero Tamburrini MD full time and Full Professor, Giada Tisci PhD student full time, Pierpaolo Ceci patent N. WO2016051340A1 entitled “A fusion protein, a nanoparticle composed by a plurality of monomers of said fusion protein and uses thereof”. The patent was sold to the company Thena Biotech S.r.l. in 2017. A Research Activity Agreement between the Institution of team#2 (C.N.R.) and Thena Biotech S.r.l. is currently operative for the projects involving the human ferritins as delivery systems., Researcher full time, Luca Massimi MD full time, Ornella Parolini Full Professor, Alessandro Arcovito Associate Professor, Wanda Lattanzi Principal Investigator AFM-Telethon (Research Grant #23607) and POR FESR Lazio 2014-2020 (Protocol: A0375-2020- 36576), MD and associate professor full time

C19.3 Modulating splicing in 5’ untranslated regions to treat rare haploinsufficient disorders

Eloise Wells ^1;2, Ruebena Dawes², Stephan Sanders^1;3, Nicola Whiffin²

¹Institute of Developmental & Regenerative Medicine (IDRM) - University of Oxford, United Kingdom; ²Nuffield Department of Medicine (NDM), University of Oxford, United Kingdom; ³UCSF Joan and Sanford I. Weill Neurosciences Building, San Francisco, United States

Background/Objectives: Rare disorders collectively affect over 300 million people worldwide, but over 95% have no effective treatments. Recent advances in oligonucleotide and CRISPR editing technologies are driving therapeutic development. It is relatively straightforward to down-regulate genes therapeutically, for example with gapmer antisense oligonucleotides (ASOs). However, while haploinsufficiency is a major cause of rare disorders, upregulating protein levels is considerably more difficult.

Methods: We explore a novel approach to upregulation: modulating splicing of 5’ untranslated regions (5’UTRs) to increase protein translation. Specifically, we aim to selectively exclude 5’UTR exons that contain upstream open reading frames (uORFs), which are negative regulators of translation.

Results: We find 2,206 ‘skippable’ 5’UTR exons (i.e. flanked by other 5’UTR encoding exons) in 1,573 MANE transcripts. Using ribosome profiling data from human tissues including brain, heart, and skeletal muscle, we identify start sites of translated uORFs in 650 of these skippable exons (29%). Sixty-three of these uORF-containing exons are in the 5’UTRs of 48 genes with a dominant loss-of-function disease mechanism (from G2P, ClinVar, GenCC, and/or PanelApp). For each candidate gene, we assessed alternative 5’UTR splicing in long-read RNA-sequencing data and predicted changes in secondary structure and uORF translation in the presence and absence of the identified skippable exon(s). Subsequently, we prioritised nine target exons with a CRISPR-editable splice-acceptor site or that could be targeted by a splice-switching ASO for on-going functional validation.

Conclusion: Up-regulation of translation through modifying 5’UTR splicing is a promising therapeutic approach for a subset of haploinsufficient disorders.

Grants: Wellcome/Royal Society (220134/Z/20/Z)

Conflict of Interest: Eloise Wells: None declared, Ruebena Dawes: None declared, Stephan Sanders S.J.S receives research funding from BioMarin Pharmaceuticals, Nicola Whiffin Partner is an exmployee and owns stock options in Vertex Pharmaceuticals., Research funding from Novo Nordisk., Has consulted for ArgoBio.

C19.4 Microparticle sustained delivery of the calcineurin inhibitor FK506 to curtail autophagy and restore vision in an ADOA mouse model

Ana Paula Miranda Mendonca ¹, ricardo rampado², paolo caliceti³, stefano salmaso³, Luca Scorano¹

¹University of Padova, Department of Biology, Padua, Italy; ²Tel Aviv University, Tel Aviv-Yafo, Israel; ³University of Padova, Department of Pharmaceutical and Pharmacological Sciences, Padua, Italy

Background/Objectives: Autosomal dominant optic atrophy (ADOA) is the most common hereditary neuropathy that affects 1:35.000 people worldwide, and treatment is not available. Mutations in OPA1 in ADOA patients promote low levels of OPA1 protein, causing Retinal Ganglion Cells (RGCs) degeneration leading to blindness. Our group showed fragmented mitochondria, excluded from the axon but combined with autophagosomes in the axonal hillock of the ADOA mouse model. Moreover, genetic autophagy inhibition restores mitochondrial distribution in the RGC axon and curtails visual loss in the ADOA animal. Given that, this project aims to develop a pharmacological therapy through vitreal injection to curtail ADOA mouse visual loss.

Methods: We developed PLGA microparticles loaded with FK506 (PLGA-FK506), an inhibitor of autophagy in RGCs. The ADOA mouse model (Opa1^fl/fl::Grik4-Cre mice) was subjected to an PLGA-FK506 intravitreal treatment over 3 months. The visual acuity of mice was measured by optokinetic tests.

Results: The ADOA mouse model loses visual acuity after 3 months. However, PLGA-FK506 treatment prevented loss of visual acuity in ADOA animals, and it was maintained over 3 months.

Conclusion: The treatment with PLGA-FK506 prevented ADOA’s vision loss for 3 months.

Grants: Marie Sklodowska-Curie grant n° 101030965 and Telethon Foundation project n° GGP19089.

Conflict of Interest: None declared

C19.5 Enhanced gene editing in patient-derived hiPSCs as a therapeutic strategy for inherited retinal dystrophies

Laura Siles ¹, Sheila Ruiz Nogales¹, Pilar Mendez¹, Paula Gaudó¹, Arnau Navinés-Ferrer¹, Esther Pomares¹

¹IMO Grupo Miranza, Genetics, Barcelona

Background/Objectives: Inherited retinal dystrophies (IRD) are a group of rare disorders triggering blindness and other ocular affectations that have no treatment. They are caused by mutations in near 300 genes resulting in retinal cells degeneration. By using gene editing we aim to precisely correct IRD-related pathogenic variants (PV) as a therapeutic approximation for IRD.

Methods: We used CRISPR/Cas9 and TALEN technologies for the correction of PV in iPSCs derived from patients affected by retinitis pigmentosa, Stargardt disease, Best disease or achromatopsia. Edited cells were differentiated into in vitro retina cell models to analyze the reversion of disease-associated phenotypes.

Results: We have corrected seven hiPS cell lines without detected off-target defects. In most assays, at least 50% of the clones exhibit PV editing without undesired genomic abnormalitites. The highest gene editing results were obtained in homozygous mutations and small insertions/deletions, achieving a maximum efficiency of the 80%. Moreover, sgRNA specificity towards mutant allele decisively influences gene editing outcomes. Surprisingly, some hiPS clones edited the PV through the use of the wild-type allele as repair template in case of heterozygous carriers. Finally, corrected clones exhibit improved retinal cell differentiation and function coinciding with reversion of the diseased condition.

Conclusion: Gene editing represents a promising approach for IRD treatment, with CRISPR/Cas9 postulating as a more robust technology, but TALENs being appropriate in cases where CRISPR not. These results strengthen the study of gene editing as a therapeutic approximation for IRD, and suggest an unexpected potential application for PV repair by using patient’s own wild-type allele.

Grants: Fi-201401

Conflict of Interest: None declared

C19.6 Preventing Cancer in Lynch Syndrome: Vaccination with Mutation-Derived Neoantigen-Loaded Dendritic Cells Eliminates Precancerous Cells

Nicoline Hoogerbrugge ¹, asima abidi², harm westdorp³, mark gorris⁴, steve boudewijns⁴, Roland Kuiper⁵, reno bladergroen⁵, Tanya Bisseling⁶, anna de Goede⁷, Michelle van Rossum⁸, Alexandra Croockewit⁹, Marjolijn Ligtenberg¹, Iris Nagtegaal¹⁰, Carl Figdor¹¹, winald Gerritsen¹², Gerty Schreibelt⁴, Jolanda de Vries⁴

¹Radboudumc, Human Genetics, Nijmegen; ²Radboud; ³Radboudumc, Mecical Oncology, Nijmegen; ⁴Radboudumc, Medical BioSciences, Nijmegen; ⁵Princess Máxima Center for Pediatric Oncology, Human Genetics, Utrecht; ⁶Radboudumc, Gastroenterology, Nijmegen; ⁷Radboudumc, Pharmacy, Nijmegen; ⁸Radboudumc, Dermatology, Nijmegen; ⁹Radboudumc, Hematology, Nijmegen; ¹⁰Radboudumc, Pathology, Nijmegen; ¹¹Radboudumc, Medical, Nijmegen; ¹²Radboudumc, Medical Oncology, Nijmegen

Background/Objectives: Individuals with Lynch syndrome (LS) have a hereditary risk of colorectal cancer and other cancers due to pathogenic variants in DNA mismatch repair genes. Mismatch repair deficiency results in the formation of neoantigens in the tumours. Dendritic cells (DC) loaded with these neoantigens can stimulate immune responses against tumour cells expressing these neoantigens, ultimately leading to tumour eradication. Our aim was to evaluate the potential of using DC-based vaccination in LS.

Methods: From 2012 to 2016, 23 individuals with LS from a phase I/II clinical trial received DC loaded with predicted neoantigen peptides from TGF-βRII, caspase-5, and carcinoembryonic antigen (CEA). Patients received up to 3 vaccination cycles, each consisting of 3 weekly injections.

Results: Neoantigen- and CEA-specific T-cells were detected in post-treatment skin biopsies in 87% of the patients. In-vitro expanded antigen specific T-cells were capable of specifically lysing tumour cells presenting 1) exogenously loaded peptides of all three (neo)antigens loaded on the DC’s and 2) endogenously processed TGF-βRII and CEA.

Higher probability of disease-free 10-year survival post vaccination was found to correlate with the presence of TGF-βRII specific T-cells.

Conclusion: LS patients who mounted a T-cell response against the neoantigen TGF-βRII did not develop cancer in almost 10 years. Within these 10 years, premalignant lesions did not exhibit the mutated form of TGF-βRII. This confirms that T-cells prevent the outgrowth of (pre)cancerous cells expressing the mutation-induced neopeptide, hence prevent the progression from adenoma to carcinoma. Prophylactic DC vaccination offers a new preventive treatment modality for patients with LS.

Conflict of Interest: None declared

C20.1 Exome-wide evidence of compound heterozygous effects across common phenotypes in the UK Biobank

Frederik Heymann Lassen ¹, Samvida Venkatesh¹, Nikolas Baya¹, Wei Zhou², Alex Bloemendal², Benjamin Neale², Benedikt Kessler³, Nicola Whiffin¹, Cecilia Lindgren¹, Duncan Palmer¹

¹University of Oxford, Oxford, United Kingdom; ²Broad Institute of MIT and Harvard, Cambridge, United States; ³University of Oxford, Target Discovery Institute, Oxford, United Kingdom

Background/objectives: Exome sequencing has identified rare protein-coding variants underlying many diseases, typically studying heterozygous and homozygous variation. However, the phenotypic impact of deleterious compound heterozygous (CH) variants on trait variation at scale remains largely unexplored.

Methods: We statistically phased rare variants (minor allele frequency (MAF) ~ 0.001%) in the UK Biobank exome sequencing data and analysed recessive effects in 175,587 individuals across 311 common diseases. Using a logistic mixed model framework, we investigated CH effects while accounting for relatedness, polygenic background, nearby common variants, and rare variant burden. We incorporated electronic health record age-at-diagnosis data to explore the longitudinal effects of bi-allelic variation. We replicate our associations in the 233,837 remaining UK Biobank participants. Finally, we developed an additive haplotype-based collapsing model, modelling the number of disrupted gene copies, comparing it with SAIGE-GENE+ burden testing.

Results: We found that 6.5% of individuals carry CH variants, and 90% these occur with a frequency below 0.34%. We find 19 significant (FDR < 0.1) recessive bi-allelic gene-trait associations, of which only six (31.1%) would have been discovered considering homozygosity alone. We find known bi-allelic gene-trait associations such as MUTYH and colorectal cancer (FDR < 0.1) and identify novel relationships such as ATP2C2-COPD (FDR < 0.1) driven by CH variation. Lastly, among our significant additive haplotype-based gene associations, 20/21 were more strongly associated than in SAIGE-GENE+ burden testing.

Conclusion: We demonstrate the utility of phasing large-scale cohorts for identification of the phenome-wide consequences of compound heterozygosity.

Grants: Wellcome Trust award 224894/Z/21/Z

Conflict of Interest: Frederik Heymann Lassen: None declared, Samvida Venkatesh: None declared, Nikolas Baya: None declared, Wei Zhou: None declared, Alex Bloemendal: None declared, Benjamin Neale B.M.N. is a member of the scientific advisory board at Deep Genomics and Neumora., Benedikt Kessler: None declared, Nicola Whiffin: None declared, Cecilia Lindgren: None declared, Duncan Palmer: None declared

C20.2 Unsupervised phenotyping and genetic discovery in 62,495 cardiac MRIs

Sara Ometto ¹, Soumick Chatterjee¹, Andrea Mario Vergani¹, Alessia Visconti¹, Emanuele Bianchi¹, Edoardo Giacopuzzi¹, Arianna Landini¹, Sodbo Sharapov¹, Francesco Cisternino¹, Emanuel Soda¹, Francesca Ieva¹, Emanuele Di Angelantonio¹, Nicola Pirastu¹, Craig Anthony Glastonbury¹

¹Human Technopole, Milan, Italy

Background/Objectives: Temporal cardiac MRI imaging is an information rich modality but manually deriving multiple phenotypes of organ function via annotation is a rate limiting step. Autoencoders can learn high level (latent) semantic concepts from images in an unsupervised manner. These latent representations can capture complex phenotypes that can then be used to drive novel genetic discovery.

Methods: We developed a 3D diffusion autoencoder (3D-diffAE) to extract robust latent phenotypes from four chamber view long-axis cardiac MRIs across 62,495 subjects in UK Biobank (UKB). Latent phenotype-clinical associations were assessed using linear regression. Testing of common SNVs associated with latent phenotypes was performed using REGENIE. Conditional analysis was conducted with COJO and colocalisation performed using multi-trait coloc. Replication was performed in a held-out subset of UKB.

Results: Using 3D-DiffAE, we inferred latent phenotypes that are reproducible, heritable (4%-18%), and significantly associated with atrial fibrillation (P-value = 1.9e-88), myocardial infarction (P-value = 4.4e-71), and other outcomes in concurrently measured EHR data. GWAS (n = 47,827) identified 76 genome-wide significant associations in 182 endophenotypes mapping to 40 independent loci, of which 6 have not been previously described. By integrating GWAS summary statistics from 56 cardiometabolic traits, we find that 10 loci colocalise (PP4 > 0.9) with at least one derived cardiac phenotype and 6 with cardiac diseases.

Conclusion: We performed the largest systematic characterisation of cardiac endophenotypes to date, demonstrating that unsupervised phenotyping of cardiac MRIs using diffusion models provides a tractable means for novel genetic discovery and interpretation.

Grants: N.A

Conflict of Interest: None declared

C20.3 Decomposing sex-different phenotypic correlations in the UK Biobank into genetic and environmental components

Amelie Fritz ^1;2;3, Liza Darrous¹, Anders G. Pedersen², Klaus Bønnelykke³, Zoltan Kutalik¹

¹UNIL - Université de Lausanne, Lausanne, Switzerland; ²Technical University of Denmark, Department of Health Tech, Kgs. Lyngby, Denmark; ³COPSAC, Gentofte, Denmark

Background/Objectives: Sexual dimorphism influences disease patterns, stature, and mental health. The role of genetic versus environmental factors in these sex-specific differences, particularly in traits linked to risk factors like smoking and lung function, are unclear but crucial for understanding disease etiology.

Methods: We developed a method to calculate the environmental correlation (rE) of trait pairs using phenotypic and genetic correlation (rG) and heritabilities. Taking advantage of big sample sizes and multiple phenotypes per individual, we analyzed 299 trait pairs from the UK Biobank exhibiting significantly different phenotypic correlations across sexes and investigated the extent to which this is influenced by environmental and genetic factors.

Results: We observed differing rG and rE across sexes for blood biomarkers e.g. c-reactive-protein~smoking with a greater contribution of genetic components in males (rG_m = -0.38;95% CI = [-0.47, -0.29]/rE_m = -0.07;95%CI = [-0.073, -0.069]) than females (rG_f = -0.15;95%CI = [-0.23,-0.08]/rE_f = -0.042;95%CI = [-0.044, -0.04]), whereas for hdl-cholesterol~education females show greater genetic contribution (rG_f = 0.38;95%CI = [0.379, 0.46]/rE_f = 0.05; 95%CI = [0.048,0.056] - rG_m = 0.11;95%CI = [0.02,0.20]/rE_m = 0.03;95%CI = [0.027,0.031]). These findings are supported by intra-trait rG between males and females being significantly different from 1, indicating differing genetic mechanisms across sexes.

Sex-specific significant differences were also observed in pairs involving testosterone with sex hormone-binding globulin, urate, waist-hip-ratio, and triglycerides.

Conclusion: We demonstrate that phenotypic correlations between trait pairs are differentially influenced by genetic and environmental effects across sexes, therefore different mechanisms may be responsible for the outcome of certain risk factors on traits or diseases. Future research will focus on clustering traits into groups representing different health aspects and assessing via Mendelian Randomization whether sex-different causal effects may give rise to the observed sex-dependent rG.

Grants: NovoNordiskFoundation

Conflict of Interest: None declared

C20.4 Large-scale exome sequencing identifies an enriched missense variant in HNF4A protective from type 2 diabetes (T2D) in South Asians

Moeen Riaz ¹, Joshua Backman¹, Bin Ye¹, Manav Kapoor¹, Jonas Bovijn¹, Sameer Malhotra¹, SR Chaudhury², Md Alfazal Khan³, Md Khalequzzaman⁴, Adam Butterworth⁵, Luca Lotta¹, Jonathan Marchini¹, Aris Baras¹, Goncalo Abecasis¹, Alan R Shuldiner¹

¹Regeneron Genetics Center, New York, United States; ²National Heart Foundation Hospital & Research Institute, Dhaka, Bangladesh; ³icddrb, Dhaka, Bangladesh; ⁴Bangabandhu Sheikh Mujib Medical University, Dhaka, Bangladesh; ⁵University of Cambridge, Hinxton, United Kingdom

Consortium: Regeneron Genetics Center, the BELIEVE study group

Background: South Asians (SA) have a fourfold higher risk of T2D compared to Europeans. Protein-coding genetic variants that affect disease risk or protection can provide clues to disease pathogenesis or therapeutics targets.

Methods: We performed genome and exome-wide association analysis in two South Asian cohorts (BangladEsh Longitudinal Investigation of Emerging Vascular Events study [BELIEVE] and UK Biobank SA) using an additive model for single variants and SKAT/SKATO for gene-based test, adjusted for age, sex and 20 genetic PCs in REGENIE. Inverse-variance weighted meta-analysis was conducted, and GWAS summary statistics from the Genes and Health Study were added, making a total effective sample size of 11,205 T2D cases and 52,135 controls for exome analysis and 21,467 T2D cases and 86,061 controls for imputed GWAS analysis.

Results: 37 genome-wide significant (P < 5e-08) index variants were associated with T2D. These include 4 novel and 33 previously known common variants (MAF > 0.01). Gene burden masks showed no genome-wide significant associations with T2D. A novel protective association of a missense variant rs150776703 (p.Pro437Ser) in HNF4A (P = 1.738e-13, effect = 0.51) was identified. Rs150776703 is enriched in Bangladeshi (MAF 0.01) compared to gnomAD Non-Finnish Europeans (MAF 1.47E-05). We also observed nominal protective association of this variant with self-reported chronic kidney disease (P = 9.5e-4, OR 0.48). HNF4A encodes a transcription factor, and loss-of-function mutations cause MODY1. p.Pro437Ser occurs in the repression domain and is putatively a gain-of-function mutation.

Conclusion: Our findings provide first human genetic evidence for activation of HNF4A as a potential therapeutic approach for T2D.

Conflict of Interest: Moeen Riaz Regeneron Genetics Center or Regeneron Pharmaceuticals., Regeneron Genetics Center and Regeneron Pharmaceuticals., Joshua Backman Regeneron Genetics Center or Regeneron Pharmaceuticals., Regeneron Genetics Center and Regeneron Pharmaceuticals., Bin Ye Regeneron Genetics Center or Regeneron Pharmaceuticals., Regeneron Genetics Center and Regeneron Pharmaceuticals., Manav Kapoor Regeneron Genetics Center or Regeneron Pharmaceuticals., Regeneron Genetics Center and Regeneron Pharmaceuticals., Jonas Bovijn Regeneron Genetics Center or Regeneron Pharmaceuticals., Regeneron Genetics Center and Regeneron Pharmaceuticals., Sameer Malhotra Regeneron Genetics Center or Regeneron Pharmaceuticals., Regeneron Genetics Center and Regeneron Pharmaceuticals., SR Chaudhury: None declared, Md Alfazal Khan: None declared, Md Khalequzzaman: None declared, Adam Butterworth: None declared, Luca Lotta Regeneron Genetics Center or Regeneron Pharmaceuticals., Regeneron Genetics Center and Regeneron Pharmaceuticals., Jonathan Marchini Regeneron Genetics Center or Regeneron Pharmaceuticals., Regeneron Genetics Center and Regeneron Pharmaceuticals., Aris Baras Regeneron Genetics Center or Regeneron Pharmaceuticals., Regeneron Genetics Center and Regeneron Pharmaceuticals., Goncalo Abecasis Regeneron Genetics Center or Regeneron Pharmaceuticals., Regeneron Genetics Center and Regeneron Pharmaceuticals., Alan R Shuldiner Regeneron Genetics Center or Regeneron Pharmaceuticals., Regeneron Genetics Center and Regeneron Pharmaceuticals.

C20.5 An atlas of quantitative trait loci for a hundred gestational phenotypes

Siyang Liu ¹, Hao Zheng^1;2, yuqin Gu¹, Zijing Yang^1;3, Yanhong Liu¹, Yuandan Wei^1;2, Xinxin Guo¹, Jianxin Zhen², Fengxiang Wei³, Shujia Huang⁴, Xiu Qiu⁴

¹Sun Yat-sen University, School of Public Health (Shenzhen), Shenzhen, China; ²Shenzhen Baoan Women’s and Children’s Hospital, Central Laboratory, Shenzhen, China; ³Longgang District Maternity & Child Healthcare Hospital of Shenzhen City, The Genetics Laboratory, Shenzhen, China; ⁴Guangzhou Women and Children’s Medical Center, Division of Birth Cohort Study, Guangzhou, China

Background/Objectives: The gestational period, spanning approximately 40 weeks from fertilization to birth, is pivotal in human reproduction. Monitoring the health of pregnant women and newborns during this period involves systematic prenatal and postpartum examinations, guided by indicators established by the national medical insurance system, collectively termed gestational phenotypes. However, our understanding of the genetic basis of these phenotypes and their intricate relationship with short- and long-term health outcomes remains markedly limited.

Methods: We conducted genome-wide association studies on over a hundred gestational phenotypes in 120,671 Chinese pregnancies, covering anthropometric metrics, non-invasive prenatal testing, and a series of biomarker measurements. Mendelian randomization was employed to explore potential causal relationships between the gestational phenotypes and birth outcomes, as well as long-term health outcomes.

Results: We identified 3,828 genetic loci, including 1,368 novel ones. We observed gestation-specific genetic effects, with a predominant enrichment in hematological pathways. Longitudinal trajectory analyses of repeated measurements of 24 phenotypes revealed significant fluctuations in genetic influences throughout gestational and postpartum periods. Intergenerational and two-sample Mendelian Randomization (MR) analyses combined with East Asian Biobanks suggested extensive causal associations between gestational phenotypes, birth outcomes, and the risk of future chronic diseases.

Conclusion: Our findings provide initial insights into the genetic foundations of human gestational phenotypes and their potential causal associations with short- and long-term health outcomes, laying the foundation for advanced population health during gestation.

Grants: This study received support from the Shenzhen Basic Research Foundation and the National Natural Science Foundation of China. The funders had no role in the study design.

Conflict of Interest: None declared

C20.6 Genome-wide association study using health registers in harmonized Nordic cohorts reveals different genetic architecture between early- and late-onset depression

John Shorter ¹, Joëlle Pasman², Andreas Jangmo³, Siim Kurvits⁴, joonas naamanka⁵, Espen Hagen⁶, Arvid Harder², Jacob Bergstedt², Patrick Sullivan², Iiris Hovatta⁵, Thomas Werge⁷, Martin Tesli³, Kelli Lehto⁴, Yi Lu², Alfonso Buil⁷

¹Roskilde University, Roskilde, Denmark; ²Karolinska Institute, Sweden; ³Folkehelseinstituttet, Norway; ⁴University of Tartu, Tartu, Estonia; ⁵University of Helsinki, Helsinki, Finland; ⁶University of Oslo, Norway; ⁷Research Institute of Biological Psychiatry, Roskilde, Denmark

Consortium: TRYGGVE, iPSYCH, MoBa, FinnGen, Estonian Biobank

Background/Objectives: Major depressive disorder (MDD) is a common and heterogeneous mental disorder. Heterogeneity in symptoms, severity, and response to treatment hampers our ability to provide a biological characterization of MDD. Studying specific subtypes of more homogeneous patient groups may increase our chances of identifying the underlying genetic causes of the disorder and facilitate our efforts of designing targeted treatment strategies.

Methods: We investigated differences in the genetic architecture of early onset (eoMDD) and late onset (loMDD) depression; two MDD subtypes with potential etiological differences. We performed a meta-analysis using nine cohorts from five countries with detailed electronic healthcare registries and accumulated around 46K early onset (eoMDD, ≤25 years old at diagnosis) and 37K late onset (loMDD, ≥50 years at diagnosis) cases to compare the genetic architecture between these two age-related subtypes of MDD.

Results: We found that eoMDD and loMDD shared no genome-wide significant loci (P < 5x10-8), that genes associated with eoMDD have a role in neurodevelopment and suicide, that the SNP-heritability for eoMDD was nearly twice as high as loMDD, the two subtypes only had a modest genetic correlation, and the polygenic score for eoMDD is a significant predictor for suicide.

Conclusion: This study is the first to identify genetic loci differentially associated with early versus late-onset MDD. eoMDD is associated with a more severe clinical profile than loMDD, suggesting prioritization of treating and preventing severe outcomes in individuals with an early onset.

Grants: US NIMH MH123724

Conflict of Interest: None declared

C21.1 De novo variants in the non-coding spliceosomal snRNA gene RNU4-2 are a frequent cause of syndromic neurodevelopmental disorders

Yuyang Chen ^1;2, Ruebena Dawes^1;2, Hyung Chul Kim^1;2, Sarah Stenton^3;4, Susan Walker⁵, Alicia Ljungdahl^6;7, Jenny Lord⁸, Vijay S. Ganesh^3;4;9, Jialan Ma³, Alexandra Martin-Geary^1;2, Gabrielle Lemire^3;4, Elston N. D’Souza^1;2, Shan Dong^6;7, Jamie Ellingford^5;10;11, John Rubenstein⁷, Eirene Markenscoff-Papadimitriou⁷, Sebastian Fica¹², Diana Baralle^13;14, Christel Depienne¹⁵, Daniel MacArthur^16;17, Joanna Howson¹⁸, Stephan Sanders^6;7, Anne O’Donnell-Luria^3;4;19, Nicola Whiffin^1;2;3

¹University of Oxford, Big Data Institute, Oxford, United Kingdom; ²University of Oxford, Centre for Human Genetics, Oxford, United Kingdom; ³Broad Institute of MIT and Harvard, Broad Center for Mendelian Genomics, Cambridge, United States; ⁴Boston Children’s Hospital, Harvard Medical School, Division of Genetics and Genomics, Boston, United States; ⁵Genomics England, London, United Kingdom; ⁶University of Oxford, Department of Paediatrics, Oxford, United Kingdom; ⁷University of California, UCSF Weill Institute for Neurosciences, San Francisco, United States; ⁸University of Sheffield, Sheffield Institute for Translational Neuroscience (SITraN), Sheffield, United Kingdom; ⁹Brigham and Women’s Hospital, Harvard Medical School, Department of Neurology, Boston, United States; ¹⁰Manchester University NHS Foundation Trust, Manchester Centre for Genomic Medicine, Manchester, United Kingdom; ¹¹University of Manchester, Faculty of Biology, Medicines and Health, Manchester, United Kingdom; ¹²University of Oxford, Department of Biochemistry, Oxford, United Kingdom; ¹³University of Southampton, Faculty of Medicine, Southampton, United Kingdom; ¹⁴University Hospital Southampton National Health Service (NHS) Foundation Trust, National Institute for Health Research (NIHR) Southampton Biomedical Research Centre, Southampton, United Kingdom; ¹⁵University Duisburg-Essen, Institute of Human Genetics, University Hospital Essen, Essen, Germany; ¹⁶Garvan Institute of Medical Research and UNSW Sydney, Centre for Population Genomics, Sydney, Australia; ¹⁷Murdoch Children’s Research Institute, Centre for Population Genomics, Melbourne, Australia; ¹⁸Novo Nordisk Research Centre, Human Genetics Centre of Excellence, Oxford, United Kingdom; ¹⁹Massachusetts General Hospital, Harvard Medical School, Center for Genomic Medicine, Boston, United States

Consortium: RNU4-2 consortium

Background/Objectives: Around 60% of individuals with neurodevelopmental disorders (NDD) remain undiagnosed after comprehensive genetic testing, primarily of protein-coding genes. Increasingly, large genome-sequenced cohorts are improving our ability to discover new diagnoses in the non-coding genome.

Methods: Using a cohort of 8,841 probands with genetically undiagnosed NDD in Genomics England (GEL), we identify the non-coding RNA RNU4-2 as a novel syndromic NDD gene. RNU4-2 encodes the U4 small nuclear RNA (snRNA), which is a critical component of the major spliceosome.

Results: We identify an 18 bp region of RNU4-2 mapping to two structural elements in the U4/U6 snRNA duplex that is severely depleted of variation in the general population, but in which we identify heterozygous variants in 119 individuals with NDD across cohorts. This region is significantly enriched for variants in GEL NDD probands compared to individuals in the UK Biobank (OR = 85.8; 95%CI:56.4-131.6; Fisher’s P = 1.84x10^-78). The majority of individuals with NDD (77.3%) have the same highly recurrent single base-pair insertion (n.64_65insT). We estimate that variants in this region explain 0.41% of individuals with NDD. We demonstrate that RNU4-2 is highly expressed in the developing human brain, in contrast to other U4 homologs, supporting RNU4-2’s role as the primary U4 transcript in the brain.

Conclusion: This work underscores the importance of non-coding genes in rare disorders. It will provide a diagnosis to thousands of individuals with NDD worldwide and catalyse development of effective treatments for these individuals.

Grants: due to word count limits, please follow this link (https://docs.google.com/document/d/1aJEbZZKtirtHutC5odTK9Wq40SKkoOSorWJBAYh_7zc/edit?usp=sharing) for grant information.

Conflict of Interest: Yuyang Chen: None declared, Ruebena Dawes: None declared, Hyung Chul Kim: None declared, Sarah Stenton: None declared, Susan Walker: None declared, Alicia Ljungdahl: None declared, Jenny Lord: None declared, Vijay S. Ganesh: None declared, Jialan Ma: None declared, Alexandra Martin-Geary: None declared, Gabrielle Lemire: None declared, Elston N. D’Souza: None declared, Shan Dong: None declared, Jamie Ellingford: None declared, John Rubenstein: None declared, Eirene Markenscoff-Papadimitriou: None declared, Sebastian Fica: None declared, Diana Baralle: None declared, Christel Depienne: None declared, Daniel MacArthur Microsoft, GlaxoSmithKline, Insitro, Overtone Theraputics, Joanna Howson Novo Nordisk, Novo Nordisk, Stephan Sanders BioMarin Pharmaceutical, Anne O’Donnell-Luria PacBio, Congenica, Tome Biosciences, Ono Pharma USA Inc., Nicola Whiffin Novo Nordisk, ArgoBio studio

C21.2 Mutations in the U4 snRNA gene RNU4-2 cause one of the most prevalent monogenic neurodevelopmental disorders

Daniel Greene ^1;2, Chantal Thys³, Els Ortibus^4;5, Andrew Mumford⁶, Kathleen Freson³, Ernest Turro^1;2;7

¹University of Cambridge, Department of Medicine, Cambridge, United Kingdom; ²Icahn School of Medicine at Mount Sinai, Department of Genetics and Genomic Sciences, New York, United States; ³KU Leuven, Department of Cardiovascular Sciences, Leuven, Belgium; ⁴KU Leuven, Department of Development and Regeneration, Leuven, Belgium; ⁵KU Leuven, Paediatric Neurology Department, Leuven, Belgium; ⁶University of Bristol, School of Cellular and Molecular Medicine, Bristol, United Kingdom; ⁷Icahn School of Medicine at Mount Sinai, Mindich Child Health and Development Institute, New York, United States

Background/Objectives: Although 1,427 genes have been identified confidently as aetiological for intellectual disability (ID), most cases remain unexplained. All but nine of the 1,427 known genes are protein-coding.

Methods: To discover non-coding aetiologies of ID, we performed a genetic association analysis comparing rare variants in 41,132 non-coding genes between 4,619 unrelated ID cases and 48,039 unrelated controls in the 100,000 Genomes Project.

Results: The most strongly associated gene was RNU4-2 (log Bayes factor = 64), one of two genes encoding the major spliceosome component U4 snRNA. We implicated variants in 47 cases in two short regions mapping to U4-U6 interactions in the aetiology of a syndrome characterised by ID, microcephaly, short stature, hypotonia, seizures and motor delay. We replicated our finding in three collections, bringing the number of cases to 72. Trio sequencing of 56 cases showed 55 alleles were de novo (one parental mosaic) and one was inherited. RNU4-2 variants explained more cases of neurodevelopmental abnormalities sequenced through the UK’s diagnostic service than any other gene except MECP2. RNA sequencing of patient fibroblasts (but not of blood) revealed abnormal splicing patterns similar to those in a recently discovered spliceosomopathy caused by loss-of-function of the U4 binding partner WBP4.

Conclusion: RNU4-2 mutations cause the most common neurodevelopmental disorder after Rett syndrome. This has remained concealed because the variants are usually de novo and the gene is soft-masked and omitted by exome sequencing. The observed lineage-specific splicing defects underscore the role of spliceosome dysfunction in the aetiologies of neurodevelopmental disorders.

Grants: R01HL161365.

Conflict of Interest: None declared

C21.3 Diagnostic Utility of Genome-wide DNA Methylation Analysis in Genetically Unsolved Developmental and Epileptic Encephalopathies and Refinement of a CHD2 Episignature

Christy LaFlamme ¹, Cassandra Rastin², Soham Sengupta¹, Helen Pennington¹, Sophie Russ-Hall³, Amy Schneider³, Emily Bonkowski¹, Edith Almanza Fuerte¹, Talia Allan³, Methylation Collaborative⁴, University of Washington Center for Rare Disease Research⁵, Sam Berkovic³, Lynette Sadleir⁶, Danny Miller⁷, Ingrid Scheffer³, Bekim Sadikovic², Heather Mefford¹

¹St. Jude Children’s Research Hospital, Center for Pediatric Neurological Disease Research, Memphis, United States; ²Western University, Department of Pathology & Laboratory Medicine, London, Canada; ³The University of Melbourne, Department of Medicine, Parkville, Australia; ⁴Multiple Centers; ⁵University of Washington, Seattle, United States; ⁶University of Otago, Wellington, Department of Paediatrics and Child Health, Wellington, New Zealand; ⁷University of Washington Medical Center, Department of Laboratory Medicine and Pathology, Seattle, United States

Background/Objectives: Sequence-based genetic testing currently identifies causative genetic variants in ~50% of individuals with developmental and epileptic encephalopathies (DEEs). Aberrant changes in DNA methylation are implicated in various neurodevelopmental disorders but remain unstudied in DEEs. Rare epigenetic variations (“epivariants”) can drive disease by modulating gene expression at single loci. DNA methylation changes at many loci across the genome can result in distinct “episignature” biomarkers for monogenic disorders in a growing number of rare diseases.

Methods: Here, we interrogate the diagnostic utility of genome-wide DNA methylation array analysis on peripheral blood samples from 582 individuals with genetically unsolved DEEs who had previously undergone extensive genetic testing.

Results: We identified rare differentially methylated regions (DMRs) and explanatory episignatures to discover causative and candidate genetic etiologies in 12 individuals. We then used long-read sequencing to identify DNA variants underlying rare DMRs, including one balanced translocation, three CG-rich repeat expansions, and four copy number variants. We also identify pathogenic sequence variants associated with episignatures; some had been missed by previous exome sequencing. Although most DEE genes lack known episignatures, the increase in diagnostic yield for DNA methylation analysis in DEEs is comparable to the added yield of genome sequencing. Finally, we refine an episignature for CHD2 using 850K metharray and at higher CpG resolution using bisulfite sequencing to investigate potential insights into CHD2 pathophysiology.

Conclusion: Our study demonstrates the diagnostic yield of genome-wide DNA methylation analysis to identify causal and candidate genetic causes as 2% (12/582) for unsolved DEE cases.

Grants: CURE Epilepsy, AES

Conflict of Interest: Christy LaFlamme: None declared, Cassandra Rastin: None declared, Soham Sengupta: None declared, Helen Pennington: None declared, Sophie Russ-Hall: None declared, Amy Schneider: None declared, Emily Bonkowski: None declared, Edith Almanza Fuerte: None declared, Talia Allan: None declared, Methylation Collaborative: None declared, University of Washington Center for Rare Disease Research: None declared, Sam Berkovic: None declared, Lynette Sadleir L.G.S. receives funding from the Health Research Council of New Zealand and Cure Kids New Zealand. L.G.S. has received research grants and consultancy fees from Zynerba Pharmaceuticals., L.G.S. has received travel grants from Seqirus and Nutricia., L.G.S. is a consultant for the Epilepsy Consortium. L.G.S. has served on Takeda and Eisai Pharmaceuticals scientific advisory panels., Danny Miller D.E.M holds stock options in MyOme., D.E.M is engaged in a research agreement with ONT., D.E.M is on a scientific advisory board at ONT and has received travel support from ONT to speak on their behalf., Ingrid Scheffer I.E.S. may accrue future revenue on pending patent WO61/010176 (filed: 2008): Therapeutic Compound; has a patent for SCN1A testing held by Bionomics Inc and licensed to various diagnostic companies; has a patent molecular diagnostic/theragnostic target for benign familial infantile epilepsy (BFIE) [PRRT2] 2011904493 & 2012900190 and PCT/AU2012/001321 (TECH ID:2012-009)., I.E.S has served as an investigator for Anavex Life Sciences, Cerevel Therapeutics, Eisai, Encoded Therapeutics, EpiMinder Inc, Epygenyx, ES-Therapeutics, GW Pharma, Marinus, Neurocrine BioSciences, Ovid Therapeutics, Takeda Pharmaceuticals, UCB, Ultragenyx, Xenon Pharmaceuticals, Zogenix and Zynerba., I.E.S. has received speaker honoraria from GlaxoSmithKline, UCB, BioMarin, Biocodex, Chiesi, Liva Nova, Nutricia, Zuellig Pharma, Stoke Therapeutics and Eisai; has received funding for travel from UCB, Biocodex, GlaxoSmithKline, Biomarin, Encoded Therapeutics Stoke Therapeutics and Eisai., I.E.S. has served on scientific advisory boards for BioMarin, Chiesi, Eisai, Encoded Therapeutics, GlaxoSmithKline, Knopp Biosciences, Nutricia, Rogcon, Takeda Pharmaceuticals, UCB, Xenon Pharmaceuticals, Cerecin. I.E.S. has consulted for Care Beyond Diagnosis, Epilepsy Consortium, Atheneum Partners, Ovid Therapeutics, UCB, Zynerba Pharmaceuticals, BioMarin, Encoded Therapeutics and Biohaven Pharmaceuticals; and is a Non-Executive Director of Bellberry Ltd and a Director of the Australian Academy of Health and Medical Sciences and the Australian Council of Learned Academies Limited., Bekim Sadikovic B.S. is a shareholder in EpiSign Inc, a company involved in commercialization of EpiSignTM software., Heather Mefford: None declared

C21.4 Pathogenicity mechanism and gene therapy of hereditary spastic paraplegia caused by recurrent UBAP1 variant

Nicolas James Ho¹, Lipeng Ding¹, Min Chen¹, Yu Shan Fung¹, Dexuan Cui¹, Linyan Meng², Shen Gu ¹

¹The Chinese University of Hong Kong, School of Biomedical Sciences, Faculty of Medicine, Hong Kong, Hong Kong; ²Baylor College of Medicine, Department of Molecular and Human Genetics, Houston, United States

Background: Hereditary spastic paraplegia (HSP) is a group of neurological disorders characterized by upper motor neuron (UMN) degeneration, leading to spasticity and weakness in the lower extremities. We and others identified an autosomal dominant HSP subtype associated with the UBAP1 (ubiquitin-associated protein 1) gene.

Methods: Using CRISPR/Cas9, we established human embryonic stem cells (hESCs) and mouse models harboring the recurrent UBAP1 delGA variant found in 80% of our patients. Neural progenitor cells (NPCs) were further differentiated from hESCs. The three mouse models included a precise delGA mutant (Ubap1^delGA/+), a knockout (Ubap1^+/-), and a humanized knock-in expressing delGA UBAP1 (Rosa26^∆UBAP1/+).

Results: Mutant UBAP1 mRNA escaped nonsense-mediated decay, resulting in stable expression of a truncated protein alongside wildtype (WT) UBAP1. Mutant NPCs exhibited accumulation of ubiquitinated proteins and abnormal endosomal dynamics. Ubap1^delGA/+ and Rosa26^∆UBAP1/+ mice demonstrated motor deficits. Single nucleus RNA sequencing analysis identified dysregulated gene expression predominantly in UMNs. In contrast, Ubap1^+/- mice were phenotypically normal, suggesting loss-of-function is unlikely to cause HSP. These models recapitulated the HSP phenotype and demonstrated a dominant negative disease mechanism.

We developed therapeutics to suppress the mutant protein while preserving WT UBAP1. siRNAs reduced mutant protein levels by 75% while maintaining WT UBAP1 expression. Antisense oligonucleotides (ASOs) targeting mutant UBAP1 mRNA showed similar results. We are evaluating these approaches to rescue the HSP phenotype in our preclinical models.

Conclusion: Our study elucidated a dominant negative mechanism underlying UBAP1-HSP and underscored the importance of allele-specific gene-silencing therapy in treating the disease.

GRF/ECS #24101921, NSFC #82202045

Conflict of Interest: None declared

C21.5 Whole genome association testing and prediction in large-scale biobank data

Al Depope¹, Jakub Bajzik¹, Marco Mondelli¹, Matthew Robinson ¹

¹Institute of Science and Technology Austria, Klosterneuburg, Austria

Background/Objectives: The recent release of whole-genome sequence (WGS) data for all UK Biobank participants facilitates investigating the impact of rare variants on complex traits. The common statistical approach of single-marker, or single-gene burden score regression, gives marginal associations that do not account for linkage disequilibrium, and in large sample size high density WGS data, even weak associations that are physically distant from causal variants will be discovered as significant. This limits our understanding of the genetic basis of human traits

Methods: We present a new algorithmic paradigm, gVAMP, that (i) directly fine-maps WGS variants and gene burden scores, conditional on all other measured DNA variations genome-wide, and (ii) provides optimal polygenic risk score prediction. This is done in one-shot, in lightning speed, allowing the analysis of datasets considered impossible before.

Results: On DNA Nexus, it takes just over one day to fine-map many thousands of human height-associated WGS variants and create a polygenic risk score with 48% prediction accuracy. We find 60 genes where rare coding mutations significantly influence phenotype, 76 X-chromosome associations, and thousands of autosomal associations localised to the single-locus level for five additional traits. Additionally, gVAMP outperforms summary statistic polygenic score methods, and outperforms REGENIE for standard association testing in a fraction of the compute time across 13 traits in imputed sequence data.

Conclusion: We showcase how to fine-map variants genome-wide in large sequencing datasets and provide open-source software.

Grants: Lopez-Loreta Prize and SNSF Eccellenza Grant PCEGP3-181181

Conflict of Interest: None declared

C21.6 Single-cell transcriptomics of post-implantation embryos: unveiling aneuploidy effects and lineage dynamics

Sheila Kwok ^1;2, Lilach Marom Haham², Stewart Russell², Hanna Balakier², Siwei Chen², Manuel Viotti³, Svetlana Madjunkova^2;4, Clifford Librach^1;2;5;6;7

¹University of Toronto, Department of Physiology, Toronto, Canada; ²CReATe Fertility Centre, Toronto, Canada; ³Zouves Fertility Center, Foster City, United States; ⁴University of Toronto, Department of Laboratory Medicine and Pathobiology, Toronto, Canada; ⁵University of Toronto, Department of Obstetrics and Gynaecology, Toronto, Canada; ⁶University of Toronto, Institute of Medical Science, Toronto, Canada; ⁷Sunnybrook Health Sciences Centre, Toronto, Canada

Background/Objectives: Embryo aneuploidy is clinically associated with pregnancy abnormalities. However, healthy live births resulting from diploid-aneuploid mosaic embryos indicate the early presence of aneuploid cells might not necessarily compromise embryo developmental competence. Here, we aim to investigate the global ploidy-specific transcriptomic signatures in post-implantation human embryos, revealing the molecular mechanisms determining aneuploid cell fates and survival.

Methods: Donated preimplantation euploid, aneuploid and mosaic embryos were cultured from day 5-6. On day 9-11, embryos were dissociated into single cells and pooled for scRNA-seq (10X Chromium). Sequencing results were analyzed using Freemuxlet to identify embryo-of-origin, Seurat for clustering and InferCNV to determine cell ploidy. Differential expression analyses were performed to identify enriched genetic markers and pathways.

Results: scRNA-seq data from 3159 cells identified five primary lineages, including epiblast, hypoblast, cytotrophoblast, syncytiotrophoblast and extravillous trophoblast. Differential expression analyses showed different intra-embryo environments facilitate distinct gene expression patterns related to energy expenditure, proteostasis, and cellular repair processes in the aneuploid cells. In homogeneous aneuploid embryos, temporary adaptive stress responses were observed, indicated by upregulated unfolded protein response (e.g. ATF4). In mosaic embryos, key machinery contributing to cell fitness is disrupted in aneuploid cells, including glycolytic process (e.g. HIF1A) and DNA damage repair (e.g. DDIT4).

Conclusion: Ploidy-specific transcriptomic analyses pointed to a diminished survival advantage in the aneuploid cells of post-implantation mosaic embryos. Differential fitness levels in the environment potentially induced self-correction by aneuploid cell depletion. These results provide the foundation for interrogating developmental competence of mosaic embryos, potentially expanding the range of viable embryos suitable for IVF treatments.

Grants:

Conflict of Interest: None declared

C22.1 Epigenome and transcriptome changes in Kabuki syndrome 1 iPSCs and neuronal progenitors

Sara Cuvertino ¹, Evgenii Martirosian¹, Peiwen Cheng², Ian Donaldson³, Terence Garner³, Adam Stevens⁴, Andrew Sharrocks⁵, Susan Kimber⁶, Siddharth Banka⁶

¹University of Manchester, Division of Evolution and Genomic Sciences, Manchester, United Kingdom; ²University of Manchester, Bioinformatics and Genomic Technologies Core Facilities, Manchester, United Kingdom; ³University of Manchester, Division of Developmental Biology & Medicine, Manchester, United Kingdom; ⁴University of Manchester, Division of Molecular and Cellular function, Manchester, United Kingdom; ⁵University of Manchester, Division of Cell Matrix Biology and Regenerative Medicine, Manchester, United Kingdom; ⁶Manchester University Foundation NHS Trust, Manchester Centre for Genomic Medicine, Manchester, United Kingdom

Background/Objectives: Kabuki syndrome type-1 (KS1) is a neurodevelopmental disorder caused by loss-of-function variants in KMT2D which encodes a H3K4 methyltransferase. The mechanisms underlying neurodevelopmental problems in KS1 are unknown.

Methods: Three control and three patient-derived induced pluripotent stem cell (iPSC) lines were differentiated into neuronal progenitors (NPs) and cortical neurons (CNs). RT-PCR, western blots, RNASeq and H3K4me1 ChIPSeq were performed at different stages of differentiation.

Results: KS1 iPSCs showed significantly lower expression of functional KMT2D transcript and protein. We also found significantly lower H3K4me1 and H3K4me2 levels but not significant changes for H3K4me3 levels.

ChIPseq in iPSCs, NPs and CNs showed that the H3K4me1 loss in KS1 is region specific and primarily affects annotated enhancer regions. RNASeq in iPSCs revealed differentially expressed genes (DEGs) related to several KS1 phenotypes such as extracellular structure organization, heart morphogenesis, ear and axon development. In contrast, RNASeq at NP and CNs showed DEGs related to axonal development and ion transport.

Comparison of ChIPSeq and RNASeq data revealed a statistically significant enrichment of differentially downregulated genes in regions corresponding to enhancer regions with H3K4me1 loss across the three differentiation stages.

Motif analysis showed that the binding sites for SUZ12, a subunit of Polycomb Repressive complex 2, were most enriched in DEGs at all three stages of differentiation.

Conclusion: KS1 is characterised by epigenomic and transcriptomic dysregulation at all stages of neuronal differentiation. We present a disease relevant human cellular model that can be used for high throughput drug screening for KS1.

Grants: Great Ormond Street Hospital (GOSH) charity, NewLife

Conflict of Interest: None declared

C22.2 Single-cell genomics and regulatory networks for 388 human brains

Matthew Jensen ^1;2, Declan Clarke^1;2, Cagatay Dursun^1;2, Prashant Emani^1;2, Chirag Gupta^3;4, Cheyu Lee⁵, Jason Liu^1;2, Shaoke Lou^1;2, Ran Meng^1;2, Jonathan Warrell^1;2, Siwei Xu⁵, Yuhang Chen^1;2, Zhiyuan Chu^1;2, Timur Galeev^1;2, Michael Gancz^1;2, Ahyeon Hwang⁵, Pengyu Ni^1;2, Yan Xia^1;2, Xiao Zhou^1;2, Trygve Bakken⁶, Tanima Chatterjee^1;2, Sophie Gaynor^7;8, Ella Henry^1;2, Ao Huang^1;2, Yunzhe Jiang^1;2, Ting Jin^3;4, Yunyang Li^1;2, Shuang Liu⁴, Shaojie Ma⁹, Eric Nguyen^1;2, Ananya Rajagopalan^1;2, Tiernon Riesenmy^1;2, Brie Wamsley¹⁰, Gaoyuan Wang^1;2, Andy Yang^1;2, Suchen Zheng^1;2, Michael Gandal^11;12, Gabriel Hoffman^13;14, Donghoon Lee^13;14, Ed Lein⁶, Panos Roussos^13;14, Nenad Šestan⁹, Zhiping Weng¹⁵, Kevin White¹⁶, Hyejung Won¹⁷, Matthew Girgenti¹⁸, Jing Zhang⁵, Daifeng Wang^3;4, Daniel Geschwind¹⁰, Mark Gerstein^1;2

¹Yale University, Program in Computational Biology and Bioinformatics, New Haven, United States; ²Yale University, Department of Molecular Biophysics and Biochemistry, New Haven, United States; ³University of Wisconsin-Madison, Department of Biostatistics and Medical Informatics, Madison, United States; ⁴University of Wisconsin-Madison, Waisman Center, Madison, United States; ⁵University of California, Irvine, Department of Computer Science, Irvine, United States; ⁶Allen Institute of Brain Science, Seattle, United States; ⁷Tempus Labs, Inc., Chicago, United States; ⁸Cornell College, Department of Biology, Mount Vernon, United States; ⁹Yale University, Department of Neuroscience, New Haven, United States; ¹⁰University of California, Los Angeles, Department of Neurology, David Geffen School of Medicine, Los Angeles, United States; ¹¹University of Pennsylvania, Department of Psychiatry, Perelman School of Medicine, Philadelphia, United States; ¹²The Children’s Hospital of Philadelphia, Lifespan Brain Institute, Philadelphia, United States; ¹³Icahn School of Medicine at Mount Sinai, Department of Psychiatry, New York, United States; ¹⁴Icahn School of Medicine at Mount Sinai, Department of Genetics and Genomic Science, New York, United States; ¹⁵UMass Chan Medical School, Worcester, United States; ¹⁶National University of Singapore, Yong Loo Lin School of Medicine, Singapore, Singapore; ¹⁷University of North Carolina at Chapel Hill, Chapel Hill, United States; ¹⁸Yale University, Department of Psychiatry, New Haven, United States

Consortium: PsychENCODE Consortium

Background/Objectives: Single-cell genomics offers a powerful means to understand the role of variants in gene expression and regulation, especially for the diverse cell types in the brain. Using these approaches, we can refine our understanding of how variants and gene regulation affect brain phenotypes, including neuropsychiatric disorders. However, population-scale cohorts with a wide range of brain disorders and traits are needed to infer significant associations between variants, regulatory elements, and expression, and to develop comprehensive models of gene regulation at the single-cell level.

Methods: Using a harmonized cell-typing scheme, we uniformly processed genotype, expression, and chromatin accessibility sequencing data in >2.8M single nuclei from the pre-frontal cortex of 388 adults with brain-related disorders (e.g., autism spectrum disorder, Alzheimer’s disease) and controls.

Results: We identified population and cell-type variation in expression and chromatin across 28 cell types for multiple gene families, including neurotransmitters and neuro-related drug targets. We found >550K cell-type-specific regulatory elements and >1.4M single-cell eQTLs, from which we built cell-type regulatory and cell-to-cell communication networks. These networks exhibit extensive cell-type-specific regulatory changes across traits, including enrichment of TF motifs during aging, and disorders, such as altered WNT and FGF signaling in schizophrenia and bipolar disorder. Using these networks, we constructed an integrative model to accurately impute single-cell expression, quantify disorder risk conditional on genotypes, and simulate perturbations, which prioritized ~250 disease-risk genes and drug targets with their associated cell types.

Conclusion: Our population-scale single-cell resource for the human brain can facilitate discovery of precision medicine approaches for neuropsychiatric disorders.

Grants: Funding provided by NIMH (U01MH116492).

Conflict of Interest: Matthew Jensen: None declared, Declan Clarke: None declared, Cagatay Dursun: None declared, Prashant Emani: None declared, Chirag Gupta: None declared, Cheyu Lee: None declared, Jason Liu: None declared, Shaoke Lou: None declared, Ran Meng: None declared, Jonathan Warrell: None declared, Siwei Xu: None declared, Yuhang Chen: None declared, Zhiyuan Chu: None declared, Timur Galeev: None declared, Michael Gancz: None declared, Ahyeon Hwang: None declared, Pengyu Ni: None declared, Yan Xia: None declared, Xiao Zhou: None declared, Trygve Bakken: None declared, Tanima Chatterjee: None declared, Sophie Gaynor Employee of Tempus Labs, Ella Henry: None declared, Ao Huang: None declared, Yunzhe Jiang: None declared, Ting Jin: None declared, Yunyang Li: None declared, Shuang Liu: None declared, Shaojie Ma: None declared, Eric Nguyen: None declared, Ananya Rajagopalan: None declared, Tiernon Riesenmy: None declared, Brie Wamsley: None declared, Gaoyuan Wang: None declared, Andy Yang: None declared, Suchen Zheng: None declared, Michael Gandal PI for PsychENCODE grants (NIMH), Gabriel Hoffman PI for PsychENCODE grants (NIMH), Donghoon Lee PI for PsychENCODE grants (NIMH), Ed Lein PI for related BICCN grants (NIMH), Panos Roussos PI for PsychENCODE grants (NIMH), Nenad Šestan PI for PsychENCODE grants (NIMH), Zhiping Weng PI for PsychENCODE grants (NIMH), Kevin White PI for PsychENCODE grants (NIMH), Hyejung Won PI for PsychENCODE grants (NIMH), Matthew Girgenti PI on related grants (NIMH), Jing Zhang PI for PsychENCODE grants (NIMH), Daifeng Wang PI for PsychENCODE grants (NIMH), Daniel Geschwind PI for PsychENCODE grants (NIMH), Mark Gerstein PI for PsychENCODE grants (NIMH)

C22.3 Local heterochromatin mitigates the impact of a transposable element insertion causing a mendelian neurodegenerative disorder

Vivien Horvath ¹, Raquel Garza¹, Marie Jonsson¹, Pia Johansson¹, Anita Adami¹, Georgia Christoforidou¹, Ofelia Karlsson¹, Laura Castilla-Vallmanya¹, Patricia Gerdes¹, Ninoslav Pandiloski¹, Christopher Douse¹, Johan Jakobsson¹

¹Lund University/Lund Stem Cell Center, Experimental Medical Science, Lund, Sweden

Background/Objectives: X-Linked Dystonia-Parkinsonism (XDP) is a mendelian neurodegenerative disorder. Recently, a polymorphic transposable element (TE) insertion in the 32nd intron of the TAF1 gene has been identified as the genetic factor responsible for this disease. The XDP-TE is associated with TAF1 mis-regulation, but the mechanisms behind this phenomenon remain elusive. We hypothesize that repressive epigenetic marks on the XDP-TE are key players in this process. Thus, here we aim to dissect the molecular intricacies that keep the XDP-TE at bay and identify how it triggers aberrant TAF1 expression, ultimately leading to XDP.

Methods: Leveraging advanced sequencing techniques and XDP patient-derived iPSCs and neural progenitor cells, we employed CUT&RUN and Oxford Nanopore Sequencing to identify epigenetic marks on the XDP-TE. To illuminate what factors establish these marks, and their effect on gene expression, we did CRISPR inhibition of various candidate genes coupled with RNA sequencing.

Results: We demonstrate that ZNF91 - a TE-binding KRAB-Zinc Finger Protein - establishes H3K9me3 and DNA methylation over the XDP-TE in a cell type specific manner in patient derived cells. Moreover, removal of DNA methylation, or both H3K9me3 and DNA methylation, severely aggravates the XDP molecular phenotype, causing a reduced TAF1 expression and increased intron retention.

Conclusion: Our study unveils how a polymorphic TE results in XDP and highlights DNA methylation as potential therapeutic target. Moreover, this work underscores the significance of studying polymorphic TEs as disease triggers, with implications for various disorders.

Grants: Collaborative Center for XDP, Torsten och Elsa Segerfalk Foundation

Conflict of Interest: None declared

C22.4 Dissecting the pleiotropic effects of MORC2 mutations through multi-omics analysis

Fatemeh Peymani ^1;2, Tomohiro Ebihara^2;3, Dmitrii Smirnov¹, Felix Distelmaier⁴, Daniele Ghezzi⁵, Robert Kopajtich^1;2, Thomas Klopstock⁶, Costanza Lamperti⁵, Kei Murayama^7;8, Fang Fang⁹, Agnès Rötig¹⁰, Christian Schlein¹¹, Polina Tsygankova¹², Aysylu Murtazina¹², Thomas Meitinger¹, Masaru Shimura^2;8, Holger Prokisch^1;2

¹Technical University of Munich, School of Medicine, Institute of Human Genetics, Munich, Germany; ²Helmholtz Munich, Computational Health Center, Institute of Neurogenomics, Oberschleißheim, Germany; ³Chiba Children’s Hospita, Department of Neonatology,, Funabashi, Japan; ⁴Heinrich Heine University Düsseldorf, Düsseldorf, Germany; ⁵Istituto Neurologico “Carlo Besta” | Fondazione IRCCS, Milano, Italy; ⁶Friedrich-Baur-Institut am LMU Klinikum, Department of Neurology, München, Germany; ⁷Juntendo University, Diagnostics and Therapeutics of Intractable Diseases, Intractable Disease Research Center, Graduate School of Medicine, Bunkyo City, Japan; ⁸Chiba Children’s Hospital, Department of Metabolism, Chiba, Japan; ⁹Beijing Children’s Hospital, Department of Neurology, China; ¹⁰Université Paris Descartes, Institut IMAGINE, Paris, France; ¹¹University Medical Center Hamburg-Eppendorf, Institute of Human Genetics, Hamburg, Germany; ¹²Research Centre for Medical Genetics, Moscow, Russian Federation

Consortium: GENOMIT consortium

Background/Objectives: Variants in the epigenetic silencing factor MORC2 are linked to a spectrum of neurological disorders, encompassing Charcot-Marie-Tooth (CMT), Leigh-spectrum syndrome (LSS), Cockayne-like syndrome (CLS) and DIGFAN. The molecular mechanism underlying diverse clinical manifestations of MORC2-disease remains largely unexplored. We leveraged multi-omics to dissect the phenotypic spectrum.

Methods: We diagnosed 30 patients with de novo mutations in MORC2 and performed a multi-omics comprising RNA-seq, proteomics, and DNA methylation from blood (n = 13) and fibroblasts (n = 13). Epigenome-wide, transcriptome-wide, and proteome-wide association studies were undertaken using >400 diseased and >20 healthy controls for each omics layer. Outliers were called using OUTRIDER. SVM classifiers were trained to establish diagnostic signatures.

Results: Amongst 100 disorders, we identified MORC2-specific signatures in DNA methylation and, for the first time, in transcriptome. MORC2-specific episignature were independently discovered by outliers and EWAS, while RNA-signature were only detected by outliers. Both signatures were driven by the same genes; hypermethylation caused lower expression. Episignature was conserved across tissues. Proteomics confirmed functional consequences of dysregulated genes. Interestingly, among the MORC2-specific outliers, we identified genes associated with LSS, CMT, CSL, and DIGFAN. The level of epigenetic silencing and gene expression of those genes strongly correlates with phenotypic manifestation. This is exemplified by hypermethylation of respiratory chain subunits, reduced expression, lower levels of protein complexes, and diminished enzymatic activity in LSS patients.

Conclusion: We demonstrated the power of multi-omics analysis in deciphering the hitherto unexplained phenotypic spectrum of MORC2-associated disorders. We provide the first example in which molecular signatures are not only diagnostic biomarkers but explain underlying pathomechanisms.

Grants: DMB-1805-0002

Conflict of Interest: None declared

C22.5 Investigating the role of SIRT6 in neurodevelopmental delay using Drosophila Melanogaster as an animal model

Clarissa Rocca ¹, Gabriel Aughey¹, Reza Maroofian¹, Henry Houlden¹, James Jepson¹

¹UCL Queen Square Institute of Neurology, London, United Kingdom

Background/Objectives: NAD-dependent histone deacetylase sirtuin-6 is an enzyme encoded by SIRT6 which is localised in the nucleus and plays a crucial role in chromatin homeostasis regulation, DNA signalling and repair, and is involved in glucose and lipid metabolism. Although the role of SIRT6 in human development remains unclear, studies suggest an association with neurodevelopmental processes. We identified SIRT6 as a novel human disease gene and investigated the effect of protein knock-down in Drosophila Melanogaster.

Methods: Exome analysis was performed on ~35,000 cases. We used the GAL4-UAS system to express the interfering shRNA (RNAi). First, we altered protein expression with the global driver actinGAL4. With the same approach, we then looked at neural stem cells (NSC) and post-mitotic effects. To investigate cell growth and division in specific clones, we employed the FLP/FRT system.

Results: We identified six different biallelic missense mutations in 20 patients from 10 unrelated families. The cases described show severe neurodevelopmental delay, including intellectual disability (n = 8) and focal and generalised seizures (n = 6). Knocking down SIRT6 in adult neurons led to decreased movement in experimental flies compared to controls. In contrast, protein knockdown in neuroblasts had a lethal effect on experimental flies.

Conclusion: In this study, targeted knockdown of the protein in adult neurons resulted in a notable reduction in movement without significant impact on overall viability. Interestingly, lethality occurred consequently to protein knockdown in NSC, a crucial component in the early and critical stages of neurodevelopment, therefore implying an essential role of SIRT6 protein in the neuronal development of Drosophila melanogaster.

Grants:

Conflict of Interest: None declared

C22.6 Long-read genome sequencing improves diagnostic sensitivity and aids gene identification

Karla Figueroa¹, Elena Buena-Atienza², Joohyun Park², Sharan Paul¹, Mandi Gandelman¹, Neseebullah Kakar³, Marc Sturm², Gunnar Schmidt⁴, Nicolas Casadei², Yorck Hellenbroich³, Jelena Pozojevic³, Kristian Händler³, Simone Zittel-Dirks⁵, Dagmar Timmann-Braun⁶, Friedrich Erdlenbruch⁶, Claudia Dufke², Lara Stühn², Arnulf Koeppen⁷, Thomas Klopstock⁸, Florian Kraft⁹, Martin Zenker¹⁰, Nadja Ehmke¹¹, Bernd Auber⁴, Ingo Kurth⁹, Malte Spielmann³, Stephan Ossowski², Stefan M. Pulst¹, Olaf Riess^2;12, Tobias Haack ^2;12

¹University of Utah, Department of Neurology,, Salt Lake City, United States; ²University of Tübingen, Institute of Medical Genetics and Applied Genomics, Tübingen, Germany; ³Institute of Human Genetics, Lübeck, Germany; ⁴University of Hannover, Institute of Human Genetics, Hannover, Germany; ⁵University Medical Center Hamburg-Eppendorf, Department of Neurology, Hamburg, Germany; ⁶University of Duisburg-Essen, Department of Neurology and Center for Translational Neuro- and Behavioral Sciences (C-TNBS), Essen; ⁷Veterans Affairs Medical Center, Albany, United States; ⁸University Hospital of Ludwig- Maximilians-Universität München, Department of Neurology with Friedrich-Baur-Institute, Munich, Germany; ⁹RWTH Aachen University Hospital, Institute for Human Genetics and Genomic Medicine, Aachen, Germany; ¹⁰Otto-von-Guericke University, Institute of Human Genetics, Magdeburg, Germany; ¹¹Charité Berlin, Institute of Human Genetics, Berlin; ¹²University Tübingen, Center for rare diseases, Tübingen

Consortium: lonGER

Background: Clinical long-read genome sequencing (LR-GS) is expected to narrow the diagnostic gap and to identify novel disease genes and mechanisms. We aimed to determine the diagnostic accuracy of LR-GS and to discover the molecular correlate of SCA4, which has escaped identification for more than 25 years.

Methods: LR-GS was benchmarked using reference samples and various disease entities prone to genomic variation that are difficult to detect, including alterations in low complexity regions, pseudogenes, structural variation, repeat expansions, and imprinting disorders. Combined linkage and LR-GS studies were performed on a large Utah SCA4 pedigree, followed by matchmaking and cellular studies.

Results: LR-GS showed advantages in the diagnostic sensitivity across all types of genomic variation, apart from small insertions and deletions, with methylation calling as an additional readout. We identified a heterozygous GGC-repeat expansion (RE) in ZFHX3 as the likely molecular cause of SCA4. A query of 6,495 genomes and exchange with collaborators identified the RE in a further 7 pedigrees. Ultra-rare DNA variants near the RE indicated a common distant founder event in Sweden. The SCA4 phenotype extends from cerebellar and sensory ataxia, to autonomic dysfunction and chronic cough. Abnormal autophagy was observed in fibroblast and iPS cells, which normalized upon siRNA-mediated ZFHX3 knockdown.

Conclusion: LR-GS has the potential to serve as an all-in-one genetic test covering all types of (epi)genetic variation. The coding GGC-RE in an extremely GC-rich region was undetectable by short-read GS and demonstrates the power of LR-GS for variant detection. SCA4 adds to neuronal intranuclear inclusion disease and FXTAS as a polyglycinopathy.

Conflict of Interest: Karla Figueroa: None declared, Elena Buena-Atienza: None declared, Joohyun Park: None declared, Sharan Paul: None declared, Mandi Gandelman: None declared, Neseebullah Kakar: None declared, Marc Sturm: None declared, Gunnar Schmidt: None declared, Nicolas Casadei: None declared, Yorck Hellenbroich: None declared, Jelena Pozojevic: None declared, Kristian Händler: None declared, Simone Zittel-Dirks: None declared, Dagmar Timmann-Braun: None declared, Friedrich Erdlenbruch: None declared, Claudia Dufke: None declared, Lara Stühn: None declared, Arnulf Koeppen: None declared, Thomas Klopstock: None declared, Florian Kraft: None declared, Martin Zenker: None declared, Nadja Ehmke: None declared, Bernd Auber: None declared, Ingo Kurth: None declared, Malte Spielmann: None declared, Stephan Ossowski Received reimbursement for travel expenses and payment for conference presentations from Illumina Inc. and Oxford Nanopore Technologies, Stefan M. Pulst: None declared, Olaf Riess: None declared, Tobias Haack: None declared

C23.1 What does a consent conversation for whole genome sequencing look like? An observational study of consent appointments in the NHS Genomic Medicine Service in England

Holly Ellard ^1;2, Celine Lewis^1;2, Huda Alfardus^1;2

¹Cardiff University, Centre for Medical Education, Cardiff, United Kingdom; ²UCL Great Ormond Street Institute of Child Health, Population, Policy and Practice Department, London, United Kingdom

Background/Objectives: Consent to whole genome sequencing (WGS) through the NHS Genomic Medicine Service follows a new ‘patient choice’ model, covering clinical testing and an invitation to the National Genomic Research Library (NGRL). Little is known about what the consent conversation looks like in practice. This study analysed the structure and content of WGS consent appointments.

Methods: Participants were healthcare professionals (HCPs) (n = 20) and parents (n = 42) of paediatric patients being offered WGS during 26 consent appointments across seven NHS Trusts. Consent appointments were audio-recorded and analysed using deductive content analysis.

Results: Consent conversations took an average of 13 minutes 59 seconds, with administrative tasks taking an average additional 5 minutes 18 seconds (total average 19 minutes 17 seconds). An average of 29% of the conversation was dedicated to the various result types from WGS, 29% to the NGRL, 17% to incidental findings, and 15% to biological concepts. HCPs frequently discussed the possibility of receiving an uncertain result (25/26) or no diagnosis (24/26), but less commonly discussed the risk of uncovering non-paternity (12/26) or the potential for reanalysis (12/26). Parents asked questions around the logistics of sampling (8/26), testing process (6/26), and data security (3/26). Health-related incidental findings were occasionally the source of questions (2/26) and fears (1/26). Whilst parents expressed few concerns, HCPs did not always ask questions to elicit their thoughts, feelings, and values (12/26).

Conclusion: Parents and patients may benefit from further opportunities to explore their views and concerns during the consent conversation.

Grants: NIHR Advanced Fellowship Grant (NIHR300099)

Conflict of Interest: None declared

C23.2 Are we doing a good job of supporting parents who are making decisions about whole genome sequencing? A cross-sectional survey study conducted in the NHS Genomic Medicine Service

Celine Lewis¹, Ria Patel ², Bettina Friedrich¹

¹UCL GOS Institute of Child Health, Population, Policy and Practice, London; ²University College London, Medical School, London, United Kingdom

Background/Objectives: Whole genome sequencing (WGS) for paediatric rare disease diagnosis is now available as a first-line test for certain clinical indications in the Genomic Medicine Service (GMS). We conducted a quantitative survey comprising validated measures and bespoke questions to understand decision-making processes. Surveys were sent to parents of children offered WGS within 4 weeks of their appointment.

Methods: 1366 surveys were sent out across 7 NHS Trusts in England; 393 completed surveys were returned (29% response rate). Descriptive and inferential statistics were conducted.

Results: Parents were satisfied with their WGS appointment (mean =24.47; range 0-28). Parents had a positive attitude towards WGS (mean = 18.17; range 0-20), although those who had discussed WGS with a genetic counsellor/genomic associate had significantly more positive attitudes than those seen by genetic consultants (p < 0.001). Decisional conflict was low (mean = 20; range 0-64). Regarding knowledge of WGS, the higher parents scored, the less conflicted they felt about their decision (p = 0.003). There was a significant correlation between parents who had spent longer looking for a diagnosis (p = 0.015), had previously undergone genetic testing (p = 0.004), were White (p = 0.003), and had higher formal educational qualifications (p = 0.005) with higher WGS knowledge scores. Type of health professional discussing WGS was not associated with improved knowledge.

Conclusion: Low decisional conflict, good knowledge, and positive attitudes (coupled with test uptake) are indicators of informed decision-making. Our results provide nuance around those parents who might need additional support. Further work to understand how to improve knowledge of WGS is important.

Grants: Celine Lewis: NIHR Advanced Fellowship Grant (NIHR300099)

Conflict of Interest: None declared

C23.3 Ethical implementation of automated genomic reanalysis: Patient, professional, and public perspectives

Danya Vears ^1;2;3;4, Fiona Lynch^1;3, Jack Harrison², Emily King¹, Savio Nona¹, Zornitza Stark^2;5

¹Murdoch Children’s Research Institute, Biomedical Ethics Research Group, Parkville, Australia; ²University of Melbourne, Department of Paediatrics, Parkville, Australia; ³University of Melbourne, Melbourne Law School, Carlton, Australia; ⁴KU Leuven, Centre for Biomedical Ethics and Law, Leuven, Belgium; ⁵Victorian Clinical Genetics Service, Parkville, Australia

Background/Objectives: Reanalysis of undiagnosed patients‘ genomic data two years after the initial test leads to a 10-15% increase in diagnostic yield. Yet reanalysis is far from routine and many patients miss out due to lack of workforce capacity and haphazard practices required to trigger this process. Automation of reanalysis using artificial intelligence (AI) is the solution proposed to provide routine reanalysis to all undiagnosed patients. However, little is known about how different stakeholders perceive this proposal.

Methods: To address this, we conducted online focus groups with members of the Australian public, patients/their carers, and genetics professionals, exploring their perspectives on the practical and ethical challenges, as well as potential solutions, for implementing automated reanalysis into clinical care.

Results: Seventy-eight stakeholders (23 patients/carers, 34 members of the public, 21 genetics professionals) participated across 20 focus groups. Overall, participants from all stakeholder groups were positive about using automated systems for reanalysis. They highlighted benefits relating to equity of access to the possibility of receiving a diagnosis and reducing the burden placed on patients to be responsible for triggering reanalysis. Genetics professionals raised concerns about how to obtain reconsent and the moral duty to return results to patients lost to follow up. In contrast, patients had mixed opinions regarding the need for reconsent and the public worried about who was accountable if results were inaccurate.

Conclusion: Our findings will be used to guide implementation of automated reanalysis, and also develop policies and educate the public about use of AI in genomics more broadly.

Grants: Australian Government MRFF GHFM APP2008820.

Conflict of Interest: Danya Vears Grant declared, Fiona Lynch: None declared, Jack Harrison: None declared, Emily King Grant declared, Savio Nona Grant declared, Zornitza Stark Grant declared

C23.4 The development and use of polygenic scores for social traits: Biobank participants’ perspectives

Margaret Waltz ¹, Kristine Kuczynski¹, Shawneequa Callier², R. Jean Cadigan¹

¹University of North Carolina at Chapel Hill, Chapel Hill, United States; ²George Washington University, Washington, United States

Background/Objectives: Using large datasets from biobanks, polygenic scores (PGS) are being created for a variety of social traits. These scores allow researchers to study gene-environment interactions, but worries abound about their uses in commercial and social settings. Scant research examines the perspectives of biobank participants, whose data could be used to develop PGS for social traits.

Methods: We conducted a survey with over 400 U.S. based biobank participants and five follow-up focus groups exploring views on various PGS for social traits and their potential uses.

Results: We will focus on respondents’ views of PGS for three traits, educational attainment, aggression, and obesity, as well as their possible uses outside of research. Survey respondents revealed mixed feelings about PGS for these traits, particularly if used to help determine which embryos to use in in vitro fertilization. In focus groups, some could understand the desire to select for traits in the otherwise uncontrollable realm of parenting. But others noted fears of scores being used to stigmatize, discriminate, or eliminate certain traits from the population, especially since scores indicate likelihood, not guarantees, and may be based on problematic measures, like body mass index. As a result, respondents questioned whether they would want their data to be used to develop PGS for social traits.

Conclusion: We will discuss the ethical implications of these findings for biobanks when handling potential researcher requests for specimens as well as the merits of engaging with or potentially reconsenting biobank participants.

Grants: U.S. National Human Genome Research Institute (#R01HG012402)

Conflict of Interest: None declared

C23.5 Emerging issues that call for more inclusive strategies in engaging underrepresented groups regarding human germline gene editing

Diewertje Houtman ¹, Wendy Geuverink^2;3, Joosje Kist¹, Isabel Retel Helmrich¹, Martina Cornel^2;4, Michelle Kasprzak⁵, Lidewij Henneman^2;3, Sam Riedijk¹

¹Erasmus Medical Center, Department of Clinical Genetics, Rotterdam, Netherlands; ²Amsterdam UMC, location Vrije Universiteit Amsterdam, Department of Human Genetics, Amsterdam, Netherlands; ³Amsterdam Reproduction and Development Research Institute, Amsterdam, Netherlands; ⁴Amsterdam Public Health Research Institute, Amsterdam, Netherlands; ⁵Rotterdam University of Applied Sciences, Willem de Kooning Academy, Rotterdam, Netherlands

Consortium: De DNA dialogen Consortium

Background/Objectives: The methods, conditions and values for public engagement regarding human germline gene editing (HGGE) are increasingly discussed. The WHO framework for governance report on human genome editing (2021) states that ‘specific strategies are needed to engage traditionally underrepresented groups, such as indigenous peoples, minority ethnic groups or faiths, or specific patient groups’. In the Dutch DNA dialogues project (2022-2026), we have started to operationalize, identify and engage underrepresented groups in dialogues about HGGE, uncovering issues that call for more inclusive strategies.

Methods: We work together with societal partners to operationalize and identify underrepresented groups and to co-create tailored dialogue strategies that employ theatre and the qualitative research method Photovoice. We conducted 14 dialogues, including individuals with genetic conditions, physical disabilities, and those identified as being difficult to reach for scientists, and analysed these using thematic analysis.

Results: In preparing and analysing the dialogues, several issues emerged: (1) Our current informed consent procedures were unfit for underrepresented groups. (2) Islamic participants were hesitant to engage in a dialogue when they were unsure about the views of their religious leaders on this new technology. (3) Experiences during the COVID-19 pandemic severely and negatively impacted participants trust in science and governance of new technologies.

Conclusion: To engage underrepresented groups in dialogue and decision-making about new disruptive technologies, researchers and engagement practitioners need more inclusive informed consent procedures, a religious knowledge base in addition to a scientific knowledge base, and to make distrust in science and governance part of the conversation.

Grants: [NWO/NWA.1389.20.075]

Conflict of Interest: Diewertje Houtman As of the 1st September 2022, the authors are involved in a Netherlands Consortium “Public Realm Entrance of Human Germline Gene Editing” funded by The Netherlands Organization for Scientific Research (NWO NWA.1389.20.075)., Wendy Geuverink As of the 1st September 2022, the authors are involved in a Netherlands Consortium “Public Realm Entrance of Human Germline Gene Editing” funded by The Netherlands Organization for Scientific Research (NWO NWA.1389.20.075)., Joosje Kist As of the 1st September 2022, the authors are involved in a Netherlands Consortium “Public Realm Entrance of Human Germline Gene Editing” funded by The Netherlands Organization for Scientific Research (NWO NWA.1389.20.075)., Isabel Retel Helmrich As of the 1st September 2022, the authors are involved in a Netherlands Consortium “Public Realm Entrance of Human Germline Gene Editing” funded by The Netherlands Organization for Scientific Research (NWO NWA.1389.20.075)., Martina Cornel As of the 1st September 2022, the authors are involved in a Netherlands Consortium “Public Realm Entrance of Human Germline Gene Editing” funded by The Netherlands Organization for Scientific Research (NWO NWA.1389.20.075)., Michelle Kasprzak As of the 1st September 2022, the authors are involved in a Netherlands Consortium “Public Realm Entrance of Human Germline Gene Editing” funded by The Netherlands Organization for Scientific Research (NWO NWA.1389.20.075)., Lidewij Henneman As of the 1st September 2022, the authors are involved in a Netherlands Consortium “Public Realm Entrance of Human Germline Gene Editing” funded by The Netherlands Organization for Scientific Research (NWO NWA.1389.20.075)., Sam Riedijk As of the 1st September 2022, the authors are involved in a Netherlands Consortium “Public Realm Entrance of Human Germline Gene Editing” funded by The Netherlands Organization for Scientific Research (NWO NWA.1389.20.075).

C23.6 Genetic counselling for caregivers of children and adolescents diagnosed with psychiatric disorders - a pilot study

Madalina Radu ¹, Andra Ciuca¹, Sebastian Pintea¹, Ramona Moldovan^1;2;3

¹Babeș-Bolyai University, Department of Psychology, Cluj-Napoca, Romania; ²University of Manchester, Division of Evolution and Genomic Sciences, Manchester, United Kingdom; ³Manchester Academic Health Science Centre, Manchester Centre for Genomic Medicine, St Mary’s Hospital, Manchester, United Kingdom

Consortium: -

Background/Objectives: The efficacy of psychiatric genetic counselling has been well documented for patients with psychiatric disorders. Family members of adult patients also appear to benefit. To date, there is no data available for parents of children and adolescents diagnosed with psychiatric disorders. This study aimed to investigate the impact of psychiatric genetic counselling on parent’s knowledge, empowerment, well-being and stigma.

Methods: Parents of young people diagnosed with psychiatric disorders in an outpatient clinic were invited to take part. Participants were assessed pre and post intervention using standardised questionnaires. A registered genetic counseller offered the parents one genetic counselling session.

Results: Of the 43 paticipants enrolled (i.e. 41 mothers and 2 fathers, mean age 40.2 ± 5.9 years old), 32 completed the post test measures. Yound people’s (mean age 9 ± 4 years old) diagnosis included ASD, ADHD, affective disorders and anxiety disorders. Statistical analyses shows genetic counselling has a significant impact on parents‘ knowledge [t(31) = -7.132, p = .000] and empowerment [t(31) = -2.723, p = .011]. Parents of children with anxiety disorders significantly improved their knowledge (r = .400, p < .05) and parents of children with ASD significantly decreased stigma (r = .379, p < .05). The sooner after the moment of diagnosis parents attended the session, the greater the impact on their knowledge (r = .388, p < .05) and well-being (r = .381, p < 0.05).

Conclusion: Psychiatric genetic counselling improves parents’ knowledge, empowerment, well-being and stigma. This can help families by providing them with the support they need to better understand, adapt, and manage the condition in their family.

Grants: -

Conflict of Interest: None declared

C24.1 Diagnostic yield of genome-wide genetic testing and demographic distribution in 741 index patients with inborn errors of immunity

Darja Gauck ¹, Marc Sturm¹, Ines Brecht², Ute Grasshoff¹, Stefanie Beck-Wödl¹, Tatjana Welzel³, Ursula Holzer², Martin Kehrer¹, Olaf Riess^1;4, Julia Skokowa⁵, Jasmin B. Kümmerle-Deschner⁶, Tobias Haack¹

¹University of Tübingen, Institute of Medical Genetics and Applied Genomics, Tübingen, Germany; ²University Children’s Hospital, University of Tübingen, Department of Pediatric Hematology and Oncology, Tübingen, Germany; ³University Children’s Hospital Basel, University of Basel, Pediatric-Rheumatology, Basel, Switzerland; ⁴University of Tübingen, Center for Rare diseases, Tübingen, Germany; ⁵University Hospital Tübingen, Department of Oncology, Hematology, Clinical Immunology, and Rheumatology, Tübingen, Germany; ⁶University Hospital Tübingen, Department of Pediatrics, Division of Pediatric Rheumatology and autoinflammation reference center Tübingen (arcT), Tübingen, Germany

Background/Objectives: Human Inborn Errors of Immunity (IEI) encompass a heterogeneous group of genetic disorders associated with versatile clinical features. In this report we provide an overview of the diagnostic yield and demographics in a prospective multi-center cohort of 741 index patients with a suspected clinical diagnosis of IEI.

Methods: Sequencing libraries were generated from genomic DNA for genome (TruSeq DNA PCR-free kit) or exome (TWIST Custom Exome IMGAG V2/SureSelectXT Human All Exon kits V6/V7) sequencing and sequenced on a NovaSeq6000 System (Illumina). Data analysis was conducted using an in-house bioinformatics pipeline (megSAP).

Results: A firm genetic diagnosis was established in 10.8% (n = 68/627) of pediatric cases, with a highest rate of solved cases (31%, n = 9/29) in a group of patients with syndromic features and/or severe congenital neutropenia <1 year of age. A definite molecular diagnosis was made in 11.7% (n = 17/145) of adult patients, mostly in the category of autoinflammatory disorders (n = 5). For 15.5% (n = 117) of the entire cohort with variants in known IEI genes and for 1.8% (n = 14) with suspected candidate genes, genetic outcome was uncertain. In 71.6% (n = 556), the underlying genetic cause of IEI remained unclear.

Conclusion: Our observations illustrate the broad genetic and clinical heterogeneity of IEI and provides insights into the dispersion of solved cases among various age groups and disease categories. The increased usage of genome sequencing has the potential to further narrow the diagnostic gap by expanding the analysis to structural variations and non-coding regions.

Conflict of Interest: None declared

C24.2 p53 Unchained – Constitutional Upregulation of p53 Caused by Germline Variants in MDM4 Leads to Bone Marrow Failure and Insufficient Telomere Maintenance

Robert Meyer ^1;2, Senthil Velan Bhoopalan³, Richa Sharma³, Lei Han³, Matthias Begemann^1;2, Marcus Jakob⁴, Daniela Dey^1;2, Tarek Hanafee-Alali⁴, Lara Heller⁵, Heinz Gabriel⁵, Kim Kricheldorf^2;6, Margherita Vieri^2;6, Susanne Isfort^2;6;7, Martin Kirschner^2;6, Jens Panse^2;6, Udo Kontny⁸, Selim Corbacioglu⁴, Shondra M. Pruett-Miller⁹, Claudia Khurana¹⁰, Miriam Erlacher^11;12, Lise Larcher^13;14;15;16, Jean Soulier^13;14;15;16, Miriam Elbracht^1;2, Ingo Kurth^1;2, Tim H. Brümmendorf^2;6, Fabian Beier^2;6, Marcin Wlodarski³

¹Institute for Human Genetics and Genomic Medicine, Medical Faculty, RWTH Aachen University, Aachen, Germany; ²Center for Integrated Oncology Aachen Bonn Cologne Düsseldorf (CIO ABCD), Germany; ³Department of Hematology, St. Jude Children´s Research Hospital, Memphis, United States; ⁴Department of Pediatric Hematology, Oncology and Stem Cell Transplantation, University Hospital of Regensburg, Regensburg, Germany; ⁵Zentrum für Humangenetik Tübingen, Tübingen, Germany; ⁶Department of Hematology, Oncology, Hemostaseology and Stem Cell Transplantation, Medical Faculty, RWTH Aachen University, Aachen, Germany; ⁷Department of Hematology, Hemostasis, Oncology, and Stem Cell Transplantation, Medical School Hannover, Hannover, Germany; ⁸Division of Pediatric Hematology, Oncology and Stem Cell Transplantation, Medical Faculty, RWTH Aachen University, Aachen, Germany; ⁹Department of Cell and Molecular Biology, St. Jude Children´s Research Hospital, Memphis, United States; ¹⁰Klinik für Kinder- und Jugendmedizin, Evangelisches Klinikum Bethel, Bielefeld, Germany; ¹¹Division of Pediatric Hematology and Oncology, Department of Pediatrics and Adolescent Medicine, Medical Center, University of Freiburg, Freiburg, Germany; ¹²German Cancer Consortium (DKTK), Freiburg, Germany; ¹³Institut de Recherche Saint-Louis (IRSL), Université Paris Cité, Paris, France; ¹⁴INSERM U944/CNRS UMR7212, Paris, France; ¹⁵Saint-Louis Hospital, Hematology Laboratory, APHP, Paris, France; ¹⁶Centre de Référence Maladies Rares “Aplasie Médullaire”, Saint-Louis and Robert Debré Hospitals, Paris, France

Background/Objectives: Loss of function of p53 – often referred to as the guardian of the genome – has been studied extensively, both in germline and cancer cells. Data on the effects of an overactive p53, however, are rare. Based on clinical and functional data, we describe the effects of constitutional dysregulation of p53 caused by germline variants in MDM4, an important regulator of p53. Until now, a single family carrying a pathogenic missense variant in MDM4 had been described.

Methods: Genome/Exome/RNA sequencing, Flow-FISH telomere length measurement, western blot, CRISPR/Cas9-guided knock-in of MDM4 variants in iPSC lines, deletion of MDM4 in CD34+ cells and mouse xenotransplantation studies.

Results: Cas9-mediated mutation of MDM4 in iPSC and CD34+ cells resulted in increased p53 activity and fewer erythroid and myeloid colonies, which was rescued by TP53 knock-out. Xenotransplant studies showed a p53-dependent bone marrow repopulation defect in MDM4-disrupted CD34+ cells. Six individual patients carrying germline variants in MDM4 presented with a varying degree of bone marrow failure (BMF), a tendency to shortened telomeres and further single features of telomere biology disorders (TBD).

Conclusion: Here, we provide strong evidence from functional and clinical data of six unrelated patients that germline variants in MDM4 cause a disorder of p53 dysregulation resulting in a variable phenotype including BMF and features of TBD. While clinical trials for MDM4 inhibitors to reactivate p53 in tumor cells are ongoing, our work allows an outlook on potential side effects of these drugs and sheds light on an unfamiliar dark side of p53.

Grants:

Conflict of Interest: None declared

C24.3 Combined long- and short-read RNA sequencing of pathogen stimulated primary immune cells reveals the expression of uncharacterized genes and transcripts

Emil Vorsteveld ¹, Renee Salz², Caspar van der Made³, Simone Kersten¹, Charlotte Kaffa², Merel Stemerdink¹, Tabea Riepe², Tsung-han Hsieh⁴, Musa Mhlanga⁴, Mihai Gheorghe Netea³, Pieter-Jan Volders⁵, Peter-Bram t Hoen², Alexander Hoischen¹

¹Radboud University Medical Center, Department of Human Genetics, Nijmegen, Netherlands; ²Radboud University Medical Center, Department of Medical BioSciences, Nijmegen, Netherlands; ³Radboud University Medical Center, Department of Internal Medicine, Nijmegen, Netherlands; ⁴Radboud University Nijmegen, Department of Cell Biology, Nijmegen, Netherlands; ⁵University of Ghent, Department of Biomolecular Medicine, Gent, Belgium

Background: Immune responses are shaped by the nature of infections and by inter-individual variability, contributing to differential susceptibility to infections and to various diseases with an inflammatory component. Dynamic transcript and protein expression in a range of cells responsible for the innate immune response is important to shape the first line of defense against a wide variety of pathogens.

Methods: Here we perturbed immune-cells with in vitro pathogen exposure. We stimulated PBMCs from 5 healthy donors for 4h or 24h with LPS, S. aureus, Poly(I:C) or C. albicans with RPMI medium as control, resulting in a total of 52 studied samples. We performed short read sequencing using Lexogen QuantSeq in all samples, as well long read sequencing using PacBio IsoSeq in a subset of samples.

Results: Short-read sequencing reveals common and distinct genes and pathways expressed during immune responses to different pathogens. Beside well-established genes, we highlight uncharacterized genes KIAA0040 and FAM49A, which show up to 2.3-fold increased expression after pathogen exposure and are co-expressed with modules of established immune genes. Long-read sequencing revealed 47.7% novelty in the transcriptomes. We found widespread isoform switching induced upon pathogen stimulation. We highlight novel transcripts of NFKB1 and CASP1 that may indicate novel immunological mechanisms.

Conclusions: Transcriptome profiling of pathogen-stimulated immune cells using paired short- and long-read approaches highlights candidate immune genes and identifies novel transcripts, revealing a more complex transcriptome landscape following pathogen exposure than previously appreciated.

Grants: MGN: ERC Advanced (#833247), Spinoza Grant (NWO). AH: Solve-RD (EU H2020 #779257).

Conflict of Interest: None declared

C24.4 A collective influence of genetic background on influenza vaccinees responses

Nhan Nguyen ^1;2, Saumya Dileep Kumar^1;2, Martijn Zoodsma^1;2, Janyn Heisig³, Stephanie Trittel³, Peggy Riese³, Cheng-Jian Xu^1;2;4, Frank Pessler^1;5, Luis Graca⁶, Carlos Alberto Guzman³, Yang Li^1;2;4;7

¹Helmholtz Centre for Infection Research (HZI), Centre for Individualised Infection Medicine (CiiM), Hannover, Germany; ²TWINCORE, Center for Experimental and Clinical Infection Research, Hannover, Germany; ³Helmholtz Centre for Infection Research (HZI), Department Vaccinology and Applied Microbiology, Braunschweig, Germany; ⁴Radboud University Medical Center, Department of Internal Medicine and Radboud Center for Infectious Diseases, Nijmegen, Netherlands; ⁵TWINCORE, Center for Experimental and Clinical Infection Research, Research Group Biomarkers for Infectious Diseases, Hannover, Germany; ⁶Instituto de Medicina Molecular João Lobo Antunes, Faculdade de Medicina, Universidade de Lisboa, Lisbon, Portugal; ⁷Hannover Medical School, Cluster of Excellence Resolving Infection Susceptibility (RESIST; EXC 2155), Hannover, Germany

Background/Objectives: While influenza infection is a significant public health threat leading to about 290,000-650,000 deaths annually worldwide, the effectiveness of influenza vaccines varies between 40 to 80% among individuals. In this study, we developed a reclassification concept to differentiate protected and unprotected vaccinees and identified a plausible genetic determination for unprotected vaccinees.

Methods: From blood samples of 286 healthy donors spanning 4 influenza seasons (45% female, age range: 18-81), we generated a multimodal dataset including genotype, whole blood transcriptome, plasma proteome, metabolome, and serological parameters. Based on the serological correlate for protection, we used pre-vaccination antibody titer and antibody titer foldchange between pre- and post-vaccination to establish a new approach to classify protected and unprotected vaccinees. Thereafter, we analyzed the differences in omics profiles between protected and non-protected vaccinees.

Results: We identified proteins and metabolites that could serve as predictive markers to identify unprotected vaccinees before their vaccine administration. In QTL analysis with post-vaccination antibody titers, we observed 3 loci (p > 10e-7) with nearby genes such as TCERG1L, SHCBP1, ORC6, GTP2, DNAJA2, ITFG1 that are implicated in multiple immune responses. Furthermore, we predicted multi-omics signatures at gene expression, protein, and metabolite levels directly from genotypes using OmicsPred models. Then, we identified omics signatures, such as PTX3 and CD5, which were genetically determined for non-protected vaccinees.

Conclusion: This study demonstrates a new approach to classify protected and unprotected individuals after influenza vaccination, potential biomarkers for predicting unprotected vaccinees, and a collective influence of genetic background on non-protected vaccinees.

Grants: ERC Starting Grant (948207)

Conflict of Interest: None declared

C24.5 A comprehensive genetic map of cytokine responses in Lyme borreliosis

Javier Botey Bataller ^1;2;3, Hedwig D. Vrijmoeth^1;4, Jeanine Ursinus^4;5, Bart Jan Kullberg¹, Cees C. van den Wijngaard⁴, Hadewych ter Hofstede¹, Ahmed Alaswad^2;3, manoj Kumar Gupta^2;3;6, Lennart Rösner⁶, Jochen Huehn^6;7, Thomas Werfel⁶, Thomas F. Schulz⁶, Cheng-Jian Xu^1;2;3, Mihai Gheorghe Netea^1;8, Joppe W.R. Hovius⁵, Leo A.B. Joosten^1;9, Yang Li^1;2;3;6

¹Radboud University Medical Center, Nijmegen, Netherlands; ²Centre for Individualised Infection Medicine (CiiM), Hannover, Germany; ³Twincore, Zentrum für Experimentelle und Klinische Infektionsforschung GmbH, Hannover, Germany; ⁴National Institute for Public Health and Environment, Bilthoven, Netherlands; ⁵Amsterdam UMC, locatie AMC, Amsterdam, Netherlands; ⁶Hannover Medical School, Hannover, Germany; ⁷Helmholtz Centre for Infection Research, Braunschweig, Germany; ⁸LIMES-Institut der Universität Bonn, Bonn, Germany; ⁹Iuliu Hațieganu University of Medicine and Pharmacy, Cluj-Napoca, Romania

Background: The incidence of Lyme borreliosis is rapidly increasing, leading to more patients experiencing persistent symptoms after antibiotic treatment. The innate immune system and related cytokines play a crucial role in the host response to infection and may be involved in the development of persistent symptoms.

Methods: Previously, we identified genetic loci that contribute to inter-individual variation in cytokine responses (cQTLs) and affect the susceptibility to infection. Here, we used a functional genomics approach to characterize the cytokine profile of 1,060 physician-confirmed Lyme borreliosis patients. We measured cytokine production upon ex vivo pathogen stimulation in samples taken before the start of antibiotic treatment and six weeks thereafter, resulting in a total of 108 cytokine-stimulation conditions.

Results: We confirmed the inhibiting effect of IL-10 on antibody production against B. burgdorferi s.l. and observed enhanced IL-1Ra responses in patients with disseminated disease. Genome-wide mapping the cytokine production allowed us to identify 34 cQTLs, 31 of which were not found in previous studies. We pinpointed the causal variant at the TLR1-6-10 locus and validated the specific regulation of this locus on IL-1Ra responses to Pam3Cys compared to B. burgdorferi s.l. at the transcriptome level using an independent cohort. Our findings show that cQTLs contribute to the susceptibility to Lyme borreliosis, and are relevant to other immune-mediated diseases such as allergy, IBD and multiple sclerosis.

Conclusions: Collectively, our results improve our understanding of the role of cytokine responses in Lyme borreliosis and provide a genetic map of general immune function as an expanded resource, made available at https://lab-li.ciim-hannover.de/apps/lyme_cqtl/.

Conflict of Interest: None declared

C24.6 Effect of parental autoimmune diseases on type 1 diabetes in offspring can be partially explained by HLA and non-HLA polymorphisms: a nationwide registry and biobank study in 7.2M Finns

Feiyi Wang ¹, Aoxing Liu^1;2;3;4, FinnGen Project¹, Tiinamaija Tuomi^1;5;6;7, Andrea Ganna^1;2;3;4

¹Institute for Molecular Medicine Finland, Helsinki, Finland; ²Broad Institute of Harvard and MIT, Program in Medical and Population Genetics, Cambridge, United States; ³Broad Institute of Harvard and MIT, Stanley Center for Psychiatric Research, Cambridge, United States; ⁴Massachusetts General Hospital, Analytic and Translational Genetics Unit, Boston, United States; ⁵HUS, Abdominal Center, Endocrinology, Helsinki, Finland; ⁶Folkhalsan Research Center, Helsinki, Finland; ⁷Lund University Diabetes Center, Malmo, Sweden

Consortium: FinnGen

Type 1 diabetes (T1D) and some autoimmune diseases (ADs) are observed to co-occur in families, yet it remains largely unknown whether offspring with ADs-affected parents would have elevated risks of T1D and to what extent such intergenerational transmission is attributed to shared genetic factors, especially when partitioning into HLA and non-HLA variations. Leveraging nationwide registers of 7.2M Finns - the population with the highest T1D prevalence globally - we defined 50 parental AIDs in 58,284 family trios and observed 10 parental ADs exerting an increased risk of T1D in offspring. The identified epidemiological associations were further evidenced by comprehensive genetic analyses performed in 470K genotyped Finns of the FinnGen study (N of trios = 10,787). Collectively, our findings suggested that intergenerational transmission of ADs to offspring T1D can be partially explained by HLA and non-HLA polymorphisms, but in a disease-dependent manner. For example, coeliac disease and psoriasis are mainly impacted by HLA while autoimmune hypothyroidism and rheumatoid arthritis also had non-HLA contributors in addition to HLA. We proposed a novel parental polygenic score, integrating variations of both HLA and non-HLA, as a useful predictor to evaluate the risk of developing T1D in offspring and observed parental ADs could impact offspring T1D incidence even among those whose parents were T1D-unaffected. While presenting the power of combining high-dimensional nationwide registers with rich genetic information in studying disease etiology, the analytical framework developed here can also be used to reveal detailed mechanical insights into the transmission of any diseases or traits across generations.

Conflict of Interest: Feiyi Wang: None declared, Aoxing Liu: None declared, FinnGen Project: None declared, Tiinamaija Tuomi: None declared, Andrea Ganna Founder of Real World Genetics

C25.1 Redefining the relevance of genetic variants in kidney stone disease through combined genetic and biochemical analyses has prognostic value

Johannes Münch¹, Joana Figueiro da Silva², Elena Cabello², Michael Papik², Ivan Ivanovski², Beatrice Oneda², Daniel Fuster³, Harald Seeger⁴, Thomas Ernandez⁵, Florian Buchkremer⁶, Grégoire Wuerzner⁷, Nasser Dhayat⁸, Alexander Ritter⁹, Catherine Stoermann⁵, Stephan Segerer⁶, Beat Roth⁷, Anita Rauch², Manuel Ferraro¹⁰, Olivier Bonny¹¹, Carsten Wagner¹, Ruxandra Bachmann-Gagescu ²

¹University of Zurich, Physiology, ZURICH, Switzerland; ²University of Zurich, Medical Genetics, Schlieren / Zurich, Switzerland; ³University Hospital Bern; ⁴Kantonsspital Baden; ⁵University hospital Geneva; ⁶Kantonsspital Aarau; ⁷University Hospital Lausanne; ⁸B. Braun Medical Care AG, Hochfelden; ⁹Kantonsspital St Gallen; ¹⁰Fondazione Policlinico Universitario A. Gemelli IRCCS; ¹¹Hôpital fribourgeois, Departement de médecine

Kidney stone disease (KSD) affects 10% of the population and is predominantly caused by complex interplay of lifestyle and genetic factors. Rare genetic variants can also cause monogenic KSD, which accounts for up to 30% in pediatric cohorts, while its prevalence in unselected adult cohorts is less clear. Challenges with variant classification and variable inheritance modes, with both mono- and bi-allelic variants in the same genes suggested to cause monogenic KSD, complicate the interpretation of their clinical relevance. To clarify this, we studied 736 kidney stone formers (KSFs) and 201 controls from the Swiss kidney stone cohort, a multicenter longitudinal observational study, with WES and detailed biochemical profiling. This combined approach and comparison with controls confirmed the role of several recessive and dominant genes involved mostly in phosphate or cystine handling in monogenic KSD. Heterozygous SLC7A9 variants increased the risk for KSD but required additional factors for disease occurrence. Notably, a common variant in SLC7A9, previously classified as LP/P, has no biochemical consequence and is not enriched in KSFs. Similarly, we show that heterozygous SLC34A1 and SLC9A3R1 LP/P variants do not perturb tubular phosphate reabsorption and are unlikely to cause monogenic KSD. Considering all factors, we find a genetic cause for KSD in 7.2% of KSFs. Importantly, we observed a steeper decline in kidney function in KSFs harboring LP/P variants compared to those without. Taken together, combined genetic and biochemical analyses clarify the clinical relevance of variants in KSD genes and have prognostic value. Funding: NCCR Kidney.CH

Conflict of Interest: None declared

C25.2 INSL3 variants and Insl3 deficiency in human and murine CAKUT

Anne Christians¹, Esra Kesdiren¹, Helge Martens ¹, Lina Werfel^1;2, Ulrike Beyer¹, Robert Geffers³, Martina Muehlenhoff⁴, Jan Hinrich Bräsen⁵, Ibrahim M. Adham⁶, Anna-Carina Weiss⁴, Dieter Haffner², Ruthild G. Weber¹

¹Hannover Medical School, Department of Human Genetics, Hannover, Germany; ²Hannover Medical School, Department of Pediatric Kidney, Liver and Metabolic Diseases, Hannover, Germany; ³Helmholtz Centre for Infection Research, Genome Analytics Research Group, Braunschweig, Germany; ⁴Hannover Medical School, Institute of Clinical Biochemistry, Hannover, Germany; ⁵Hannover Medical School, Nephropathology, Department of Pathology, Hannover, Germany; ⁶University Medical Centre, Institute of Human Genetics, Göttingen, Germany

Background/Objectives: Variants in the insulin-like 3 (INSL3) gene encoding a ligand of relaxin family peptide receptor 2 (RXFP2) and Insl3 deficiency cause cryptorchidism. Here, a cryptorchidism-associated INSL3 missense variant co-segregated with the CAKUT phenotype in females of an index family, instigating the investigation of the role of INSL3 in human and murine CAKUT.

Methods: Whole-exome or targeted INSL3 sequencing was done in 312 CAKUT families. Variant carriers were subjected to reverse phenotyping. INSL3 expression was determined in human fetal and adult tissues and the developing murine kidney by qRT-PCR. A tubulo-morphogenesis assay was done on CRISPR/Cas9-derived Insl3-deficient mIMCD3 cells. Insl3 knock-out mice were characterized morphologically and histologically with respect to urogenital anomalies.

Results: Rare INSL3 variants predicted to be deleterious were identified in 6 of 312 (1.9%) CAKUT families, significantly more frequently than in controls (0.8%, p = 0.04). The four different INSL3 variants detected were heterozygous, and almost exclusively maternally inherited. Cryptorchidism was significantly more frequent in male CAKUT patients with versus without INSL3 variants (67% vs. 9%, p = 0.02). INSL3 was expressed in the fetal and adult human kidney. Insl3 and Rxfp2 transcripts were present in the developing murine kidney. In Insl3^+/- and Insl3^-/- mIMCD3 cells, fewer tubular structures were observed indicating impaired tubulogenesis processes. The CAKUT spectrum in Insl3^+/- and Insl3^-/- mice included kidney hypo(dys)plasia, fusion, segmental nephron losses, hydronephrosis, duplex ureter, a narrowed ureteric lumen.

Conclusion: Heterozygous INSL3 variants associate with human CAKUT, and heterozygous and homozygous Insl3 knock-out can cause CAKUT in mice.

Grants: Deutsche Forschungsgemeinschaft (KO5614/2-1, MA9606/1-1), Else Kröner-Fresenius-Stiftung (2018_Kolleg.12, TITUS-Clinician Scientist Program)

Conflict of Interest: None declared

C25.3 Targeted RNA sequencing in monogenic renal diseases

Clément SAUVESTRE ¹, Marc Bras², Zaina Ait Arkoub¹, Vincent Moriniere¹, Christelle Arrondel³, Nicolas Cagnard², Patrick Nischke², Manon Mautret Godefroy¹, Laurence Heidet^3;4, Corinne Antignac³, Guillaume Dorval^1;3

¹APHP, Université Paris-Cité, Service de Médecine Génomique des Maladies Rares, Paris, France; ²Institut Imagine, Université de Paris, Plateforme Bio-informatique, Inserm UMR 1163, Paris, France; ³Institut Imagine, Université Paris Cité, Laboratoire des Maladies Rénales Héréditaires, Inserm UMR 1163, Paris, France; ⁴Hôpital Necker-Enfants Malades, Assistance publique, Hôpitaux de Paris (AP-HP), Service de néphrologie pédiatrique Centre de Référence MARHEA, Paris, France

Background/Objectives: RNA sequencing (RNAseq) is a crucial tool for functional assessment of certain variants affecting splicing in the post-genomic era. However, it can be challenging to use when studying genes expressed in specific, inaccessible tissues, such as kidney tissue. Additionally, transcript analysis requires significant depth of coverage, which can be difficult to achieve.

Methods: This study applies Targeted RNAseq technology to capture a panel of 228 genes involved in monogenic kidney diseases in various tissues (kidney, blood, fibroblasts, and urine) from patients with a previously identified intronic variant affecting splicing. The expression profiles of the captured genes in each tissue were compared, and the impact of each identified variant on RNA splicing was interpreted.

Results: Several samples were studied for each tissue. On average, 11,176 reads were aligned with the exon-exon junctions identified, providing sufficient reading depth to interpret the splicing events (approximately 354 reads per junction). The analysis of different samples in clusters demonstrated the homogeneity and repeatability of analysis in tissues. For instance, the analysis of RNA from the complete urine of three patients allowed for the examination of splicing of 222 out of 237 captured genes in all cases. Specific studies of the transcripts of interest were conducted and confirming all the variants identified.

Conclusion: This work demonstrates that studying RNA from various tissues, enables functional interpretation of variants affecting splicing. In the field of nephrogenesis, RNA from urine can be used to interpret events in almost all of the genes studied.

Grants: No funding

Conflict of Interest: None declared

C25.4 Metabolome and exome (WES) data integration to stratify Non-Alcoholic Fatty Liver disease (NAFLD) patients

Rebecca Filomena ¹, Carla Debernardi¹, Cecilia Di Primio¹, ANGELO SAVOCA¹, Gian Paolo Caviglia², Chiara Catalano¹, clara viberti¹, Elisabetta Casalone¹, Chiara Rosso², Angelo Armandi², Alessandra Allione¹, Marcello Manfredi³, Miriam Rosselli¹, elton jalis herman¹, Alessia Russo¹, Elisabetta Bugianesi², Giuseppe Matullo^1;4

¹University of Turin, Unit of Genomic Variability and Complex diseases, Department of Medical Sciences, Torino, Italy; ²University of Turin, Division of Gastroenterology and Hepatology, Department of Medical Sciences, Torino, Italy; ³University of Piemonte Orientale, Department of Translational Medicine, Novara, Italy; ⁴AOU Città della Salute e della Scienza, Medical Genetics Unit, Torino, Italy

Background/Objectives: Non-alcoholic fatty liver disease (NAFLD) is defined as the accumulation of excessive fat in the liver in the absence of excessive alcohol consumption or other liver diseases. It is the most common cause of liver disease worldwide and includes a wide range of liver damage such as simple steatosis (non-alcoholic fatty liver; NAFL) and non-alcoholic steatohepatitis (NASH). The aim of this study is to find a potential correlation between the genetic and metabolic component of the disease than can better differentiate the phenotypes based on the severity.

Methods: We performed untargeted metabolomics analysis and Whole Exome Sequencing (WES) on 154 patients with biopsy-proven NAFLD status, enrolled at the Liver Unit of the Department of Medical Sciences, University of Turin.

Results: From the untargeted metabolomics analysis, 8 metabolites out of 380 resulted to be significantly differentially expressed (p-value < 0.05, 0.769<Fold Change<1.3). Pathway analysis showed that all of them are involved in the Aminoacyl-tRNA biosynthesis (FDR = 6.82×10⁻⁵).

As a preliminary analysis, aimed to identify metabolite-QTL, we selected the exome variants in the 44 genes related to this pathway. We observed 230 non-synonymous variants, with an average of 14 variants per sample in the investigated genes. Moreover, a higher metabolite expression correlated with an increased average number of variants.

Conclusion: The integration of metabolomics and WES could help to identify genes and metabolites involved in NAFLD progression, improve patients’ stratification and potentially develop new therapies.

Grants: Ministero dell’Istruzione, dell’Università e della Ricerca (MIUR) Project “Dipartimenti di Eccellenza 2018–2022. Project n° D15D18000410001”

Conflict of Interest: None declared

C25.5 The molecular study of 407 Chinese children with 46,XY Disorders of Sex Development (DSD) by Next-generation sequencing (virtual)

Xiaojun Tang ¹, Xiaoping Lan¹, Xiaozhen Song¹, wuhen xu¹, Yongchen Yang¹, Wenhao Weng¹, Shengnan Wu¹

¹Shanghai Children’s Hospital Luding Rd Yuanqu-Blood Bank, Shanghai Children’s Hospital, Shanghai, China

Objectives 46,XY Disorders of Sex Development (DSD) is the most common and complicated subtype, accounting for approximately 50% of all DSD cases. Molecular diagnoses provide guidance for suitable treatments, prognosis evaluation, and genetic counseling. This study aims to evaluate the diagnostic yield and clinical value efficacy of next-generation sequencing (NGS) in children with 46,XY DSD.

Methods Single/trio whole exome sequencing or targeted panel sequencing were performed on 407 Chinese children with 46,XY DSD. Single-nucleotide variants (SNVs), insertions/deletions and copy number variants (CNVs) were concurrently analyzed and interpreted according to American College of Medical Genetics guideline. The variants detected by NGS were confirmed by Sanger sequencing, QPCR or MLPA accordingly. The clinical presentation and family history were collected.

Results The overall diagnostic yield of this cohort stood at 39.56%, comprising 155 with SNVs, 5 with CNVs and one with both. SNVs were most frequently found in SRD5A2, AR, NR5A1, FGFR1, CYP17A1, HSD17B3 and ANOS1. Notably, 56 variants have not been previously reported, expanding the mutation spectra of DSD related genes. It’s crucial to highlight that intragenic exons deletion/duplication were identified in 6 patients. Interestingly, 35 variants were categorized as variants of uncertain significance due to lack of evidence currently, for which further functional experiments and additional patients are needed to clarify the pathogenicity of these variants.

Conclusion The early diagnosis can provide an accurate molecular etiology, which enables to detect potential phenotypes, guide appropriate therapies and clinical follow-up, which benefiting the clinical management of patients.

Grants Shanghai Children’s Hospital Funding (2021YGZQ07)

Conflict of Interest: None declared

C25.6 Added value of clinical exomes in the management of male infertility - study on ESTonian ANDrology (ESTAND) cohort.

Kristiina Lillepea¹, Anna-Grete Juchnewitsch¹, Laura Kasak¹, Anu Valkna¹, Avirup Dutta¹, Kristjan Pomm², Erik Tamp³, Stanislav Tjagur², Mailis Sütt², Olev Poolamets², Margus Punab², Maris Laan ¹

¹University of Tartu, Institute of Biomedicine and Translational Medicine, Tartu, Estonia; ²Tartu University Hospital, Andrology Clinic, Tartu, Estonia; ³East-Tallinn Central Hospital, Center of Pathology, Tallinn, Estonia

Background/Objectives: Infertility affects ~10% men and a large fraction of idiopathic cases are thought to have an unknown genetic cause. Although hundreds of genes have been reported as plausible candidates linked to spermatogenic failure (SPGF), current management of infertile men does not include exome sequencing (ES). The main objective was to analyze a large nationwide cohort of SPGF patients for likely pathogenic and pathogenic variants (LP/P) in infertility candidate genes to evaluate the diagnostic yield, genotype-phenotype links and contribution of ES to clinical management, counselling of patients for reproductive and overall health.

Methods: ES was performed to 521 SPGF patients (total sperm counts 0 - 10 million per ejaculate) recruited to ESTAND cohort and the analyzed gene panel comprised of >600 infertility linked loci. Variant pathogenicity was evaluated using the ACMG guidelines and variants were confirmed by Sanger sequencing. We also piloted providing feedback to patients about the finding(s) and their broader impact.

Results: More than 10% of patients received molecular diagnosis linked to diverse spectrum of molecular etiologies, including isolated and syndromic forms of SPGF. The findings were distributed to all inheritance modes and some patients carried >1 LP/P variants. Notably, over 20% of identified variants represented recurrent findings in male infertility either reported in other cohorts and/or carried by two ESTAND patients. Link to SPGF was validated for several recently proposed candidate genes (e.g. ASZ1, RBM5, GLUD2).

Conclusion: Introducing ES to andrology practice will support improved diagnostics and multidisciplinary management of SPGF patients.

Grants: PRG1021 (Estonian Research Council)

Conflict of Interest: None declared

C27.1 Enhancing prenatal prognostic accuracy and understanding intrafamilial variability in osteogenesis imperfecta: insights from a retrospective study of 269 cases.

Cristina Peduto ¹, sophie monnot^1;2, Caroline Michot^1;2, Elise Schaefer³, Melanie FRADIN⁴, Carine Abel⁵, Genevieve Baujat^1;6, Valérie Cormier-Daire^1;2

¹Centre de Référence Maladies Osseuses Constitutionnelles, Fédération de médecine génomique des maladies rares, APHP, Hôpital Necker-Enfants Malades,, Paris, France; ²Institut Imagine, Université Paris Cité, INSERM UMR1163, Paris, France; ³Service de Génétique Médicale et Centre de Référence Anomalies du Développement et Syndromes Malformatifs, Hôpitaux Universitaires de Strasbourg, Institut de Génétique Médicale d’Alsace, Strasbourg, France; ⁴Service de Génétique Médicale, CHU de Rennes, Rennes, France; ⁵Service de Génétique et Centre de Diagnostic Anténatal, CHU de Lyon HCL-GH Nord-Hôpital de La Croix Rousse, Lyon, France; ⁶Institut IMAGINE and Hôpital Universitaire Necker-Enfants Malades, Département de Génétique, Paris, France

Background/Objectives: Osteogenesis Imperfecta (OI) is due mainly by mutations in COL1A1 and COL1A2 genes. OI exhibits diverse phenotypes, from severe prenatal forms to mild postnatal ones, posing challenges in prognosis, particularly when diagnosed during pregnancy.

Methods: This retrospective study was based on a cohort of 269 cases from the French Constitutional Bone Diseases Reference Centre OI database (January2017-March2023), with the aim of improve prenatal prognostic accuracy by the correlation of prenatal isolated femoral bowing with disease progression, likewise intrafamilial variability.

Results: Out of 119 individuals (COL1A1 = 55; COL1A2 = 64) with prenatal OI diagnosis only, 25 (COL1A1 = 7;COL1A2 = 15) has isolated femoral bowing. COL1A1 patients (1 missense,5 slicing,1 frameshift) mostly display moderate severity (6/7,85%), however none showed favorable progression. COL1A2 patients (7 missense,2 splicing,2 indels) exhibit a range of phenotypes from mild to severe. Notably, two COL1A2 gene variants (c.2026-2A > G;p.? and c.1109G>C;p.Gly370Ala) showed regressive outcomes, highlighting outcome variability with improved femur bowing and fracture numbers. Otherwise, to assess intrafamilial variability 9 patients inherited the condition from asymptomatic parents among 107 cases: 4 COL1A1 (2 missense,1 frameshift,1 splicing) and 5 COL1A2 (4 missense,1 splicing). We roll out a mosaicism by testing on two separate blood samples, grandparent segregation and analysis on buccal cells and urine. We found 5 parents with de novo variants (COL1A1 = 3, COL1A2 = 2), 3 inherited from a grand-parent (COL1A2 = 3) and 1 mosaic.

Conclusion: These findings suggest that modifier genes may influence OI expression, enhancing our understanding of OI phenotype variability and providing clearer guidance for family counseling concerning prognosis.

Grants: No

Conflict of Interest: None declared

C27.2 First elements of comparison between invasive and non-invasive exomes strategies in prenatal diagnosis when ultrasound reveals developmental anomalies: towards implementation in routine strategy?

Frédéric Tran Mau-Them ^1;2, Jean Muller^3;4;5, Julian Delanne⁶, Virginie Haushalter^3;4, Anne-Sophie Denommé-Pichon^1;2, Audrey Schalk^3;4, Aurore Garde⁶, sophie scheidecker^3;4, Hana Safraou^1;2, Cecile Lang^3;4, Caroline Racine⁶, Ange-Line Bruel^1;2, Antonio Vitobello^1;2, Sophie Nambot⁶, Nicolas Bourgon⁷, Valentin Bourgeois^1;2, Victor Couturier^1;2, Charlotte Poe^1;2, Martin Chevarin^1;2, Thierry Rousseau⁷, Paul Sagot⁷, Emmanuel Simon⁷, Sophie Naudion⁸, Estelle Colin⁹, Agnes Guichet¹⁰, Yosra Halleb¹¹, Alice Goldenberg¹², Anne-Marie Guerrot¹², Clara Houdayer¹², Juliette Coursimault¹², Audrey Putoux¹³, Mathilde Nizon^14;15, Solène Conrad^14;15, Benjamin Durand¹⁶, elise brischoux-boucher¹⁷, Laetitia Lambert¹⁸, Marie-Laure Humbert¹⁹, Melodie Soto², Thi-Thu Creusvaux², Céline Bernard², Christine Binquet¹⁹, Yannis Duffourd^1;2, Benedicte GERARD^3;4, Christophe Philippe^1;2, Nadège Calmels^3;4, Laurence Faivre^2;6, Caroline Schluth-Bolard^3;4, Christel Thauvin-Robinet^1;2;6

¹CHU Dijon, Unité Fonctionnelle Innovation en Diagnostic Génomique des maladies rares, Dijon, France; ²INSERM, UMR1231 GAD, Dijon, France; ³Laboratoire de Génétique Médicale, INSERM, UMR_S 1112, CRBS, Université de Strasbourg, Strasbourg, France; ⁴Nouvel Hôpital Civil, Laboratoires de Diagnostic Génétique, Strasbourg, France; ⁵Hôpitaux Universitaires de Strasbourg, Unité Fonctionnelle de Bioinformatique Médicale appliquée au diagnostic (UF7363),, Strasbourg, France; ⁶CHU Dijon, Centre de Référence Maladies Rares “Anomalies du développement et syndromes malformatifs”, Dijon, France; ⁷CHU Dijon, Service de Gynécologie Obstétrique, Médecine Fœtale et Stérilité Conjugale, Dijon, France; ⁸CHU Bordeaux, Service de Génétique Médicale, Bordeaux, France; ⁹Université Angers, Biochemistry and Genetics Department, Angers; ¹⁰CHU Angers, Service de génétique, Plateau de Biologie Hospitaliere, Angers, France; ¹¹CHU Angers, Centre Robert Debré, Angers, France; ¹²CHU Rouen, Service de génétique - Unité de génétique clinique, Rouen, France; ¹³GH Est-Hôpital Femme Mère Enfant, Service de génétique, Lyon, France; ¹⁴CHU Nantes, Service de Génétique Médicale, Nantes, France; ¹⁵INSERM, Institut du thorax, Nantes, France; ¹⁶Hopital Universitaire de Strasbourg, Institute Genetic Medical D’alsace Igma, Strasbourg, France; ¹⁷Hôpital Saint Jacques, Centre de génétique humaine, Pole de biologie et anatomie pathologie,, Besançon, France; ¹⁸CHU de Nancy, Hôpitaux de Brabois, Service de génétique clinique et médicine infantile, hôpital d’enfants, Vandoeuvre les Nancy, France; ¹⁹UFR des Sciences de Santé, Université de Bourgogne-Franche-Comté,, Centre d'Investigation Clinique CIC-EC Inserm CIC1432, Dijon, France

Background/Objectives: Prenatal invasive exome sequencing (pES) has become a first-line diagnostic tool to refine fetal prognosis and guide couples in pregnancy management when ultrasound abnormalities are detected. Available non-invasive prenatal testing (NIPT) of circulating-free fetal DNA is currently limited to specific genetic indications. Leveraging our experience with pES, we compared NIPT-ES and pES approaches when ultrasound abnormalities are detected.

Methods: We performed trio pES (70X) and NIPT-ES (300X) in 70 pregnancies with ultrasound abnormalities (10-32 weeks of gestation). Only class 5, 4, and 3+ variants in pES, associated with the ultrasound phenotype, were reported during pregnancy. The results from both blindly interpreted strategies were compared and disclosed after the pregnancy outcome.

Results: Preliminary results showed that pES identified the causal diagnosis in 15/43 (35%), whereas NIPT identified it in 9/43 (21%) pregnancies respectively. pES and NIPT-ES were concordant in 36/43 (84%). Compared with pES, NIPT had a false-positive (FP) and false-negative (FN) rates of 3.5% (1/28) and 40% (6/15), respectively. FP and FN results were due to technical limitations of NIPT (two CNV detection) or the complexity of the prenatal situation (incomplete clinical information).

Conclusion: Preliminary results demonstrate the feasibility of NIPT-ES. However, the 3.5% FP rate suggests that its first-line use may lead to few unnecessary invasive procedures and the 40% FN rate suggests that diagnosis may be missed in some cases. All the 70 inclusions will be presented, but further evaluations of NIPT-ES are essential given the FP and FN rates before considering NIPT-ES as a routine strategy and its positioning in pregnancy management.

Conflict of Interest: None declared

C27.3 Comprehensive noninvasive fetal screening by deep trio-exome sequencing

Ieva Miceikaite ^1;2, Qin Hao², Charlotte Brasch-Andersen², Christina Fagerberg², Pernille Torring², Britta Schlott Kristiansen³, Lilian Ousager¹, Lene Sperling³, Mette Holm Ibsen⁴, Katrin Löser⁵, Martin Larsen^1;2

¹University of Southern Denmark, Clinical Research, Odense, Denmark; ²Odense University Hospital, Clinical Genetics, Odense, Denmark; ³Odense University Hospital, Gynecology and Obstetrics, Odense, Denmark; ⁴University Hospital of Southwestern Jutland, Esbjerg, Denmark; ⁵Hospital of Southern Jutland, Aabenraa, Denmark

Background/Objectives: Fetal genetic diagnosis is vital for prenatal care. While invasive prenatal trio-exome sequencing is effective, it’s limited to fetuses with structural anomalies, leaving many monogenic disorders undiagnosed. Noninvasive prenatal screening using maternal plasma cell-free DNA (cfDNA) revolutionized screening but mainly targets chromosomal disorders due to its low resolution. desNIPT, a novel noninvasive screening method using deep trio-exome sequencing, aims to address this gap.

Methods: In our study involving 36 pregnancies, deep trio-exome sequencing was performed on maternal plasma cell-free DNA and parental blood samples. The approach leverages ultra-deep, error-corrected, trio-exome sequencing for accurate detection of all fetal de novo variants, including single-nucleotide variants, indels, copy-number variants, and chromosomal aneuploidies. Results were compared with invasive prenatal analysis outcome.

Results: desNIPT showed 100% concordance with invasive methods in detecting pathogenic fetal de novo variants, including single-nucleotide variants, deletions, trisomies, and an unbalanced translocation. Mean sensitivity for detecting fetal sequence variants across protein-coding region was 95.12%. The method also enables additional carrier screening in both parents.

Conclusion: desNIPT offers a comprehensive, noninvasive approach for prenatal screening of a wide range of genetic disorders. Integrating this approach into routine prenatal care could improve early detection rates, reduce invasive procedures, and facilitate timely interventions, with a low false-positive rate and quick turnaround time. These findings support further exploration of deep-exome sequencing of cell-free DNA for screening fetuses without structural anomalies.

Grants: The Region of Southern Denmark (18/17848 and 20/14085), the Aase and Ejnar Danielsens Foundation (19-10-0259), the A.P. Møller Foundation for the Advancement of Medical Science (19-L-0281).

Conflict of Interest: Ieva Miceikaite Nonacus Ltd. is sponsoring my travel to ESHG, no sponsorship or involvement was provided for the study itself., Qin Hao: None declared, Charlotte Brasch-Andersen: None declared, Christina Fagerberg: None declared, Pernille Torring: None declared, Britta Schlott Kristiansen: None declared, Lilian Ousager: None declared, Lene Sperling: None declared, Mette Holm Ibsen: None declared, Katrin Löser: None declared, Martin Larsen: None declared

C27.4 Novel Utility of Non-Invasive Fetal Exome Sequencing

Christopher Whelan ^1;2;3, Harrison Brand^1;2;3;4;5, Michael Duyzend^1;2;6, John Lemanski², Jessica Giordano⁷, Monica Salani², Stephanie Hao^2;5, Elise Valkanas^1;2, Caroline Cusick³, Diane Lucente², Krithika Nathamuni^1;2, Casie Genetti⁶, Lori Dobson⁸, Courtney Studwell⁸, Kathleen Gianforcaro⁸, Louise Wilkins-Haug⁹, Stephanie Guseh⁹, benjamin Curral², Ronald Loosen⁷, Kathryn Gray¹⁰, Wendy Chung⁶, Ronald Wapner⁷, Michael Talkowski^1;2;3;4

¹Broad Institute of MIT and Harvard, Program in Medical and Population Genetics, Cambridge, United States; ²Massachusetts General Hospital, Center For Genomic Medicine, Boston, MA, United States; ³Broad Institute of MIT and Harvard, Stanley Center for Psychiatric Research, Cambridge, MA, United States; ⁴Massachusetts General Hospital and Harvard Medical School, Department of Neurology, Boston, MA, United States; ⁵Massachusetts General Hospital, Pediatric Surgical Research Laboratory, Boston, MA, United States; ⁶Boston Children’s Hospital and Harvard Medical School, Department of Pediatrics, Boston, MA, United States; ⁷Columbia University Medical Center, Department of Obstetrics and Gynecology, New York, NY, United States; ⁸Brigham and Women’s Hospital, Center for Fetal Medicine and Reproductive Genetics, Boston, MA, United States; ⁹Brigham and Women’s Hospital and Harvard Medical School, Division of Maternal-Fetal Medicine, Boston, MA, United States; ¹⁰University of Washington, Department of Obstetrics & Gynecology, Seattle, WA, United States

Background/Objectives: Analysis of cell-free DNA (cfDNA) has revolutionized prenatal diagnostics, providing a noninvasive screen for aneuploidies. We devised a higher-resolution approach with the ability to capture all coding fetal genetic variation: noninvasive fetal exome sequencing (NIFS). We applied NIFS to 51 pregnancies (gestational ages 9 – 40 weeks) and developed novel methods for detecting single nucleotide variants (SNVs), indels, and copy number variants (CNVs). We are currently performing a validation study that includes over 100 participants with matched NIFS and whole genome sequencing (WGS).

Methods: NIFS exome libraries were prepared from cfDNA extracted from maternal plasma and sequenced to 200x mean coverage on ~22,000 genes. We detected SNVs and indels using machine learning-based variant filtration and a Bayesian mixture model that disentangles maternal and fetal variants. We identified CNVs via analysis of read depth.

Results: In 11 samples with matched invasive sequencing, NIFS detected de novo and paternal variants with high sensitivity (median 96.3%) irrespective of fetal fraction. Across all variants, NIFS demonstrated high median sensitivity for fetal SNVs (93.0%) and indels (77.1%). All 5 clinically confirmed pathogenic variants were detected. NIFS genotyped mothers with high average accuracy (97.6% SNVs; 86.4% indels) and 57% had a reportable carrier variant. Our validation study permits further extensive benchmarking.

Conclusion: NIFS extends prenatal cfDNA screening to SNVs, indels, and CNVs, and potentially offers a rapid reflex test in the presence of ultrasound abnormalities. NIFS demonstrates that fetal exomes are accessible from maternal blood samples collected for prenatal screening.

Grants: NICHD (HD081256, HD105266, HD112084); NIMH (MH115957).

Conflict of Interest: None declared

C27.5 Fetal genome sequencing: what do we miss by prenatal exome sequencing ?

Agnese Feresin ¹, Boris Keren^2;3, Delphine Heron⁴, Benjamin Cogne^5;6, Sarah Grotto^7;8, Christele Dubourg^6;9, Anne Dieux¹⁰, Sandra Whalen¹¹, Chloe Quelin¹², Jonathan Levy^6;13, Daphné Lehalle⁸, Paul Gueguen^6;14, Mathilde Nizon¹⁵, melanie rama^6;16, LYSE RUAUD¹³, Sophie Rondeau^6;17, Anne Brehin¹⁸, Corinne Collet^6;17, Leila Ghesh^15;19, Laurence Heidet^6;20, YLINE CAPRI¹³, lydie BURGLEN^6;21, Laurence Perrin¹³, severine drunat^6;13, Catherine Delorme²², Fabienne Escande^23;24, Claire Beneteau¹⁵, Lucie Boutaud^6;25, Florence Petit²⁶, Pascale Saugier-Veber^6;27, Audrey Briand-Suleau^6;28, Stéphanie Arpin^29;30, Andrée Delahaye^6;31, Manon Godin³², Marie Faoucher^6;9, Anne-Marie Guerrot¹⁸, Nathalie Couque^6;33, Alice Goldenberg¹⁸, Yoann Vial^6;13, Alinoe Lavillaureix³⁴, Adeline Bonnard¹³, Agnes Guichet^6;35, Alban Lermine⁶, Pierre Blanc^6;36, Thomas Smol^6;37, Odile Boute³⁸, Tania Attie-Bitach^6;39

¹Service de Médecine Génomique des Maladies Rares, UF MP5, Hôpital Necker-Enfants Malades, Assistance Publique-Hôpitaux de Paris, Paris, France; ²Centre de Génétique Moléculaire et Chromosomique, Hôpital de la Pitié-Salpêtrière, UF de génomique du développement, Assistance Publique-Hôpitaux de Paris, Sorbonne Université, Paris, France; ³Laboratoire de Biologie Médicale multisites SeqOIA – PFMG2025, Paris, France; ⁴Centre de Génétique Moléculaire et Chromosomique, Hôpital de la Pitié-Salpêtrière, UF de génomique du développement, Assistance Publique-Hôpitaux de Paris, Paris, France; ⁵Service de Génétique Médicale, Centre Hospitalier Universitaire de Nantes, Nantes, France; ⁶Laboratoire de Biologie Médicale multisites SeqOIA – PFMG2025, Paris, France; ⁷Maternité Port-Royal, Hôpital Cochin, Assistance Publique-Hôpitaux de Paris, Paris, France; ⁸UF de génétique clinique, Centre de référence anomalies du développement et syndromes malformatifs, Hôpital Trousseau, Assistance Publique-Hôpitaux de Paris, Paris, France; ⁹Service de Génétique Moléculaire et Génomique, Centre Hospitalier Universitaire de Rennes, Rennes, France; ¹⁰Service de Génétique Clinique, Centre Hospitalier Régional Universitaire de Lille, Lille, France; ¹¹UF de génomique du développement, Centre de Génétique Moléculaire et Chromosomique, Hôpital de la Pitié-Salpêtrière, Assistance Publique-Hôpitaux de Paris, Paris, France; ¹²Service de Génétique Clinique, CLAD Ouest, Centre Hospitalier Universitaire de Rennes, Rennes, France; ¹³Département de Génétique, Hôpital Robert Debré, Assistance Publique-Hôpitaux de Paris, Paris, France; ¹⁴Laboratoire de Génétique Moléculaire · Service de génétique · CHRU de Tours - Hôpital Bretonneau, Tours, France; ¹⁵Service de Génétique Médicale, Centre Hospitalier Universitaire de Nantes, Nantes, France; ¹⁶Institut de Génétique Médicale, Centre Hospitalier Régional Universitaire de Lille, Université de Lille, Lille, France; ¹⁷Service de Médecine Génomique des Maladies Rares - GHU Necker- Enfants Malades, Assistance Publique-Hôpitaux de Paris, Paris, France; ¹⁸Service de Génétique-Unité de Génétique Clinique, Centre Hospitalier Universitaire de Rouen, Rouen, France; ¹⁹UF de Foetopathologie et Génétique, Centre Hospitalier Universitaire de Nantes, Nantes, France; ²⁰Service de Néphrologie Pédiatrique, Hôpital Necker-Enfants Malades, Assistance Publique-Hôpitaux de Paris, Université Paris Cité, Centre de Référence MARHEA, Paris, France; ²¹Centre de Référence des Malformations et Maladies Congénitales du Cervelet, Département de Génétique et Embryologie Médicale, Assistance Publique-Hôpitaux de Paris, Hôpital Trousseau, Paris, France; ²²Pôle De Biologie Pathologie Génétique, Hôpital Jeanne de Flandre, Centre Hospitalier Régional Universitaire De Lille, Lille, France; ²³Laboratoire de Biochimie et Biologie Moléculaire, Centre Hospitalier Régional Universitaire de Lille, Lille, France; ²⁴Laboratoire de Biologie Médicale multisites SeqOIA – PFMG2025; ²⁵Service de Médecine Génomique des Maladies Rares, Hôpital Necker-Enfants Malades, Assistance Publique-Hôpitaux de Paris, Paris, France; ²⁶Service De Génétique Clinique, Centre Hospitalier Régional Universitaire De Lille, Lille, France; ²⁷Département de génétique et Centre de référence anomalies du développement, CHU de Rouen, Rouen, France; ²⁸Service de Médecine Génomique des Maladies de Système et d’Organe, Hôpital Cochin, AP-HP Centre-Université de Paris Cité, Hôpital Cochin, Assistance Publique-Hôpitaux De Paris, Paris, France; ²⁹Service De Génétique, Centre Hospitalier Universitaire De Tours, Tours, France; ³⁰Université De Tours, INSERM, Tours, France; ³¹Unité Fonctionnelle De Médecine Génomique Et Génétique Clinique, Hôpital Jean Verdier, Assistance Publique Des Hôpitaux De Paris, Université Sorbonne Paris Nord, Paris, France; ³²Départemet de génétique, Centre de reference anomalies du développement, Centre Hospitalier Universitaire De Caen, Caen, France; ³³Service de Génétique, Hôpital Robert-Debré, Assistance Publique-Hôpitaux de Paris, Paris, France; ³⁴Service de Génétique Clinique, Centre Hospitalier Universitaire de Rennes, Rennes, France; ³⁵Génétique, Centre Hospitalier Universitaire D’Angers, Université d’Angers, Angers, France; ³⁶Département de Génétique, Hôpital Armand Trousseau, Assistance Publique-Hôpitaux de Paris City: Paris, Université Sorbonne Université, Paris, France; ³⁷Centre Hospitalier Régional Universitaire de Lille, Institut de Génétique Médicale, Université, Lille, France; ³⁸Service de Génétique Clinique, Hôpital Jeanne de Flandre, CLAD Nord, Lille, France; ³⁹Service de Médecine Génomique des Maladies Rares, UF MP5, Hôpital Necker-Enfants Malades, Assistance Publique-Hôpitaux de Paris, Université Paris Cité, Paris, France

Background/Objectives: Prenatal cytogenetics and chromosomal microarray (CMA) solve 40-45% of congenital anomalies and exome sequencing (ES) improves diagnostic yield in 26-36% of cases. The study aims to describe the input of genome sequencing (GS) in a cohort of postmortem fetuses/newborns with structural anomalies and to compare the efficiency of GS versus prenatal CMA + ES.

Methods: We collected information of 482 index cases sequenced at the French Laboratory SeqOIA (2020-April 2023) including terminations of pregnancy (418), in-utero deaths (39), fetal demises (4), and deceased newborns (21). Prior to GS: all had CMA, pathological examination; some cases had ES (39) or targeted analysis (59).

Results: GS solved 162 cases (34%). Chromosomal abnormalities (5) included mosaic imbalances in fetal tissue (3) - absent in amniocytes (2) or villi (1) – and balanced inversion (1). Monogenic disorders (157) involved 128 disease genes due to structural variants (15), variants in coding (194), and in non-coding regions (2). Eleven solved cases (7%) had previous ES. Eleven variants initially classified as of uncertain significance (VUS) were reclassified, mainly following RNA analysis (7). GS was non conclusive in 298 cases (62%), and with a VUS in 22 cases (4%).

Conclusion: GS is effective for etiology determination of fetal structural anomalies. Overpowering CMA + ES by identifying non coding variants, balanced structural variants, CNVs below the resolution of CMA, and overcoming limitations of ES detection, GS should be rapidly proposed in prenatal diagnosis. Fetal pathology is crucial for deep phenotyping, genotype/phenotype correlations, and tissue sampling/analysis. New strategies are needed for non conclusive cases.

Grants:

Conflict of Interest: None declared

C27.6 Non-invasive cell-free DNA methylation profiling of pregnant women for the diagnosis and early risk determination of preeclampsia

Machteld Baetens¹, Bram Van Gaever ², Stephanie Deblaere³, Andries De Koker⁴, Nico Callewaert⁴, Kristien Roelens³, Ellen Roets³, Isabelle Dehaene³, Björn Menten⁵

¹Ghent University Hospital, Center for Medical Genetics, Ghent, Belgium; ²Ghent University, Medical Genetics, Ghent, Belgium; ³Ghent University Hospital, Women’s Clinic, Ghent, Belgium; ⁴Flanders Institute for Biotechnology, Ghent, Belgium; ⁵Ghent University Hospital - Ghent University, Medical Genetics, Ghent, Belgium

Background/Objectives: Aberrant embryo implantation and suboptimal placentation can lead to (severe) pregnancy complications, like preeclampsia. Current identification of high-risk pregnancies relies on combining risk factors, biomarkers, and ultrasound examinations, a relatively inaccurate approach. An improved first trimester screening could enhance understanding placental development and pathology, to provide better prenatal care. Given the identification of disease-associated methylation changes in cfDNA, this study explores this potential of prenatal, non-invasive methylation screening.

Methods: We performed genome-wide cfDNA methylation profiling to identify significantly differentially methylated regions (DMRs) between uneventful pregnancies and those complicated with preeclampsia, with two objectives: diagnose preeclampsia at time of maternal symptoms and predict preeclampsia during the pre-symptomatic first trimester. Furthermore, we performed a longitudinal cfDNA methylation profiling of women with a complicated pregnancy and with a normal-term birth to investigate methylation evolution and to identify differences associated with placental dysfunction.

Results: We detected DMRs in pregnancies complicated with preeclampsia as early as 12 weeks of gestation. DMRs could be categorized into two groups, with an observed difference in genomic methylation patterns between early and late pregnancy. We were able to build two separate classification models which performed well in predicting or diagnosing preeclampsia on a small set of validation samples. Investigation of methylation patterns throughout the pregnancy yielded several methylation trends that showed to be significantly different between unaffected and affected pregnancies.

Conclusion: This study demonstrated the efficacy of non-invasive cfDNA methylation profiling in understanding pathology of preeclampsia. Our results offer promising potential for integrating this approach into routine prenatal care.

Grants: Vlaams Industrieel Onderzoeksfonds (F2020/IOF_STarTT035)

Conflict of Interest: None declared

C28.1 Integration of protein language models, regulatory CNNs and other nucleotide-level scores improves genome-wide variant prediction of CADD

Max Schubach ¹, Thorben Maass², Lusiné Nazaretyan¹, Sebastian Röner¹, Martin Kircher^1;2

¹Berlin Institute of Health at Charité – Universitätsmedizin Berlin, Exploratory Diagnostic Sciences, Berlin, Germany; ²University Hospital Schleswig-Holstein, University of Lübeck, Institute of Human Genetics, Lübeck, Germany

Background/Objectives: Machine Learning-based scoring and classification of genetic variants aids the assessment of clinical findings and is employed to prioritize variants in diverse genetic studies and analyses. Combined Annotation-Dependent Depletion (CADD) is one of the first methods for the genome-wide prioritization of variants across different molecular functions and has been continuously developed and improved since its original publication.

Methods: Here, we present our most recent release, CADD v1.7. We explored and integrated new annotation features, among them protein language model scores (Meta ESM), regulatory variant effect predictions (from sequence-based convolutional neural networks) and sequence conservation scores (Zoonomia). We evaluated the new version on data sets derived from ClinVar and gnomAD. For coding effects, we tested CADD on 31 Deep Mutational Scanning (DMS) datasets, and, for regulatory effect prediction, we used saturation mutagenesis reporter assay data of promoter and enhancer sequences.

Results: The individual inclusion of new features in CADD improved performance on variant prioritization, respectively. Further, the integration of all features improved the overall performance, and we released the improved model as CADD v1.7. As with previous releases, all data sets, genome-wide CADD v1.7 scores, scripts for on-site scoring and an easy-to-use webserver are provided via https://cadd.bihealth.org/.

Conclusion: With v1.7 we moved CADD further into the direction of prioritizing regulatory variants effects and deep learning models for protein and regulatory grammar. We conclude that CADD highly benefits from these sequence models, keeping it up-to-date with the latest research.

Grants: Deutsche Forschungsgemeinschaft (DFG) [464313370]; Impact of Genomic Variation on Function/NHGRI [1UM1HG011966-01];

Conflict of Interest: None declared

C28.2 NERINE: Network-based association test for rare variants in complex traits and diseases

Sumaiya Nazeen ^1;2, Xinyuan Wang², Autumn Morrow^1;2, Vikram Khurana², Shamil Sunyaev^1;3

¹Harvard Medical School, Biomedical Informatics, Boston, United States; ²Brigham and Women’s Hospital, Neurology, Boston, United States; ³Brigham and Women’s Hospital, Medicine/Division of Genetics, Boston, United States

Background: Rare protein-coding variants provide a direct path from genetics to biology but are difficult to detect. Traditional single-variant and gene-level tests have limited power. A promising strategy to increase power is to combine variants from multiple genes in a network. However, gene networks can have noisy definitions, with many genes having no effect or varying effects in different directions, posing additional challenges.

Methods: We present NERINE, a new statistical framework that infers the effect of rare variants aggregated across a gene network on a trait using a maximum-likelihood framework. We encode gene-gene relationships in a symmetric positive-definite matrix that can accommodate data from different screens. Genes can have either zero effect or varying degrees of trait-increasing and trait-decreasing effects. We test for network significance by nested hypothesis testing, providing an asymptotic p-value.

Results: NERINE outperforms existing methods in simulations, especially on noisy networks and imbalanced case-control datasets. It also helps us select the most relevant functional screen for a trait. When applied to LDL, HDL, and early MI phenotypes in the UK Biobank, NERINE recovers networks of lipid genes with expected directions of effects. Additionally, it identifies a novel gene module involved in Parkinson’s disease in UKBB and AMP-PD datasets, which we validate in CRISPRi experiments in a synucleinopathy iPSC model.

Conclusion: Overall, NERINE can identify novel disease-associated gene networks with rare variant burden, provide evidence of a screen’s relevance to a given disease, and facilitate a mechanistic understanding of the disease process.

Grants: This project is supported by grants NIH R35-GM127131 and ASAP 124423.

Conflict of Interest: Sumaiya Nazeen Full-time postdoctoral fellow at Harvard Medical School, Consultant for Dacapo Brainscience, Inc., Xinyuan Wang Full-time postdoctoral fellow at Brigham and Women’s Hospital, Consultant at Dacapo Brainscience, Inc., Autumn Morrow Full-time research assistant at Brigham and Women’s Hospital, Vikram Khurana: None declared, Shamil Sunyaev: None declared

C28.3 LDAK-PBAT: A Pathway-Based Analysis Tool for Decoding the Genetics of Complex Diseases

Takiy Berrandou¹, Doug Speed¹

¹Aarhus University, Center for Quantitative Genetics And Genomics (QGG), Aarhus, Denmark

Background/Objectives: The quest for personalized medicine necessitates sophisticated analysis tools capable of deciphering complex disease genetics. We introduce LDAK-PBAT, a novel tool that offers a holistic systems biology approach by simultaneously estimating variant effects.

Methods: Employing a heritability-based framework, LDAK-PBAT collectively analyzes multiple genes within pathways, refined through simulations and validated against datasets like the UK Biobank. Its performance was benchmarked against established tools, including MAGMA.

Results: LDAK-PBAT demonstrates exceptional accuracy in pathway heritability estimation, achieving RMSE values as low as 0.001084 and correlation coefficients up to 0.98. Comparative performance analysis reveals its enhanced sensitivity and precision, outperforming MAGMA and Hypergeometric testing with an F1 score of 0.734 versus 0.636 and 0.570, respectively. At the stringent threshold of 2.78e-06, it maintains an F1 score of 0.518, significantly higher than MAGMA’s 0.206. In analyzing 6000 pathways across 10 UK Biobank traits at the Bonferroni threshold, LDAK-PBAT identifies 344 significant pathways, markedly surpassing MAGMA’s 6, and demonstrating its superior pathway detection capabilities. Validation across increasing sample sizes further confirms LDAK-PBAT’s efficacy, with a significant increase in pathways retaining significance, affirming its reliability in genetic pathway analysis. Moreover, its robustness against variations in SNP lists and reference panels underscores the tool’s consistent performance.

Conclusion: LDAK-PBAT emerges as a potent tool for genetic research, offering advanced insights into complex diseases. Its precise estimation of pathway enrichment and heritability, along with superior performance in detecting significant pathways, underscores its potential to revolutionize genetic analysis and further the advancement of personalized medicine.

Conflict of Interest: None declared

C28.4 Adjusting for environmental factors in multi-ancestry meta-regression improves power to detect genetic associations

Siru Wang ¹, Oyesola Ojewunmi², Tinashe Chikowore^3;4;5, Abram Kamiza^6;7;8, Michele Ramsay⁶, Andrew Morris⁹, Segun Fatumo^2;7, Jennifer Asimit¹

¹University of Cambridge, MRC Biostatistics Unit, Cambridge, United Kingdom; ²London School of Hygiene and Tropical Medicine, Department of Non-Communicable Disease Epidemiology, London, United Kingdom; ³University of the Witwatersrand, MRC/Wits Developmental Pathways for Health Research Unit, Department of Paediatrics, Faculty of Health Sciences, Johannesburg, South Africa; ⁴Brigham and Women’s Hospital, Channing Division of Network Medicine, Boston, United States; ⁵Harvard Medical School, Boston, United States; ⁶University of the Witwatersrand, Sydney Brenner Institute for Molecular Bioscience, Faculty of Health Sciences, Johannesburg, South Africa; ⁷MRC/UVRI and LSHTM, The African Computational Genomic (TACG) Research Group, Entebbe, Uganda; ⁸Malawi Epidemiology and Intervention Research Unit, Lilongwe, Malawi; ⁹The University of Manchester, Centre for Genetics and Genomics Versus Arthritis, Centre for Musculoskeletal Research, Manchester, United Kingdom

Background/Objectives: Meta-analysis of genome-wide association studies (GWAS) across diverse populations offers gains in power to identify associations with small effect sizes through sample size increases. Common models for meta-analysis are grouped into fixed-effects (FE), random-effects (RE), and multi-ancestry meta-regression (MR-MEGA). However, with the growing number of genome-wide association (GWAS) data collected from highly diverse populations, it is likely for heterogeneous effect sizes across the populations to be correlated with not only ancestry but also environmental exposures, such as dietary and lifestyle factors. The current models cannot easily accommodate these factors in a meta-analysis framework that uses GWAS summary statistics.

Methods: To address this challenge, we present an environment-adjusted meta-regression model (env-MR-MEGA) to detect genetic variant associations by adjusting for differing environmental exposures between populations. Additionally, env-MR-MEGA quantifies the extent of heterogeneity due to ancestral and environmental effects.

Results: Through a series of simulations, we demonstrate that env-MR-MEGA increases power over MR-MEGA as high as 16% to detect genetic associations. In genetic association studies with LDL cholesterol in 20,000 individuals across twelve sex-stratified GWAS from Africa, adjusting for sex and overweight status within env-MR-MEGA leads to the detection of three novel loci, of which one has evidence of heterogeneity due to sex and overweight status, and another for ancestral heterogeneity.

Conclusion: The proposed meta-analysis method provides a way to adjust and test for environmental effects using summary-level data, making it a useful tool for meta-analyses without the need to share individual-level data.

Grants: Funded by an MRC-NIHR Better Methods, Better Research grant (MR/W02098X/1).

Conflict of Interest: None declared

C28.5 Understanding nurture via untransmitted genetic alleles

Robin J. Hofmeister ^1;2;3, Liza Darrous^1;2;3;4, Zoltan Kutalik^1;2;3

¹Unisanté I Policlinic Medicine Générale Du Bugnon, Lausanne, Switzerland; ²UNIL - Université de Lausanne, Lausanne, Switzerland; ³SIB Swiss Institute of Bioinformatics, Lausanne, Switzerland; ⁴University of Geneva, Geneva, Switzerland

Background/Objectives: Studies distinguishing transmitted and untransmitted genetic variants revealed that the genetic basis of educational attainment (EA) is substantially driven by the effect of parental alleles acting through parental rearing. While most studies focussed on accounting for these indirect genetic effects (IGEs), our aim is to accurately estimate them, in order to establish parental nurturing effects via Mendelian Randomisation (MR).

Methods: To boost sample size, we introduce a novel approach for estimating IGEs without the need for parental information. For each target individual, we identify close relatives and pinpoint Identity-By-Descent sharing among them to infer untransmitted parental alleles. These alleles are then employed to estimate associations with various offspring phenotypes. Finally, we employ MR to understand how parental characteristics can influence offspring traits.

Results: Using UK Biobank data, we inferred untransmitted alleles for ~130’000 individuals (vs ~5’000 parent-offspring pairs). The inference accuracy depends on relatedness: median R-squared of 0.81, 0.33, and 0.25 for inference from sibling, second-degree and third-degree relatives, respectively. We tested associations between a range of traits and untransmitted parental alleles and successfully replicated these associations using 178’000 sibling pairs from 19 cohorts. Plugging these estimates into MR, we identified a widespread impact of parental characteristics on offspring traits, notably highlighting the influence of parental obesogenic environment and socio-economic status on offspring EA and body mass index.

Conclusion: Overcoming the traditional requirement for parental information by leveraging inter-individual relatedness in biobanks allows us to evaluate nurturing effects at an unprecedented scale.

Grants: Swiss National Science Foundation 310030-189147

Conflict of Interest: None declared

C28.6 Multi-trait reverse regression for gene burden and rare variant association tests.

Andrey Ziyatdinov ¹, Joseph Herman¹, Benjamin Geraghty¹, Karl Landheer¹, Joelle Mbatchou¹, Jonathan Marchini¹

¹Regeneron Genetics Center, Tarrytown, NY, United States

Background/Objectives: Understanding the genetic basis of complex traits often requires analyzing multiple interrelated phenotypes. Most genetic association studies are performed for a single quantitative or binary trait at a time. However, joint analysis of multiple traits can boost power and aid interpretation. Existing multi-trait approaches have focused on analysis of common variants, with gene burden and rare variant multi-trait tests having received less attention. The advent of large scale exome sequencing studies of deeply phenotyped cohorts motivates the need for new approaches.

Methods: We have developed a reverse regression multi-trait test that works on rare variants, gene burden and common variants, and on any combination of quantitative and binary traits. The test is implemented within the REGENIE framework so utilizes a whole-genome regression model that accounts for population structure, relatedness and polygenicity. To control Type I error at tests of rare variants and binary traits with imbalanced case control ratios, we use Firth logistic or proportional odds models fitted with a robust pseudo-response step-halving algorithm.

Results: We use UK Biobank imputed genotype and exome sequencing data in up to 414,000 individuals together with simulated phenotypes to demonstrate the Type I error control and power of this approach. We further illustrate the approach using multiple phenotypes derived using MRI and DXA images from the UK Biobank imaging study.

Conflict of Interest: Andrey Ziyatdinov Regeneron Genetics Center or Regeneron Pharmaceuticals, Regeneron Genetics Center or Regeneron Pharmaceuticals, Joseph Herman Regeneron Genetics Center or Regeneron Pharmaceuticals, Regeneron Genetics Center or Regeneron Pharmaceuticals, Benjamin Geraghty Regeneron Genetics Center or Regeneron Pharmaceuticals, Regeneron Genetics Center or Regeneron Pharmaceuticals, Karl Landheer Regeneron Genetics Center or Regeneron Pharmaceuticals, Regeneron Genetics Center or Regeneron Pharmaceuticals, Joelle Mbatchou Regeneron Genetics Center or Regeneron Pharmaceuticals, Regeneron Genetics Center or Regeneron Pharmaceuticals, Jonathan Marchini Regeneron Genetics Center or Regeneron Pharmaceuticals, Regeneron Genetics Center or Regeneron Pharmaceuticals

C29.1 GestaltMatcher Database - A global reference for the phenotypic variability of rare diseases in humans

Hellen Lesmann ¹, Gholson Lyon², Pilar Caro³, Ibrahim Abdelrazek⁴, Alexander Hustinx¹, Hannah Klinkhammer¹, Shahida Moosa⁵, Jean Tori Pantel⁶, Annabelle Arlt¹, Elaine Marchi², Hannah Weiland¹, Merle ten Hagen¹, Sophia Kaptain¹, Alexej Knaus¹, Behnam Javanmardi¹, Rami Abou Jamra⁷, Rebekah Waikel⁸, Sebastien Küry⁹, Frédéric Ebstein⁹, Sylvia Safwat⁴, Franziska Schnabel⁷, Artem Borovikov¹⁰, Claudio Graziano¹¹, Amira Nabil⁴, Vincent Strehlow⁷, Clara Velmans¹², Miriam Elbracht¹³, Cordula Knopp¹³, Annette Uwineza¹⁴, Luisa Averdunk¹⁵, Tinatin Tkemaladze¹⁶, Himanshu Goel¹⁷, Stefan Barakat¹⁸, Nergis Güzel¹³, Julia Suh¹³, Theresa Brunet¹⁹, Sabine Rudnik²⁰, Christian Schaaf³, Benjamin D. Solomon⁸, Christoffer Nellaker²¹, Alain Verloes²², Koen Devriendt²³, Martin Mücke⁶, Aleksandra Jezela²⁴, Karen W. Gripp²⁵, Christian Netzer¹², Ebtessam Abdallah⁴, Markus M. Nöthen²⁶, Peter Krawitz¹, Tzung-Chien Hsieh¹

¹Institute for Genomic Statistics and Bioinformatics, University Hospital Bonn, Bonn, Germany; ²Department of Human Genetics, New York State Institute for Basic Research in Developmental Disabilities, Staten Island, New York, United States; ³Institute of Human Genetics, Heidelberg University, Heidelberg, Germany; ⁴Department of Human Genetics, Medical Research Institute, Alexandria University, Alexandria, Egypt; ⁵Division of Molecular Biology and Human Genetics, Stellenbosch University and Medical Genetics, Tygerberg Hospital, Stellenbosch, South Africa; ⁶Institute for Digitalization and General Medicine, University Hospital RWTH Aachen, Aachen, Germany; ⁷Institute of Human Genetics, University of Leipzig Medical Center, Leipzig, Germany; ⁸Medical Genomics Unit, Medical Genetics Branch, National Human Genome Research Institute, Bethesda, United States; ⁹Service de Génétique Médicale, Nantes Université, CHU Nantes, Nantes, France; ¹⁰Research Centre for Medical Genetics (RCMG), Moscow, Russian Federation; ¹¹Medical Genetics Unit, Ausl Romagna, Cesena, Italy; ¹²Institute of Human Genetics, University of Cologne, Faculty of Medicine and University Hospital Cologne, Cologne, Germany; ¹³Institute for Human Genetics and Genomic Medicine, Medical Faculty, RWTH Aachen University, Aachen, Germany; ¹⁴Department of Clinical Biology, University of Rwanda, Kigali, Rwanda; ¹⁵Clinic for General Paediatrics, Neonatology and Paediatric Cardiology, University Hospital Düsseldorf, Düsseldorf, Germany; ¹⁶Department of Molecular and Medical Genetics, Tbilisi State Medical University, T’bilisi, Georgia; ¹⁷Hunter Genetics, Waratah Campus Newcastle, Newcastle, Australia; ¹⁸Department of Clinical Genetics, Erasmus MC University Medical Center, Rotterdam, Netherlands; ¹⁹Institute of Human Genetics, Technische Universität München, München, Germany; ²⁰Institute Of Human Genetics, Medical University Innsbruck, Innsbruck, Austria; ²¹Big Data Institute, Li Ka Shing Centre for Health Information and Discovery, Nuffield Department of Women’s & Reproductive Health, University of Oxford, Oxford, United Kingdom; ²²Department of Clinical Genetics, Robert-Debré Hospital, Paris, France; ²³Center for Human Genetics, Uni Leuven, Leuven, Belgium; ²⁴Department of Genetics and Clinical Immunology, National Institute of Tuberculosis and Lung Diseases, Warsaw, Poland; ²⁵Division of Medical Genetics, A.I. du Pont Hospital for Children/Nemours, USA, Wilmington, United States; ²⁶Institute of Human Genetics, University Hospital Bonn, Bonn, Germany

Background: Characteristic facial dysmorphism is more difficult for both humans and AI to recognize when there are no cases of the same ethnicity to learn from. Although portraits in the literature are predominantly of Europeans, most children with rare disorders are now born in Asia and Africa, indicating a need for action.

Methods: We designed GestaltMatcher Database (GMDB) - a platform compliant with the FAIR Data principles. Each case in GMDB can get a Digital Object Identifier (DOI), becoming a citable micro-attribution. We curated data of individuals with dysmorphic features from 2,224 publications to initialize GMDB. Then, we encouraged clinician scientists and patient support groups from everywhere to contribute cases with molecular diagnoses of diverse ethnic backgrounds. Finally, we analyzed the impact of the novel data on the performance of GestaltMatcher AI.

Results: We have compiled a collection of 10,189 frontal images from 7,695 patients representing 683 disorders. These data contain 1,409 novel medical images that registered users can query. Our global collaborators increased the data of underrepresented ethnicities to 20% from Asia and 5% from Africa. Our analysis with next-generation phenotyping (NGP) indicate that higher ethnic diversity enhances the overall performance of GestaltMatcher AI. For all ethnic groups, incorporating non-European cases into the training set increased the top-1-accuracy by 32% and top-5-accuracy by 13%.

Conclusion: GMDB is a resource for medical image data of individuals with rare disorders from diverse ethnic backgrounds that can contribute to the training of dysmorphologists and NGP technology worldwide.

Grants: Eva Luise und Horst Köhler Stiftung. Arbeitsgemeinschaft für Gendiagnostik

Conflict of Interest: None declared

C29.2 The true recurrence risk: parental mosaicism in a large family cohort with de novo variants

Sonja Neuser ¹, Natalie Ahmad¹, John Wiedenhöft¹, Denny Popp¹, Johannes Lemke¹, Knut Krohn², Ulrike Klotz³, Toralf Kirsten³, Maike Karnstedt¹, Julia Hentschel¹, Stephan Drukewitz¹, Rami Abou Jamra¹

¹Institute of Human Genetics, University of Leipzig Medical Center, Leipzig, Germany; ²Core Unit DNA Technologies, Medical Faculty, University Leipzig, Leipzig, Germany; ³Medical Data Science, University of Leipzig Medical Center, Leipzig, Germany

Background/Objectives: Parents of children with genetic disorders due to de novo variants are usually counselled on a recurrence probability estimate of 1-3% for further affected siblings, while the actual probability varies between 0 and 50%. This discrepancy is well-known, but barely investigated.

Methods: We prospectively enrolled 138 families, where a pathogenic putative de novo variant had previously been identified in an affected child. Covering two germ layers, we collected blood (n = 276), buccal (n = 228) and nail samples (n = 229) of both parents and a semen sample of the father (n = 92). We performed deep sequencing achieving >1000x coverage after unique molecular identifiers (UMI) collapsing for all variants across all tissues (n ≈ 130,000 measured values). Sophisticated Bayesian analyses regarding low mosaicism, correlations of allele fractions in different tissues, and probability models are ongoing. Using Nanopore long read sequencing, variants were phased to identify the parental allele of origin.

Results: We identified clear (>1%) mosaicism for six variants in either the mother (n = 1) or the father (n = 5). In two cases, only the semen sample was affected. The alternative allele fraction varied from 1.3% to 23%. Phasing revealed one quarter of variants are located on the maternal allele and, three quarters on the paternal allele.

Conclusion: Our observation suggests that parental mosaicism could be less common than assumed. For families with a child carrying a de novo variant, testing the semen - and not the blood of the father - is an indispensable analysis to give a reliable answer.

Grants: DFG AB393/11-1 and NE2706/2-1

Conflict of Interest: Sonja Neuser DFG NE2706/2-1, Natalie Ahmad: None declared, John Wiedenhöft: None declared, Denny Popp: None declared, Johannes Lemke PTC Therapeutics, GRIN Therapeutics, UCB Pharma, Angelini Pharma, Knut Krohn: None declared, Ulrike Klotz: None declared, Toralf Kirsten: None declared, Maike Karnstedt: None declared, Julia Hentschel: None declared, Stephan Drukewitz: None declared, Rami Abou Jamra DFG AB393/11-1

C29.3 Towards population-scale transcriptomics — RNA-seq using decades-old blood spot samples

Victor Yakimov ¹, Andres Ingason², Christine Søholm¹, Marie Bækvad-Hansen¹, Preben Bo Mortensen², merete nordentoft², Ole Mors², Anders Børglum², David Michael Hougaard¹, Bjarke Feenstra³, Thomas Werge², Alfonso Buil², Jonas Bybjerg-Grauholm¹

¹Statens Serum Institut, Department for Congenital Disorders, København, Denmark; ²The Lundbeck Foundation Initiative for Integrative Psychiatric Research; ³Statens Serum Institut, Department of Epidemiology Research, Copenhagen, Denmark

Background/Objectives: Biobanks contain millions of dried blood spot (DBS) samples obtained for routine neonatal screening. These samples create an opportunity to study biological signals preceding clinical presentation of disease. To date, no large-scale RNA-seq study in DBS samples has been published. Here, we show that such studies are feasible in samples stored at -20ºC for up to four decades.

Methods: We examine the quality of RNA-seq data obtained from DBS by inspecting transcriptomic profiles as a function of time they have spent in storage, and replicate differential expression analysis on sex, as well as expression-quantitative trait loci (eQTL) analysis performed by the Genotype-Tissue Expression (GTEx) consortium.

Results: We find no evidence that fewer genes can be identified in our samples compared to RNA-sequencing done in fresh whole blood samples. We also show that robust biological signals can be detected by closely replicating two high-quality studies performed by GTEx. While we find evidence of transcript degradation associated with a sample’s age, we find that it does not impact our ability to detect meaningful biological signals, and it can be controlled for by using standard statistical approaches.

Conclusion: Our results open the door to a new wave of experiments that can measure transcriptomic profiles in the blood of newborns and integrate that information with genetic data and life-time medical history of thousands of individuals.

Grants: This study was supported by The Lundbeck Foundation. This research has been conducted using the Danish National Biobank resource, supported by the Novo Nordisk Foundation.

Conflict of Interest: None declared

C29.4 Harnessing RNA-Sequencing in rare disease diagnostics: Insights from a multi-regional pilot project.

Carolina Jaramillo Oquendo ¹, Matthew Lyon², Dave Bunyan², Jacob Abbott², Jessica Woodley³, Samantha Court³, Jennie Bell³, Andrew Douglas⁴, Sarah Willis⁵, N Simon Thomas², Anthony Williams⁵, Sarah Ennis¹, Diana Baralle¹

¹University of Southampton, Southampton, United Kingdom; ²Wessex Genomics Laboratory Service, Salisbury, United Kingdom; ³West Midlands Regional Genetics Laboratory, Birmingham, United Kingdom; ⁴Oxford Centre for Genomic Medicine, Oxford, United Kingdom; ⁵Central & South Genomic Medicine Service Alliance, United Kingdom

Background/Objectives: Whole exome and genome sequencing (WES/WGS) have revolutionized the genetic testing of rare disease, however, even with WGS, nearly half of patients remain without a conclusive diagnosis. In 2022, the Central and South Genomic Medicine Service Alliance (GMSA) initiated a project with regional collaborations aiming to evidence RNA-based testing as a technique that increases diagnostic yield and develop proposals for how RNA-based testing could be incorporated into NHS England’s National Genomics Test Directory.

Methods: Clinical leads identified three patient cohorts: patients with splicing variants of uncertain significance (VUS) [n = 76], patients with no genetic diagnosis following DNA testing [n = 41] and patients with a recessive disorder and only one pathogenic variant identified [n = 3]. RNA was extracted from blood collected in PAXgene tubes and sent for sequencing in batches of 16. FASTQ files were aligned to the reference genome using STAR. Aligned files were visually inspected using the Integrative Genomics Viewer to detect splicing abnormalities.

Results: To date 39 VUS cases have been analysed, resolving 28 with RNA-Sequencing alone and 10 using RT-PCR. RNA-based testing aided in the re-classification of 68% of VUSs. For 33% of VUSs, RNA-sequencing identified complex splicing abnormalities missed by RT-PCR. In at least one case results enabled preimplantation genetic diagnosis. We will compare RNA-seq and RT-PCR, noting pros and cons, and share insights on infrastructure and data governance challenges.

Conclusion: We implemented a high-throughput RNA-sequencing workflow within the Central & South GMSA, with forthcoming steps aimed at collating results across regions to inform clinical integration.

Grants: NHSE, NIHR

Conflict of Interest: None declared

C29.5 Using Multiplexed Functional Data to Reduce Variant Classification Disparities in Populations Underrepresented in Genomic Medicine (virtual)

Moez Dawood ^1;2;3, Shawn Fayer^4;5, Mason Post⁵, Divya Kalra¹, Karynne Patterson⁴, Eric Venner^1;2, Lara Muffley^4;5, Douglas Fowler^4;5, Alan Rubin^6;7, Jennifer Posey², Sharon E Plon^1;2, James Lupski^1;2, Richard Gibbs^1;2, Lea Starita^4;5, Carla Daniela Robles Espinoza^8;9, Willow Coyote-Maestas^10;11, Irene Gallego Romero^12;13;14

¹Baylor College of Medicine, Human Genome Sequencing Center, Houston, United States; ²Baylor College of Medicine, Molecular and Human Genetics, Houston, United States; ³Baylor College of Medicine, Medical Scientist Training Program, Houston, United States; ⁴University of Washington, Genome Sciences, Seattle, United States; ⁵University of Washington, Brotman Baty Institute, Seattle, United States; ⁶The Walter and Eliza Hall Institute of Medical Research, Bioinformatics Division, Parkville, Australia; ⁷University of Melbourne, Medical Biology, Melbourne, Australia; ⁸Wellcome Sanger Institute, CASM, Hinxton, United Kingdom; ⁹Universidad Nacional Autónoma de México, Laboratorio Internacional de Investigación sobre el Genoma Humano, Ciudad de México, Mexico; ¹⁰University of California San Francisco, Bioengineering and Therapeutic Sciences, San Francisco, United States; ¹¹University of California San Francisco, Quantitative Biosciences Institute, San Francisco, United States; ¹²University of Tartu, Center for Genomics, Evolution and Medicine, Tartu, Estonia; ¹³St. Vincent’s Institute of Medical Research, Human Genomics and Evolution, Fitzroy, Australia; ¹⁴The University of Melbourne, School of BioSciences and Melbourne Integrative Genomics, Parkville, Australia

Background/Objectives: The saturation-nature of Multiplexed Assays of Variant Effects (MAVEs) may offer a systematic experimental approach to address variant classification disparities between different populations, especially for variants of uncertain significance (VUS).

Methods: We analyzed the distribution of clinical significance classifications in genome sequencing data from 245,394 individuals in All of Us v7 and replicated our findings in two independent population-sale databases of the Genome Aggregation Database (gnomAD), 123,709 exomes of gnomAD v2.1.1 and 51,535 genomes of gnomAD v3.1.2 (non v2). Then, we incorporated clinically calibrated MAVE data into the ClinicalGenome’s Variant Curation Expert Panel rules and automated VUS reclassification.

Results: Using two orthogonal statistical approaches, we show across all 3 databases a higher prevalence (p ≤ 5.95e-06) of VUS in individuals of non-European genetic ancestry across all medical specialties assessed. Further, in the non-European genetic ancestry group, higher rates of Benign or Likely Benign and variants with no clinical designation (p ≤ 2.5e-05) were found across many medical specialties, whereas Pathogenic or Likely Pathogenic assignments were increased in individuals of European genetic ancestry (p ≤ 2.5e-05).

Using publicly-accessible MAVE data, we then reclassified 80.3% of all VUS in BRCA1, TP53, and PTEN across the three databases and reclassified non-European VUS at a significantly higher proportion in comparison to European VUS (p ≤ 8.7e-06) effectively compensating for the original VUS disparity.

Conclusion: These findings demonstrate the potential of MAVEs to reduce disparities in variant classification for underrepresented populations to make genomic medicine more informative and inclusive for all.

Grants: NHGRI-U01HG011758-01, NHGRI-UM1HG011969-03S, CPRIT-RP210027

Conflict of Interest: None declared

C29.6 A benchmarking database of SNPs, indels and SVs created using a four-generation, 28-member pedigree sequenced with multiple technologies

Zev Kronenberg ¹, Cillian Nolan¹, David Porubsky², Katy Munson², William Rowell¹, Michelle Noyes², Brent Pedersen³, Sean Mcgee², Josh Smith², Cairbre Fanslow¹, primo baybayan¹, Nidhi Koundinya², William Harvey², Kendra Hoekzema², Jordan Knuth², Gage Garcia², Tom Mokveld¹, Egor Dolzhenko¹, Scott Watkins³, Deborah Neklason³, Aaron Quinlan³, Lynn Jorde³, Evan Eichler², Michael Eberle¹

¹PacBio, Menlo Park, United States; ²University of Washington, Genome Sciences, Seattle, United States; ³The University of Utah, Salt Lake City, United States

Background/Objectives: Accurate long-read sequencing makes it possible to survey the full spectrum of genetic variation. Developing and validating new variant calling methods requires a comprehensive benchmarking dataset but current benchmark datasets have been developed primarily from short-read technologies and cannot adequately quantify the performance of long-read methods.

Methods: We have created a more comprehensive variant truthset by utilizing the power of genetic inheritance within a four-generation family (CEPH 1463) sequenced with multiple technologies (PacBio, ONT, Illumina) from blood-derived DNA. Large kinship pedigrees allow us to unambiguously determine the genetic inheritance of the parental haplotypes. Filtering genotypes with the genetic inheritance provides a means to adjudicate variant accuracy for all the sequencing technologies in our study. Combining these ‘truth’ callsets from three distinct technologies creates a comprehensive truthset that is not affected by the limitations of any individual datatype.

Results: Our approach yielded a truthset containing over six million SNPs and indels, an 18% increase over the NA12878 Genome in a Bottle dataset. Benchmarking revealed HiFi-based variant calls exhibit the highest recall for SNVs and INDELs (0.985/0.952), compared to ONT (0.969/0.733) and Illumina (0.935/0.914). Expanding beyond small variants, we identified 327,226 non-reference tandem repeats and >29,000 structural variants that were inherited. Importantly, we increased the rate of de novo SNVs, STRs and SVs by more than 25% in previously inaccessible regions of the genome.

Conclusion: Combining technologies and the inheritance from a 28-member pedigree we built a much more comprehensive benchmarking dataset that will be a valuable resource for tool developers and geneticists alike.

Grants:

Conflict of Interest: Zev Kronenberg PacBio, and Phase Genomics, PacBio, Cillian Nolan PacBio, PacBio, David Porubsky: None declared, Katy Munson: None declared, William Rowell PacBio, PacBio, Michelle Noyes: None declared, Brent Pedersen: None declared, Sean Mcgee: None declared, Josh Smith: None declared, Cairbre Fanslow PacBio, PacBio, primo baybayan PacBio, PacBio, Nidhi Koundinya: None declared, William Harvey: None declared, Kendra Hoekzema: None declared, Jordan Knuth: None declared, Gage Garcia: None declared, Tom Mokveld PacBio, PacBio, Egor Dolzhenko PacBio, PacBio, Scott Watkins: None declared, Deborah Neklason: None declared, Aaron Quinlan: None declared, Lynn Jorde: None declared, Evan Eichler SAB: VariantBio, Michael Eberle PacBio, PacBio

C30.1 A unified genome constraint, pathogenicity, and pLoF model identifies new genes associated with epilepsy, autism, and schizophrenia

Oscar Aguilar ¹, Manuel Rivas²

¹Stanford University, Management Science & Engineering, Stanford, United States; ²Stanford University, Biomedical Data Science, Stanford, United States

Background/Objectives: Highly heterogeneous disorders like epilepsy, autism, and schizophrenia are thought to have strong genetic components. However, identifying these risk factors using whole-exome sequencing studies requires very large sample sizes and good signal-to-noise ratio in order to assess the association between rare variants in any given gene and disease.

Methods: We present a novel approach for predicting constraint in the human genome – sections of the genome where any mutation can cause a severe disorder. Through application of a Hidden Markov Model (HMM) to the Regeneron Genetics Center Million Exome dataset and the AllofUs whole genome sequencing data, we predict the probability of observing no variants across the population for each position in the genome. Next, we aggregate the constraint predictions by gene and assess its association to epilepsy, schizophrenia, and autism, and compare against published results. Finally, we extend our analysis model to incorporate pathogenicity predictions from AlphaMissense (AM) and pLoFs.

Results: We identified a set of (p < 1x10^-4) genes with stronger signals than previously published studies including KDM5B, KCNQ2, CACNA1A, CACNA1B, RYR2, and ATP2B2. Our models allow us to evaluate the contribution of constraint, protein structure based pathogenicity prediction from AM, and pLoFs jointly.

Conclusion: We showed that relatively simple sequence-dependent constraint prediction models can complement structure-based missense variant pathogenicity predictions and pLoFs for population cohort studies which require additional statistical power in the identification of gene-based signals for neurogenetic and psychiatric disorders. Finally, we explore the application of transformers to incorporate other features for constraint prediction.

Conflict of Interest: Oscar Aguilar Some of the computing for this project was performed on the Sherlock cluster., Graduate research assistant for Stanford University., Manuel Rivas Co Founder of Broadwing Bio, Some of the computing for this project was performed on the Sherlock cluster., M.A.R. is in part supported by National Human Genome Research Institute (NHGRI) under award R01HG010140, and by the National Institutes of Mental Health (NIMH) under award R01MH124244 both of the National Institutes of Health (NIH), Stanford University, Consultant for Curie.Bio and Mubadala Ventures

C30.2 Unlocking the genetic code of cerebral palsy: exploring clinical and genetic insights in modern era

Tracy Lau ¹, charles steward², Jozef Gecz³, david rowitch⁴, Jana Vandrovcova¹, Reza Maroofian¹, Henry Houlden¹