P20

Discovery of New Therapeutic Targets for Duchenne Muscular Dystrophy by Inference of Gene Regulatory Networks

E Guillot(1,5,6) E Mozin(3) Q Fort(1,5) B Robert(3) C Lièvre(3) S Luttrell(2,4) D L Mack(2) A Bonnaffoux(1,5) J B Dupont(3)

1:Centre de Recherche en Cancérologie de Lyon, Inserm U1052-CNRS UMR5286, Centre Léon Bérard, Université Claude Bernard Lyon 1, 69008 Lyon, France; 2:Institute for Stem Cell and Regenerative Medicine, University of Washington Department of Rehabilitation Medicine, Seattle, WA 98109, USA; 3:Nantes Université, INSERM, TARGET, F-44000 Nantes, France; 4:Curi Bio, Seattle, WA 98121, USA; 5:Département de Biologie Computationnelle, Centre Léon Bérard, 69008 Lyon, France; 6:Fondation Synergie Lyon, 69008 Lyon, France

Duchenne Muscular Dystrophy (DMD) is an X-linked muscular disorder caused by mutations in the DMD gene. Patients are usually diagnosed at 3-4 years of age but early phenotypes have been described in fetuses and in animal models before the symptoms appear. Progressive degeneration of DMD muscles leads to severe disabilities, wheelchair dependency and a median life expectancy of ~30 years. In animal models, gene therapy with adeno-associated virus (AAV) vectors carrying a micro-dystrophin transgene leads to an almost complete disease correction. In DMD patients however, despite promising intermediary results, high doses of AAV have limited efficacy and lead to serious adverse events. This highlights the need for innovative, patient-centered gene therapy strategies together with relevant models to identify new therapeutic targets. In this context, our team uses induced pluripotent stem cells (iPSCs) to understand the initiation of DMD during muscle development. We first evaluated the impact of dystrophin deficiency on the transcriptome at single-cell resolution during the differentiation of iPSCs into skeletal muscle. Our data showed that DMD iPSCs diverge from control cells at the somite stage and acquire a different transcriptomic profile. Using a retro-engineering approach together with an augmented learning algorithm, we integrated the scRNA-Seq time series with RNA half-life and proteomic data and we inferred a gene regulatory network (GRN) composed of 6 genes interacting in a genotype-dependent topology. Modulation of the genes within the DMD GRN using RNA interference is currently under investigation to validate their potential as therapeutic targets in future gene therapy applications.