Résumé
Objective Mutations in superoxide dismutase 1 gene ( SOD1) , encoding copper/zinc superoxide dismutase protein, are the second most frequent high penetrant genetic cause for amyotrophic lateral sclerosis (ALS) motor neuron disease in populations of European descent. More than 200 missense variants are reported along the SOD1 protein. To limit the production of these aberrant and deleterious SOD1 species, antisense oligonucleotide approaches have recently emerged and showed promising effects in clinical trials. To offer the possibility to any patient with SOD1-ALS to benefit of such a gene therapy, it is necessary to ascertain whether any variant of unknown significance (VUS), detected for example in SOD1 non-coding sequences, is pathogenic. Methods We analysed SOD1 mutation distribution after SOD1 sequencing in a large cohort of 470 French familial ALS (fALS) index cases. Results We identified a total of 27 SOD1 variants in 38 families including two SOD1 variants located in nearsplice or intronic regions of the gene. The pathogenicity of the c.358–10T>G nearsplice SOD1 variant was corroborated based on its high frequency (as the second most frequent SOD1 variant) in French fALS, the segregation analysis confirmed in eight affected members of a large pedigree, the typical SOD1-related phenotype observed (with lower limb onset and prominent lower motor neuron involvement), and findings on postmortem tissues showing SOD1 misaccumulation. Conclusions Our results highlighted nearsplice/intronic mutations in SOD1 are responsible for a significant portion of French fALS and suggested the systematic analysis of the SOD1 mRNA sequence could become the method of choice for SOD1 screening, not to miss these specific cases.