Abstract

Context: Neurofibromatosis type 1 (NF1) is an autosomal dominant genetic condition caused by loss of function variants in the tumor suppressor gene NF1. The gene encodes neurofibromin, a negative regulator of the RAS-MAPK pathway. NF1 is characterized by a highly variable expressivity. International studies of genotype-phenotype correlations have shown that a few NF1 pathogenic variants are associated either with a milder or with a more severe presentation of the condition. For the remaining variants, no significant correlation was evidenced. Heritability studies in NF1 families have identified a genetic component unlinked to the NF1 gene, possibly explaining the phenotypic variability observed in patients.

Objectives: The aim of this work was to identify the genetic factors linked to the NF1 gene or the modifier genes, implicated in the phenotypic variability of NF1, in the French cohort.

Methods: We exmained the phenotypic data recorded using a standardized questionnaire filled out by expert clinicians in the context (i) of three national hospital clinical research programs dedicated to the study of NF1 modifiers (NF-France cohort, 2002-2013) and (ii) of the routine molecular diagnosis performed until 2020 in the Service de Médecine Génomique at the Cochin hospital in Paris. NF1 pathogenic variants were studied using various methods to identify and delineate the causal genomic variations and to implement genotype-phenotype correlations. Additionally, the patients from the NF-France cohort were genotyped for more than 7 million common genomic variants and grouped into a discovery cohort (918 patients) and a replication cohort (415 patients) to perform a genome-wide association study (GWAS).

Results: (i) The genotype-phenotype correlation study for the NF1 point variants previously associated with a specific presentation of the condition, obtained results concordant with those of the literature. We confirmed a milder form of NF1 in patient with an inframe deletion of the Met992 or a missense variant at Met1149 or Arg1809. Those with a missense variant at positions Arg1276, Lys1423 or affecting codons 844 to 848 of neurofibromin have a higher risk of developing a severe condition. Our study demonstrated that patients with a complete deletion of the NF1 locus exhibit a particularly severe disease presentation. We developped a new tool based on the digital droplet PCR technology to quickly and efficiently characterize the recurrent deletions of the NF1 locus.

(ii) The GWAS revealed several genomic regions showing an association with the occurrence of cutaneous, subcutaneous, and plexiform neurofibromas. They included about 170 candidate modifier genes. Using a competition assay on isogenic WT and NF1-KO Schwann cells, we demonstrated that at least two of these genes (GAS1 and SPRED2) might have a modifier function on the proliferation of NF1-KO Schwann cells, which are at the origin of neurofibromas.

Conclusions and perspectives: Our results confirm previous genotype-phenotype correlation studies. The establishment of such correlations and the development of complementary molecular tools are key elements of clinical management and genetic counselling. Haploinsufficiency for genes neighbouring NF1 might explain the more severe phenotype identified in patients with large deletions encompassing NF1. The GWAS suggests several candidate modifier genes for neurofibromas. Setting up functional screening assays should enable us to confirm the involvement of these genes, eventually bringing about new therapeutic avenues for the management of neurofibromas.