Anorexia Nervosa (AN) is a chronic psychiatric disorder whose pathophysiology is largely unknown but has a significant genetic component. The objective of this work was to identify rare genetic variants of predisposition for anorexia nervosa.

We studied 10 families of 15 affected women, 11 cases of restrictive ANR (ANR) and 4 cases of binge-purging AN, and analyzed in trio by new generation sequencing. The variants were filtered by quality controls (blanket, IGV viewer), segregation in familial forms, deleterious in silico prediction of variants (Cadd>15), and absence or rarity in unaffected controls and in the databases (Gnomad, “Déjà vu” Paris Descartes) (<1%).

The analysis of the de novo variants did not make it possible to identify deleterious candidate variants (with the exception of p.L206V, deleterious in the LUMAN/CREB3 gene) and recurrent ones. In index cases of familial forms, potential de novo variants have been identified. Dominant model analysis identified 349 potential variants distributed in 110 genes. Analysis of this list of genes by String, David and GeneOntology has identified several biological pathways that may be altered. Three major KEGG pathways could be identified: “neuroactive ligand-receptor interaction” 9 genes, “cholinergic synapse” 6 genes, and “dopaminergic synapse” 6 genes. In addition, the literature analysis identified 24 variants in genes contributing to neuropsychiatric disorders.

We identified specific damaging variant in several genes than can play a role in AN. Targeted analyzes of a large cohort of patients should be conducted to confirm our results. This study was supported by Fondation Maladies Rares (#11632) Program high throughput sequencing and rare diseases.