TRIM25 nonsense mutation (p.C168) as the probable cause of early‐onset autosomal dominant Alzheimer’s disease

Background Patients with early‐onset dementia (EOD) often have a high genetic burden, and EOD kindreds with an autosomal dominant pattern of inheritance (ADPI) are particularly powerful for gene identification studies. Methods We present the clinical phenotype and a whole‐exome sequencing (WES) study of a Spanish EOD family with ADPI, including five affected and two unaffected siblings. The proband case was negative for mutations in a Next Generation Sequencing panel of genes associated with neurodegenerative dementias. Candidate pathogenic mutations were prioritized according to frequency in population databases, impact/pathogenicity scores, and segregation (including a disease‐free second branch of the family). Final candidate genes were also examined in a set of Spanish‐origin FTD cases (n=583) and controls (n=493), plus in a Belgian WES dataset of AD (n=236) and FTD (n=269) cases. Results Affected siblings had an average disease onset of 60.2 years (range 56‐63 years). Two had a behavioral rather than an amnestic phenotype, developing some psychotic features and significant parkinsonism after treatment with risperidone. However, CSF biomarkers in one patient were indicative of an AD pathology. WES was conducted in four affected and one unaffected siblings (all of them APOEε 4,4). Variant prioritization pointed to 16 candidate variants, including three loss of function variants, two in‐frame deletions, and 11 missense. Additional genetic analysis of another unaffected sibling, a fifth affected sibling with incipient disease, plus four unaffected relatives from the second branch narrowed the list to two candidate genes: TRIM25 (p.C186*, CADD score 36) and EEA1 (p.V693A, CADD 18.2). TRIM25 is a cytoplasmic protein that functions as an ubiquitin E3 ligase and is involved in the innate immune response, while EEA1 is an endosomal trafficking protein. Potentially pathogenic (CADD>20) rare variants were found in the studied populations in EEA1 (0,7‐3,8%; Spanish‐Belgian) and TRIM25 (0‐0,7%). All except one were missense variants of uncertain significance. Importantly, in TRIM25 which is highly intolerable for loss of function mutations, we identified one additional nonsense mutation in TRIM25 in a frontal‐type dementia case. Conclusion We conclude that in this EOD family, the nonsense mutation in TRIM25 is the most likely causal mutation.