Endocrine Abstracts

Introduction: ARMC5 germline and somatic inactivating mutations were discovered in patients treated by adrenalectomy for hypercortisolism due to primary bilateral macronodular adrenal hyperplasia (PBMAH). Since then, several ARMC5 germline variants have been described in PBMAH patients. Genetic alterations are spread all over ARMC5 coding sequence and many are missense variants. For them, geneticist conclusions are based on in silico predictions. As for now, no functional assay is routinely performed to study their pathogenic consequences. ARMC5 is considered as a tumor suppressor gene with proapoptotic function. Our center performs routine germline ARMC5 sequencing by next generation sequencing for patients all over the country. New missenses have been recently identified. We aimed to develop a functional test to confirm bioinformatics predictions.

Method: In a series of 352 French index cases, 10 new exonic missense variants, not classified as benign polymorphisms by SIFT^® and Polyphen2^® softwares have been recently found. For in vitro analysis, constructs encoding the potentially pathogenic ones are made in a pIRES vector. Transfections are performed in HEK293 cells to study recombinant ARMC5 protein levels by western blot and cell apoptosis using AnnexinV and propidium iodide staining for flow cytometry (FACS) analysis. Loss of the proapoptotic function of ARMC5 was considered as a marker of mutant pathogenicity.

Results: Seven of these variants were predicted as deleterious or possibly damaging by both SIFT^® and Polyphen2^®; for the three others, the two softwares predictions were discordant and they were not described in exomes databases. The known variants Leu548Pro, p.Leu331Pro, p.Cys139Arg were previously selected for in vitro studies and compared with the predicted benign p.Phe14Tyr variant and wild-type ARMC5. Kinetic analysis were done to study cells apoptosis and ARMC5 protein levels in each condition at three different times between 10 and 16 h of transfection. Preliminary results for these damaging mutants showed decreased markers of apoptosis by FACS assay in comparison with the wild-type and the “benign” variant. Interestingly, the protein levels of the p.Leu548Pro, p.Leu331Pro and p.Cys139Arg “damaging” variants were higher at the longest transfection time than wild-type ARMC5 protein, consistent with more cell death in ARMC5 wild-type expressing cells. Method must be applied to the newly identified missense variants.

Perspective: This approach will help to demonstrate the pathogenicity of the missense variants found in PBMAH patients. This demonstration would be important along with in silico analysis to improve genetic conselling for PBMAH patients and their relatives.