Endocrine Abstracts

Nicotinamide nucleotide transhydrogenase (NNT) is an integral protein of the inner mitochondrial membrane and plays a major role in maintaining the redox balance by catalysing the trans-hydrogenation between NADH and NADP+ and proton translocation across the mitochondrial membrane.

Genetic variants in NNT have recently been reported in patients with familial glucocorticoid deficiency (FGD), combined mineralocorticoid and glucocorticoid deficiency and combined adrenal failure and testicular adrenal rest tumours. Moreover, knockout animal models suggest that NNT is involved in the pathogenesis of diabetes mellitus and obesity. Impaired NNT activity is also thought to be involved in the aging process and the development of neurological disorders and cancer.

In this study, we generated a 3D structural model of human NNT (H-NNT) by homology modelling using bacterial NNT as templates. We identified key structural and functional residues in H-NNT, such as those participating in NAD binding and in H-NNT homodimerization. Moreover, we mapped 14 amino acid substitutions causing adrenal disorders and 6 rare genetic variants reported in the ExAC database. This new model allowed us to demonstrate that deleterious variants affect H-NNT structure by altering its structure (p.Gly200Ser, p.Thr357Ala, p.Tyr388Ser, p.Pro437Leu, p.Ala533Val, p.Leu977Pro), its ability to dimerize (p.Phe215Ser, p.His365Pro), its ability to bind NAD (p.Ser193Asn) or NADP (p.Ala1008Pro and p.Asn1009Lys) or its ability to correctly fold within the mitochondrial inner membrane (p.Gly664Arg, p.Gly678Arg, p.Gly862Asp). Without the 3D H-NNT model, molecular mechanisms could only be identified for the two variants located in the NADP binding site.

In conclusion, availability of a 3D H-NNT model allowed us to decipher the mechanisms by which genetic variants causing adrenal disease affect NNT structure and function. Structural biology can provide valuable information on the structure-function relationship of proteins, and integration of genetic analysis with protein 3D modelling can greatly enhance prioritization and interpretation of human genetic variants.