Phenotypic prediction in glutaric aciduria type 1 combining in silico and in vitro modeling with real-world data

Abstract

Glutaric aciduria type 1 (GA1) is caused by inherited deficiency of glutaryl-CoA dehydrogenase (GCDH). To further understand the unclear genotype–phenotype correlation, we transfected mutated GCDH into COS-7 cells resembling known biallelic GCDH variants of 47 individuals with GA1. In total, we modeled 36 genotypes with 32 missense variants. Spectrophotometry demonstrated an inverse correlation between residual enzyme activity and the urinary concentration of glutaric acid and 3-hydroxyglutaric acid, confirming previous studies (Pearson correlation, r = −0.34 and r = −0.49, p = 0.045 and p = 0.002, respectively). In silico modeling predicted high pathogenicity for all genotypes, which caused a low enzyme activity. Western blotting revealed a 2.6-times higher GCDH protein amount in patients with an acute encephalopathic crisis (t-test, p = 0.015), and high protein expression correlated with high in silico protein stability (Pearson correlation, r = −0.42, p = 0.011). The protein amount was not correlated with the enzyme activity (Pearson correlation, r = 0.09, p = 0.59). To further assess protein stability, proteolysis was performed, showing that the p.Arg88Cys variant stabilized a heterozygous less stable variant. We conclude that an integration of different data sources helps to predict the complex clinical phenotype in individuals with GA1.