Abstract
Familial hypercholesterolemia (FH) is a metabolic disease caused by the inherited pathogenic mutations in PCSK9 gene. This study has evaluated the potential of diverse computational methods in predicting the PCSK9 clinical variants and also studied the impacts of mutation on protein phenotype and function. Our findings show the superior prediction ability of FATHMM over CADD, M-CAP, SIFT and Polyphen methods, in screening PCSK9 missense mutations causative to FH. Computational 3D mapping of PCSK9 variants located in prodomain (K83T), catalytic domain (D204N) and c-terminal domains (K494E) revealed mutant residue induced disturbances in the tertiary structure of protein. These variants are also predicted to impair the free energy dynamics, hence stability of PCKS9 as per the consensual predictions made by protein structure-based prediction algorithms like SAAFEC and MAESTRO. Molecular docking assay by PIZMA algorithm showed the increased binding affinity between PCSK9 variants and LDLR molecules. Furthermore, the results from PDBsum method have also suggested changes in interfacing residues, interface area, salt bridge and ionic bond interactions confirming the findings from docking analysis. The PCSK9 functional domain mutations could increase its binding affinity with LDLR, eventually promoting LDLR degradation in lysosomes and elevate circulating cholesterol levels in body. This study supports the application of comprehensive computational assessment of FH causative PCSK9 mutations before undertaking labor intensive functional biology investigations.