Deleterious single nucleotide polymorphisms of protein kinase R identified by the computational approach

The human protein kinase R (PKR) recognizes invading RNA viruses and mediates the antiviral immune response by phosphorylating the eukaryotic translation initiation factor 2α (eIF-2α), thus blocking protein translation in infected cells and thus preventing viral replication. The observation that individuals show different degrees of susceptibility to viral infections gives rise to the hypothesis that single nucleotide polymorphisms (SNPs) in the protein kinase R may alter the response to an infection. Using different available servers (e.g. SIFT, PROVEAN, Polyphen2, SNAP2, SNP&GOs, SNP-PhD, I-Mutant Suite), 14 SNPs were identified that were predicted to have deleterious effects on the protein kinase R. Five SNPs, namely D266Y, Y323D, I398 K, Y465C and Y472C, were selected for homology modeling and the generated models were investigated with regard to their secondary structure, residue fluctuations and eIF-2α binding. Analysis with computational tools POLYVIEW-MM, SAAPdap, SRIDE, CMView, elNémo, NMsim and PatchDock revealed structural changes in all mutants yielding a more stable structure at the cost of reduced flexibility (except Y465C) and less conformational freedom compared to the native protein. The conformational changes in the mutant protein structures and the displacement of functional residues from their strategic positions are predicted to affect the functionality of PKR, and consequently will affect the efficiency of the individual's antiviral immune response negatively. This study will aid the physicians in precision medicine field to tailor optimal treatment for the patients.