FIGURE 2 . Different sizes of C14odf166 proteins in fusion with yeast GAL4-AD, encoded by C14orf166 cDNAs isolated from positive yeast colonies (A ), and reporter β-galactosidase activities measured in the yeast cells harboring these proteins along with the GAL4-BD. PA bait protein (B ). N , amino acid residues corresponding to the non-translated region of the gene.
The Y2H screening assay was also performed using a bait protein consisting of the amino-terminal moiety (359 amino acid residues) of influenza A virus PA fused to yeast GAL4-BD as bait (data not shown). Although many positive colonies were obtained on the SD selective medium, no colonies carrying C14orf166 cDNA were found among them, suggesting that the amino-terminal half of the PA protein was not a target for the C14orf166 protein.
Predicted models for viral PA protein and C14orf166 interaction
Protein-protein interactions allow complex metabolic events to take place harmoniously in the cells. Therefore, determining the interactions between proteins and the interaction models allows us to understand biological events. Deciphering the interrelationships of viral and cellular proteins is of vital importance for revealing the replication/transcriptional mechanisms of viruses, which depend on the host cells for all biosynthesis events, and developing strategies to cope with pathogenic viruses. However, it is quite difficult to predict the three-dimensional (3D) structures of proteins based on their primary structures and protein-protein interactions. Peptide bonds between amino acids are strong and highly rigid chemical bonds. In contrast, the bonds connecting the alpha carbon (α-carbon) of the amino acids to the carboxyl group (-COOH) and the amino group (-NH2) are very flexible, and the polypeptide can rotate freely around these two bonds30. The flexibility of these groups allows variable three-dimensional structures for a protein containing 20 different amino acids and makes predicting the folding patterns much more complicated, depending on the size of the protein. Therefore, the carboxy-terminal part of the C14orf166 protein covering 69 amino acid residues (C69) which interacts with the viral PA in the yeast cells was used as the basic structure for the prediction of the PA protein interaction patterns. De novo 3D fold models of the C69 peptide were generated using an online I-TASSER server. The top three models with the highest C-score are given in Figure 3. Despite bending in different directions, the most prominent structural features of the models are the alpha helix structure at the carboxy-terminal ends.