FIGURE 8. The effects of overexpression of C14orf166 proteins on influenza A WSN and DkPen virus RdRp activity. The HEK293 cells were grown in a 24-well plate and co-transfected with a 200 ng of pCHA-C14orf166 expression plasmid and influenza A virus mini-replicon plasmids, and the reporter enzyme activities were defined as previously described23.
DISCUSSION
The C14orf166 protein is referred to as RNA Transcription,Translation and Transport Factor (RTRAF) according to recent data. This protein is a nucleo-cytoplasmic protein having a binding activity to RNA molecules and the RNA polymerase II complex31. As a result of these activities, it has a positive regulatory effect on RNA Pol II transcription and is involved in processes related to RNA metabolism such as tRNA splicing as part of the tRNA splicing ligase complex. In addition, it is thought that it also functions in RNA transport. The C14orf166, a multifunctional protein with these properties, is associated with some cancer types. It has been suggested that the overexpression of C14orf166 has a role in the development of non-small-cell lung cancer (NSCLC), and it would be beneficial to consider it in the treatment of high-risk patients32. It has been reported that this protein is overexpressed in breast cancer33. The expression of both the C14orf166 mRNA and protein was also found to be upregulated in cervical cancer cell lines and tissues18. Therefore, it has been suggested that it can be used as a biomarker for the diagnosis and prognosis of some cancer types34.
It has been determined that the C14orf166 protein, which has important functions in cellular RNA metabolism, also participates in the replication/transcription processes of some viruses. The C14orf166 protein interacts with the hepatitis C virus core protein and upregulates the viral infection22. Another important virus type with which the C14orf166 protein is associated is the influenza A virus. It has been stated that this protein binds to the influenza A virus RbRP enzyme PA subunit and stimulates the virus replication21. Our results with the minireplicon models also showed that C14orf166 stimulates the influenza viral polymerases (Figure 8).
In this study, we demonstrated that the full-size and truncated C14orf166 proteins in different sizes interacted with the influenza virus PA protein in the Y2H assay using the PA protein as a bait and the HEK293 Matchmaker cDNA (Figures 1 and 2). Among these proteins, the smallest fragment that interacts with PA protein consists of 69 amino acid residues that form the carboxy-terminal end of the C14odf166 (AD-C69). Therefore, it was concluded that the carboxy-terminal end of C14orf166 is important and sufficient for interaction with the viral PA protein. In a previous report, it was shown that the C14orf166 protein interacts not only with PA alone but also with the influenza virus RdRP complex and co-localizes with viral ribonucleoproteins12. Thus, the C14orf166 protein is not thought to bind to the regions of PA that interact with PB2 and PB1 subunits. In the Y2H assay where we used the amino-terminal half of the viral PA protein as a bait, no C14orf166 protein could be detected among the host interactor proteins, which led us to conclude that the amino-terminal region of PA is most likely not a target for this protein. Therefore, we focused on the carboxy-terminal moiety of PA for binding of the C14orf166 protein in docking analyses. Although the most probable 3D structures were selected in terms of free energies among the AD-C14orf166 protein models interacting with PA, these models may not reflect the real structures of the proteins. Here, we focused on the C14orf166:PA interaction model rather than the 3D structures. The docking results of the PA with the predicted 3D structures of the C69 peptide (non-fused with GAL4-AD) and AD-C69 (fused with GAL4-AD) expressed in the yeast cells indicated that this peptide/protein binds to the ligand (peptide) recognition site6 located in the carboxy-terminal region of the PA. In docking models, it was observed that C69 and AD-C69 clustered in this region, mostly covering the 610th to 630th amino acid residues on the PA (Figures 5A, 6A, and 6C). The docking results of AD-C209:PA and AD-C244+4:PA proteins synthesized in yeast cells showed that the carboxy-terminal ends of these fusion proteins have a high affinity for the peptide binding site of the PA protein such as C69 and AD-C69 (Figure 7). One the docking models of AD-C244+4, which involves full-length C14orf166 and the PA protein, the prominent amino acids in the interaction areas of these two were evaluated with PyMol software and PDBsum algorithm. In this model, there is the possibility of forming three salt bridges and fourteen hydrogen bonds between AD-C244+4 and the PA protein (Figure 7 E). Although the amino acid residues in the interaction area of the PA protein show a distribution within the 450th to 700th amino acids, the intense interaction site is seen in the 610th to 630th amino acid positions. Huarte et al., as a result of their mapping study, suggested that amino acid residues in positions 493 to 512 and 557 to 574 on the PA protein are important for binding to C14orf16621. Although this region of PA agrees with our results, docking analysis data point to amino acids within the 610th to 630th position as the most intense interaction site in our models.
In conclusion, possible interaction models between influenza A virus PA and human C14orf166 protein were revealed by reconciling the experimental results of the Y2H assay carried out with the influenza A virus PA as a bait, existing literature information, and the results of in silico 3D prediction/protein-protein docking analysis. It is very difficult to accurately predict protein-protein interaction patterns similar to those in the natural environment with in silico tools ignoring the factors in the cellular environment. It was concluded that the data of the Y2H assay will make an important contribution to predicting the 3D structures and possible interaction models in a form close to its native structure. The interaction models presented here for human C14orf166 and viral PA protein will allow us to better understand the influenza A virus replication and virus-host relationship.