Structural and functional characterization of human DDX5 and its interaction with NS5B of hepatitis C virus
Hepatitis C was first recognized as a transfusion-associated liver disease not caused by hepatitis A or hepatitis B virus after serological tests were developed to screen for their presence in the blood. The infectious agent was finally identified with the cloning of the cDNA of hepatitis C virus (HCV) using random polymerase chain reaction (PCR) screening of nucleic acids extracted from plasma of a large pool of chimpanzee infected with non-A non-B hepatitis. NS5B, a membrane-associated RNA-dependent RNA polymerase essential in the replication of HCV, initiates the synthesis of a complementary negative-strand RNA from the genomic positive-strand RNA so that more positive-strand HCV RNA can then be generated from the newly synthesised negative-strand template. The crystal structure of NS5B presented typical fingers, palm and thumb sub-domains encircling the GDD active site, which is also seen in other RNA-dependent RNA polymerases, and is similar to the structure of reverse transcriptase of HIV-1 and murine Moloney leukaemia virus. The last 21 amino acids in the C-terminus of NS5B anchor the protein to the endoplasmic reticulum (ER)-derived membranous web. NS5B has been shown to interact with the core, NS3/NS4A, NS4B and NS5A proteins, either directly or indirectly. Numerous interactions with cellular proteins have also been reported. These proteins are mainly associated with genome replication, vesicular transport, protein kinase C-related kinase 2, P68 (DDX5), α-actinin, nucleolin, human eukaryotic initiation factor 4AII, and human VAMP-associated protein. Previous studies have confirmed that NS5B binds to full-length DDX5. By constructing deletion mutants of DDX5, we proceeded to characterize this interaction between DDX5 and HCV NS5B. We report here the identification of two exclusive HCV NS5B binding sites in DDX5, one in the N-terminal region of amino acids 1 to 384 and the other in the C-terminal region of amino acids 387 to 614. Proteins spanning different regions of DDX5 were expressed and purified for crystallization trials. The N-terminal region of DDX5 from amino acids 1 to 305 which contains the conserved domain I of the DEAD-box helicase was also cloned and expressed in Escherichia coli. The cloning, expression, purification and crystallization conditions are presented in this work. Subsequently, the crystal structure of DDX5 1-305 was solved and the high resolution three-dimensional structure shows that in front of domain I is the highly variable and disordered N terminal region (NTR) of which amino acids 51-78 is observable, but whose function is unknown. This region forms an extensive loop and supplements the core with an additional α-helix. Co-immunoprecipitation experiments demonstrated that the NTR of DDX5 1-305 auto-inhibit its interaction with NS5B. Interestingly, the α-helix in NTR is essential for this auto-inhibition and seems to mediate the interaction between the highly flexible 1-60 residues in NTR and NS5B binding site in DDX5 1-305, presumably located within residues 79-305. Furthermore, co-immunoprecipitation experiments revealed that DDX5 can also interact with other HCV proteins, besides NS5B.